Docstoc

Morgan Stanley - Local Markets Guidebook

Document Sample
Morgan Stanley - Local Markets Guidebook Powered By Docstoc
					RESEARCH




Emerging Markets
Local Markets Guidebook   March 2012
                                                      MORGAN STANLEY RESEARCH

                                                      March, 2012
                                                      EM Local Markets Guide




                                                                                     Key Contributors
EM Local Markets                                      Morgan Stanley Asia
                                                      Limited+
                                                                                     AXJ

Guidebook                                                                            Pieter Van Der Schaft
                                                                                     Pieter.Van.Der.Schaft@morganstanley.com
                                                                                     +852 3963 0550

This is the second edition of the Morgan Stanley
                                                                                     Kritika Kashyap
EM Local Markets Guidebook. In this Guidebook,                                       Kritika.Kashyap@morganstanley.com
we feature details of currency, bond and interest                                    +852 2239 7179

rate markets across AXJ, CEEMEA and Latin             Morgan Stanley & Co.           CEEMEA
America. Moreover, we present some of our latest      International plc+
                                                                                     Mihail Bozinov
work on EM local markets. We start with a review                                     Mihail.Bozinov@morganstanley.com
of the development of the asset class and include                                    +44 20 7677 1150

the outlook for bond supply this year, a                                             Meena Bassily
presentation of our risk premia models for Latin                                     Meena.Bassily@morganstanley.com
America, as well as an outline of our new                                            +44 20 7677 0031

S$NEER Model, among other features. We trust          Morgan Stanley & Co. LLC       LatAm
that you will find this reference valuable for your
                                                                                     Vitali Meschoulam
investment decision-making process, and we look                                      Vitali.Meschoulam@morganstanley.com
forward to hearing from you about what content                                       +(1) 212 761 1889

and improvements we can bring to the next                                            Sian Griffiths
edition.                                                                             Sian.Griffiths@morganstanley.com
                                                                                     +(1) 212 761 1884


Rashique Rahman
Head of EM Macro Strategy
Morgan Stanley & Co. International plc+

EM Strategy Team
March 29, 2012




                                                        Morgan Stanley does and seeks to do business with
                                                        companies covered in Morgan Stanley Research. As a
                                                        result, investors should be aware that the firm may have a
                                                        conflict of interest that could affect the objectivity of Morgan
                                                        Stanley Research. Investors should consider Morgan
                                                        Stanley Research as only a single factor in making their
                                                        investment decision.
                                                        For analyst certification and other important
                                                        disclosures, refer to the Disclosure Section,
                                                        located at the end of this report.
                                                        += Analysts employed by non-U.S. affiliates are not registered with FINRA, may not be associated
                                                        persons of the member and may not be subject to NASD/NYSE restrictions on communications with a
                                                        subject company, public appearances and trading securities held by a research analyst account.
                                                                               MORGAN STANLEY RESEARCH


                                                                               March, 2012
                                                                               EM Local Markets Guide




Table of Contents
Introduction to EM Local Markets
Changing Patterns in EM Rates ...................................................................................................... 3
2012 Government Financing Outlook ............................................................................................ 18
Asia Markets.................................................................................................................................. 32
CEEMEA Markets.......................................................................................................................... 63
LatAm Markets .............................................................................................................................. 85
Product and Market Focus........................................................................................................... 103
EM as an Asset Class since 1994 ............................................................................................... 104
EM Flows – Should I Stay or Should I Go?.................................................................................. 109
Valuing Latin America Rates ....................................................................................................... 111
The Risk Premia in CEEMEA ...................................................................................................... 117
A Guide to Latin America Interest Rate Options .......................................................................... 126
Hiking Cycles and Term Spreads in Latin America...................................................................... 136
Hiking Cycles and Flattening Trends in CEEMEA ....................................................................... 139
How to Value EM Breakeven Inflation? ....................................................................................... 145
SGD: The New S$NEER Model .................................................................................................. 148
A Primer in the EM FX Probability Analyzer ................................................................................ 153
Country Reference Sheets .......................................................................................................... 164




Morgan Stanley and its affiliates do not render advice on tax and tax accounting matters to clients. This material was not intended or written to be
used, and it cannot be used or relied upon by any recipient, for any purpose, including the purpose of avoiding penalties that may be imposed on
the taxpayer under US federal tax laws or the tax laws of any other countries discussed herein. Each client should consult his/her personal tax
and/or legal advisor to learn about any potential tax or other implications that may result from acting on a particular recommendation.



                                                                                                                                                       2
                                                                                    MORGAN STANLEY RESEARCH

                                                                                    March, 2012
                                                                                    EM Local Markets Guide




Changing Patterns in EM Rates
Vitali Meschoulam, Sian Griffiths, Rashique Rahman
                                 th
First Published February 17 , 2012: EM rates (R)evolution, Part 1 - Changing Patterns in EM Rates

This report reflects one in a series of investigations into the                              proxy – has been a significant driver of this change in
changing nature of EM local markets as an asset class. In                                    market dynamics for EM local rates.
this installment, we examine the changing nature of the
                                                                                            Second, for the period since 2009, we note that funding
trading dynamics of EM local rates and bonds. We find there
                                                                                             market stress has been a crucial driver linked to
has been a clear evolution in the manner in which EM rates
                                                                                             changes in the trading behavior of EM rates. Our
have traded since 2009.
                                                                                             analysis showed that although EM rates are trading
Prior to 2009, EM rates traded as a risky asset class – yields                               much more like DM since 2009, this behavior still breaks
increased as risk markets (i.e., equities) sold off and EM                                   down in times of elevated funding market stress.
currencies depreciated.
                                                                                     The development of the local debt markets made it possible
This contrasts to trading behavior of core rate markets such                         for EM rates to trade increasingly as defensive assets, but
as US Treasuries or German Bunds, where a decrease in                                the level of funding stress acts as an ‘on and off’ switch that
risk appetite tends to be reflected in a sell-off in risky assets.                   determines if rates will indeed trade as defensive or risky
                                                                                     assets at any given point in time. 2
Generally speaking, all the markets we looked at in this
report 1 exhibit the same general pattern. Pre-2009, EM rates                        Moreover, those markets that exhibit higher external
performed in line with other risky assets, selling off as                            vulnerability – those with lower external coverage ratios 3 –
broader risk markets sold off.                                                       tend still to trade as risk markets, with the rate markets
                                                                                     exhibiting high correlation with the local currency, such as
Since 2009, however, this relationship has changed – EM
                                                                                     Hungary and Turkey 4.
rates have been performing more like DM rates – yields
decrease as risk markets sell off and EM currencies                                  Exhibit 2
depreciate.                                                                          Rising Domestic Debt in Emerging Markets
Exhibit 1                                                                                10000     Brazil           Mexico
3m Rates, Equity and FX Performance in Brazil                                            9000      Turkey           Poland
     50%                                                                                 8000
                                                                                                   South Africa     Korea
     40%                                                                                 7000
     30%                                                                                 6000
     20%                                                                                 5000
     10%                                                                                 4000
      0%                                                                                 3000
    -10%                                                                                 2000
    -20%                                                                                 1000
    -30%                                                                                    0
    -40%            BRL 3m %                 2y Swap 3m %         Local Equity                   2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
    -50%                                                                             Source: BIS, Morgan Stanley Research
       Apr-07     Jan-08     Oct-08      Jul-09      Apr-10    Jan-11      Oct-11    The size of the local debt market – what we consider a proxy
Source: Bloomberg, Morgan Stanley Research                                           for its level of development – we find is a statistically
We see two contributing factors to this change in behavior for                       significant contributor to the evolution of EM rates markets.
EM local rate markets.

       First, empirical analysis reveals that the development of                    2
                                                                                       We found that high levels of funding stress (defined as our EM Funding
        the local debt market – using the size of the market as a                      Stress Index at a level greater than 20) are associated with a breakdown in
                                                                                       the co-movement between EM and DM rates.
                                                                                     3
                                                                                       We define the external coverage ratio as the ratio of external funding
                                                                                       requirements to central bank liquid FX reserves.
1                                                                                    4
    We examined the performance of Brazil, Mexico, South Africa, Poland,               We have explored these aspects in EM Profile: Channels of Financial Market
    Turkey                                                                             Contagion – Funding Risks, Aug 11, 2011).


                                                                                                                                                                3
                                                                        MORGAN STANLEY RESEARCH

                                                                        March, 2012
                                                                        EM Local Markets Guide




Manifestations of the development of the market are                      More specifically, (1) an expansion in the investor base;
improved liquidity and an expansion in the investor base,                (2) improved domestic economic stability and management;
both of which may have contributed to this change in regime              and (3) a ‘trial by fire’ in 2008 all likely contributed in the
since 2009.                                                              evolution of the EM rates market.

Related to the development of the market are fundamental                 We investigate these aspects in turn.
drivers such as more robust fiscal and economic policy, the
                                                                         Exhibit 4
institutionalization of inflation targeting regimes and more
                                                                         Size of EM Domestic Bond Market (US$ million)
freely floating exchange rates.
                                                                          7000
We think all of these features have contributed to the regime
                                                                          6000
shift for local markets in recent years (Exhibit 3).                                                        CEEMEA
                                                                          5000                              LatAm
Exhibit 3
                                                                          4000                              AXJ
EM Rates Have ‘Co-moved’ with DM More Since
                                                                                                            EM
2009                                                                      3000

         % time spent in 'region of comovement'                           2000
                         Pre 2009     Since 2009                          1000
Brazil                        39%           81%                               0
Mexico                        62%           86%
                                                                             95

                                                                                      96

                                                                                               98

                                                                                                         99

                                                                                                                  01

                                                                                                                          02

                                                                                                                                    04

                                                                                                                                            05

                                                                                                                                                     07

                                                                                                                                                            08

                                                                                                                                                                      10

                                                                                                                                                                               11
Turkey                        52%           32%
                                                                           1

                                                                                   3

                                                                                             1

                                                                                                      3

                                                                                                               1

                                                                                                                        3

                                                                                                                                   1

                                                                                                                                          3

                                                                                                                                                    1

                                                                                                                                                           3

                                                                                                                                                                    1

                                                                                                                                                                            3
                                                                          Q

                                                                                  Q

                                                                                            Q

                                                                                                     Q

                                                                                                              Q

                                                                                                                       Q

                                                                                                                                  Q

                                                                                                                                         Q

                                                                                                                                                   Q

                                                                                                                                                          Q

                                                                                                                                                                   Q

                                                                                                                                                                           Q
South Africa                  63%           82%                          The data for Asia includes the Local Currency Government Bonds only, i.e., Treasury
                                                                         Bonds for China, Treasury Bonds and Recaps for Indonesia, Government Bonds for India,
Poland                        68%           91%                          Treasury Bonds and Central Bank Bills for Korea, Government Securities and Islamic
Source: Bloomberg, Morgan Stanley Research                               Issues for Malaysia, Government bonds for Singapore, Thailand and Taiwan.
                                                                         Source: BIS, Morgan Stanley Research
Indeed, the financial crisis itself was a crucial ‘trial by fire’ for
EM rates. Investors observed how quickly EM economies                    Larger Local Bond Market, Improved Access and
shrugged off the effects of the crisis, and how strong their             Tradability
relative financial positions were compared to DM. The                    Based on EMTA, trading volumes for EM local bonds now
improved depth and resiliency of the market also likely                  outpace those of external debt (see Exhibit 5).
favorably altered perceptions of the asset class.
                                                                         Exhibit 5
This change in sentiment – which is observable in falling risk           Local Debt Trades in Greater Volumes than
premia – was, we believe, an important driver of the change              External
in EM rates behavior since that time.
                                                                            1,600

Slow Transformation                                                         1,400
                                                                                                    Local Debt
                                                                            1,200
The role of an expanding local bond market in positively                                            External Debt
affecting a change in the behavior of EM rates is consistent                1,000

with a number of additional features. Improved access and                     800
tradability (a greater number of instruments and fewer                        600
investment restrictions such as capital controls) likely led to
                                                                              400
increased attractiveness of EM local debt markets as an
                                                                              200
investment destination – especially considering robust
growth and declining inflation dynamics for emerging                              0

economies.
                                                                                                       3

                                                                                                              4

                                                                                                                    5

                                                                                                                              6

                                                                                                                                    6

                                                                                                                                         7

                                                                                                                                                 8
                                                                                           2

                                                                                                 3




                                                                                                                                                        9

                                                                                                                                                               9

                                                                                                                                                                      0

                                                                                                                                                                           1
                                                                                  1



                                                                                               '0




                                                                                                                          '0

                                                                                                                                  '0




                                                                                                                                                            '0

                                                                                                                                                                   '1
                                                                                                                                               '0




                                                                                                                                                                          '1
                                                                               '0

                                                                                       '0



                                                                                                      '0

                                                                                                            '0

                                                                                                                    '0




                                                                                                                                         '0



                                                                                                                                                     '0
                                                                              3

                                                                                       2

                                                                                             1

                                                                                                     4

                                                                                                            3

                                                                                                                  2

                                                                                                                          1

                                                                                                                                4

                                                                                                                                        3

                                                                                                                                               2

                                                                                                                                                     1

                                                                                                                                                           4

                                                                                                                                                                  3

                                                                                                                                                                         2
                                                                                      Q

                                                                                            Q

                                                                                                    Q

                                                                                                           Q

                                                                                                                 Q

                                                                                                                         Q

                                                                                                                               Q

                                                                                                                                       Q

                                                                                                                                              Q

                                                                                                                                                    Q

                                                                                                                                                          Q

                                                                                                                                                                 Q

                                                                                                                                                                        Q
                                                                             Q




The benefits associated with the improved tradability of a               Source: EMTA

larger local bond market are, we find, important in the                  Greater trading volumes tend to have a salutary effect on
evolution of EM rates towards a defensive asset class.                   volatility (i.e., the market is better equipped to handle larger
                                                                         flows). Exhibit 6 shows a clear correlation between trading
                                                                         volume in local bonds and the yield volatility. For example,


                                                                                                                                                                                4
                                                                                MORGAN STANLEY RESEARCH

                                                                                March, 2012
                                                                                EM Local Markets Guide




during the financial crisis in 2008, volume plummeted and                        The growth in the number of funds – and assets under
yield volatility (measured as the 90-day realized volatility of                  management (AUM) for index-oriented local bond funds –
an aggregate index of EM bonds) increased substantially.                         has been dramatic (Exhibit 7).
Financial crisis aside, the chart shows that even small                          Exhibit 7
reductions in trading volume are associated with an increase                     EM Local Currency-Bond Funds (USDmn, AUM)
in yield volatility.
                                                                                  60,000
Exhibit 6
                                                                                  50,000
Yield Volatility Spikes as Trading Volume
Decreases                                                                         40,000

  3                                                                     1,500
                                                                                  30,000
  5
                                                                        1,300
  7                                                                               20,000
  9
                                                                        1,100
 11                                                                               10,000
 13                                                                     900
                                                                                        0
 15
                                                                                                2006     2007          2008            2009       2010       2011
                                                                        700
 17
                                                                                 Source: Bloomberg, Morgan Stanley Research
 19
                                                                        500      These investors manage against an index, with their relevant
 21
 23                                                                     300
                                                                                 market exposure indicated by their allocation relative to their
   Mar-05      Mar-06    Mar-07      Mar-08       Mar-09       Mar-10            benchmark. This in practice means that there is less need for
                   EM Loca Bond Trading Volumes (RHS)                            these investors to buy or sell as much of their holdings for a
                   EM Aggregate Bond Yield Volatility (LHS inverted)
                                                                                 given allocation change.
Source: Bloomberg, Morgan Stanley Research
A potential explanation for this relationship is that as the                     Exhibit 8
depth of the market increased, bond markets were better                          Falling Yield Volatility Associated with Growth in
equipped to deal with potential inflows and outflows, allowing                   EM Local Index Funds (USDmn, AUM)
for less ‘panicky’ price action that had led to high correlations                 25                                                                              170000
with risk markets in the past.
                                                                                                                 Yield Volatility
                                                                                                                                                                  150000
Furthermore, because these markets are increasingly less                          20
                                                                                                                 Index Oriented Fund AUM
                                                                                                                                                                  130000
prone to swings, a virtuous cycle exists as investors do not
feel the need to exit en masse when sentiment turns                               15                                                                              110000
negative.
                                                                                  10                                                                              90000
Expansion in the Investor Base
                                                                                                                                                                  70000

Coinciding with this broad development of the market, EM                           5
                                                                                                                                                                  50000
local rates have attracted an increasing number of foreign
                                                                                   0                                                                              30000
investors.
                                                                                   Jun-08    Jan-09     Aug-09      Mar-10          Oct-10    May-11     Dec-11
Independent of funds drawn by the higher yields on offer in                      Source: EPFR
EM compared to core-market rates, our analysis suggests                          Exhibit 8 suggests that this growth in the indexed investor
that diversified portfolios should have 30% allocated to EM                      base is associated with a dampening in volatility for EM rate
local assets (currency unhedged) based on the historical                         markets and less correlation with broader risk markets.
risk/return profile of the asset class (see EMQSU: EM as an                      Particularly since the 2008 financial crisis, increases in the
Asset Class since 1994, June 23, 2011).                                          AUM of index funds coincides with falling EM yield volatility.
The increase in foreign investor participation in local rates                    The increasing proportion of foreign investors may also play
markets also reflects the rise of index-oriented investors                       a further role in changing the trading behavior of EM rates.
participating in the market.                                                     The World Bank suggests that where sound macroeconomic




                                                                                                                                                                      5
                                                                                                                                                          MORGAN STANLEY RESEARCH

                                                                                                                                                          March, 2012
                                                                                                                                                          EM Local Markets Guide




and financial policies are in place, foreign investment can                                                                                                Exhibit 10
catalyze the development of domestic debt markets 5.                                                                                                       Pension Fund Assets Have Grown over the Past 10
                                                                                                                                                           Yrs
Foreign investment can increase the depth, breadth and
liquidity of domestic markets, while enhancing their efficiency                                                                                                800
                                                                                                                                                                       US$ billion
through the development of financial instruments, portfolio                                                                                                    700
diversification and the promotion of international standards
                                                                                                                                                               600
(see EM Profile: The Case for EM Local Markets, April 8,
                                                                                                                                                               500
2011).
                                                                                                                                                               400
Some argue that because most foreign investors hail from
                                                                                                                                                               300
developed markets where they are used to trading rates as
defensive assets, they provide a positive influence on the                                                                                                     200

local investor.                                                                                                                                                100

Exhibit 9                                                                                                                                                       0
                                                                                                                                                                      2001      2002   2003     2004    2005   2006    2007   2008     2009
Foreign Ownership of Local Bonds
                                                                                                                                                                     Chile                       Czech Republic           Hungary
                                                                                                                                                                     Israel                      Korea                    Mexico
    45%                                                                                                                                                              Poland                      Turkey                   South Africa
    40%                                       Brazil                                           Mexico                                                      Source: OECD, Morgan Stanley Research

    35%                                       Poland                                           Turkey                                                      Lastly, compared to pre-financial-crisis levels, arguably there
    30%
                                                                                                                                                           is less pure duration available in the market given that
    25%                                                                                                                                                    Peripheral Europe – Greece, Portugal, Ireland and to a
    20%                                                                                                                                                    lesser extent Italy and Spain – now trades more akin to
    15%                                                                                                                                                    sovereign credit than interest rates.
    10%                                                                                                                                                    In the context of this scarcity of rate markets, EM is a good
    5%
                                                                                                                                                           substitute. This dynamic has supported the growth of
    0%
                                                                                                                                                           external investors.
                   Mar-07




                                              Mar-08




                                                                         Mar-09




                                                                                                    Mar-10




                                                                                                                               Mar-11
          Nov-06




                                     Nov-07




                                                                Nov-08




                                                                                           Nov-09




                                                                                                                      Nov-10




                                                                                                                                                 Nov-11
                            Jul-07




                                                       Jul-08




                                                                                  Jul-09




                                                                                                             Jul-10




                                                                                                                                        Jul-11




                                                                                                                                                           Solid Macroeconomic Management
Note: Brazil uses a different methodology, using percent of total outstanding and not                                                                      Beyond the advantages already discussed, a well-developed
percent of specific bonds (such as NTN-Fs, LTNs) as in the case of the other countries.
Using a similar measure, foreign ownership in Brazil would be closer to Mexico’s.                                                                          local bond market has secondary benefits that improve the
Source Morgan Stanley Research, Brazil National Treasury, Mexico Treasury, Central Bank
of Turkey, Central Bank of Poland                                                                                                                          ability of policymakers to manage the economy 6.
In addition to growth in the external investor base, over the                                                                                              Exhibit 11
past decade, financial sector and pension system reform and
                                                                                                                                                           Inflation Falling in All Regions
investment by domestic institutional investors have swelled
                                                                                                                                                               20
the pool of investible local cash.                                                                                                                                    CPI, %Y
                                                                                                                                                               18
These EM pension funds tend to invest the majority of their                                                                                                    16
                                                                                                                                                               14
assets in government bonds – and pension fund assets have
                                                                                                                                                               12
grown significantly in EM economies (see Exhibit 10).                                                                                                          10
These institutional investors, constrained by mandate, do not                                                                                                   8
                                                                                                                                                                6
withdraw their capital from their country during times of
                                                                                                                                                                4
stress. Instead, if anything, they may tend to shift their                                                                                                      2
allocations away from domestic equities towards domestic                                                                                                        0
bonds.                                                                                                                                                               1999       2001     2003        2005     2007       2009      2011
                                                                                                                                                                       EM                CEEMEA                  AXJ               Latam
                                                                                                                                                           Source: Bloomberg, Morgan Stanley Research



5                                                                                                                                                          6
    ‘Global Development Finance: Mobilizing Finance and Managing                                                                                               See EM Profile: The Case for EM Local Markets, April 8, 2011) for a more
    Vulnerability’ The World Bank, 2005, p78.                                                                                                                  detailed discussion of this topic.


                                                                                                                                                                                                                                           6
                                                                                                        MORGAN STANLEY RESEARCH

                                                                                                        March, 2012
                                                                                                        EM Local Markets Guide




A well-developed liquid bond market can be used as a tool to                                             This improved fiscal soundness associated with a deeper
target inflation, manage shocks and help guide consumption                                               bond market helps improve macroeconomic stability, which
and investment cycles. By managing the issuance and                                                      in turn has increased the prospects for local rates to trade
repurchase of sovereign debt, a country can increase or                                                  more like their DM counterparts – as a defensive asset.
decrease the amount of liquidity in the financial system. A                                              Exhibit 12 suggests a positive correlation between the risk
country can use this mechanism to inject liquidity when it                                               level associated with sovereign bonds (risk level is derived
wishes to promote investment and spending, and remove it                                                 from the S&P credit rating) and the volatility of the yield on
to tame inflation and overheated growth.                                                                 the bonds.
This improved economic management is reflected in                                                        2008 Financial Crisis – A ‘Trial by Fire’ for EM
improving fundamentals in EM, including a long-term
                                                                                                         We believe that the change seen since 2009 indicates a
downtrend in inflation (Exhibit 11) and a steady improvement
                                                                                                         change in sentiment with respect to EM rates after the
in creditworthiness in recent years, resulting in rising
                                                                                                         markets and economies in EM had been seriously tested
sovereign credit ratings (Exhibit 13).
                                                                                                         during the financial crisis.
Exhibit 12
                                                                                                         The Lehman crisis that brought many DM economies to their
Local Bond Yield Volatility Positively Correlated
                                                                                                         knees, and whose effects are still weighing on growth, left
with Local Currency Sovereign Credit Rating
                                                                                                         most EM economies relatively unscathed. Economies such
   40                                                                                                    as Brazil saw one quarter of negative growth and then
                                                                              Hungary
   35                                                                                                    bounced back rapidly.
   30                                                                                                    After the crisis, analysts observed that using metrics like debt
   25                                                                        Turkey                      burden, budget financing needs and current account
   20
                                                   Mexico Russia                                         balance, EM was in much better shape than its DM
                                       Poland                                                            counterparts.
   15                       South Africa                 Peru
                                                         Thailand                    Indonesia
                           Korea                                                                         It was this trial by fire in 2008 that changed the way in which
   10
                  Israel                                                  India                          investors view EM rates – as defensive rather than risky
                                                   Brazil                            Phillipines
    5
         Chile              China
                                        Malaysia
                                                            Colombia                                     assets. This change is evident in the decrease in risk premia
    0                                                                                                    since 2009.
     AA+         AA    AA-         A+      A        A-       BBB+      BBB    BBB-      BB+        BB
Source: Bloomberg, Morgan Stanley Research.                                                              After spiking dramatically higher at the heights of the crisis,
Exhibit 13                                                                                               the risk premia quickly decreased and has been dropping
EM Hard Currency Sovereign Credit Ratings                                                                ever since (Exhibit 14). The decreasing risk premia
 BBB
                                                                                                         (compared to increasing risk premia in many DM countries),
                                                                                                         reflects this improving investor sentiment towards EM rates.

 BBB-                                                                                                    Exhibit 14
                                                                                                         Risk Premia Has Decreased Since 2009

  BB+
                                                                                                           7
                                                                       SCRM forecast

   BB
                                                                                                           5
                                                                                                                                                      Brazil
  BB-
        '96 '97 '98 '99 '00 '01 '02 '03 '04 '05 '06 '07 '08 '09 '10 '11 '12
                                                                                                           3
SCRM = Sovereign Credit Rating Model, which forecasts sovereign credit ratings (EM
Profile: Sovereign Credit - Sensitiity to Macro Fundamentals, June 6th 2011
Source: Morgan Stanley Research.
                                                                                                           1
                                                                                                          Nov-08                 Oct-09      Sep-10            Aug-11
                                                                                                         Source: Morgan Stanley Research


                                                                                                                                                                          7
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




Since 2009 – Funding Market Stress as a Driver                                         Exhibit 16
                                                                                       Percentage of EM Currency Selloffs Where Yields
Since 2009, we find that the level of funding market stress is
                                                                                       Increased
the most significant determinant of whether EM rates trade
as a risky or defensive asset.                                                                                                   pre 2009   post 2009
                                                                                       Brazil                                        85%         10%
EM funding market stress – as measured by our EM Funding                               Mexico                                        50%         36%
Stress Index – has predictive power on a three-month                                   Turkey                                       100%          0%
forward-looking basis for EM currencies and EM credit                                  South Africa                                  89%         53%
spreads (see EM Quantitative Strategy: EM Rates - EM or                                Poland                                        72%         14%
Rates?, Jan 5, 2012 and EM Profile: Monitoring Funding                                 Source: Bloomberg, Morgan Stanley Research.

Markets - EM Funding Stress Index, Dec 6, 2011).                                       Since 2009, however, a sell-off in currencies was more
                                                                                       associated with a decrease in yields.
After the 2008 financial crisis, EM rates showed a greater
tendency to trade like defensive assets unless funding                                 This is consistent with the evolution of EM rates to a more
market stress was high.                                                                defensive asset – a sell-off in the local currency reflects a
                                                                                       movement out of risky assets, and the subsequent decrease
Exhibit 15                                                                             in EM yields indicates capital reallocating into these
Z-Score Difference and EM Funding Stress Index                                         defensive assets, driving yields down.
(6m)
                                                                                       But even here the relationship is imperfect and may be more
    300                                                                          45
                                                                                       about magnitudes – material depreciation of the currency is
    200
                                                                                 40    ultimately likely to lead to higher rates, as prospects for lower
    100
      0
                                                                                       policy rates are undermined.
                                                                                 35
 -100                                                                                  However, unlike with the behavior of EM rates as a risky
 -200                                                                                  versus defensive assets, funding stress does not appear to
                                                                                 30
 -300
                                                                                       catalyze a breakdown in rates and FX behavior since 2009.
 -400                                                                            25
 -500                                                                                  Although there are some incidents in Mexico and Poland of
 -600                                                                            20    EM yields breaking a trend of decreasing during an FX sell-
        1-    15-   29-    12-     26-   10-    24-       7-   21-    5-   19-
       Aug    Aug   Aug    Sep     Sep   Oct    Oct      Nov   Nov   Dec   Dec
                                                                                       off during times of high funding stress (see Exhibit 45 and
          Mexico          Poland          South Africa          EMFSI (RHS)            Exhibit 47), this is a limited occurrence.
Source: Bloomberg, Morgan Stanley Research
In times of escalating funding market stress, EM rates return                          Perhaps one explanation for increasing divergence between
to trading like risky assets (Exhibit 15).                                             EM currency and EM rate performance can be gleaned by
                                                                                       the episode of global market volatility in 2H11.
EM Rates and EM Currencies                                                             An intensification of the European crisis eventually led to a
Not only have EM rates tended to trade in a more mature                                significant bout of EM currency depreciation (versus the
fashion over time, the relationship between EM rates and EM                            USD). But notably, the weakness in EM currencies was not
currencies also seems to have changed since 2009.                                      accompanied by a traditional widening of interest rates – at
                                                                                       least not in the same magnitude that would have been
Using the methodology described in EM Profile: Hedging                                 expected given past performance.
Against A European Selloff, Jan 24, 2012, we analyzed the
performance of EM rates during selloffs in the currency 7.                             Indeed what we saw this time around was hedging of long
                                                                                       local bond positions. Investors held their bond positions
Prior to the 2009 inflection point (discussed earlier), EM                             whilst they bought USD (sold EMFX) as a means of hedging
yields tended to widen as EM currencies sold off in the                                their exposure.
majority of cases (Exhibit 16).
                                                                                       A potential hypothesis about why this has been the case
                                                                                       effectively deals with some of the same factors that helped
7                                                                                      EM rates trade as more defensive assets.
    For our purposes, we defined selloffs as periods of 2-65 trading days when
    the cumulative loss is greater than 95% of the same trading days in the
    sample. See Exhibit 45-Exhibit 50 for full tables.


                                                                                                                                                        8
                                                                 MORGAN STANLEY RESEARCH

                                                                 March, 2012
                                                                 EM Local Markets Guide




That is, deeper, more liquid markets, much more robust            Assuming, as we do, that we have not seen the last of
fundamentals and more importantly an expectation that             funding stress due to ongoing potential disruptions arising
these positive trends were likely to continue.                    from the ongoing European malaise, we would look to treat
                                                                  EM rates as positively correlated with funding costs. That is,
Strategic Implications
                                                                  in times of high funding stress, we would expect (and
Since 2009, our analysis has shown that funding stress is         position for) a bear-steepening of local rates curves as risk
positively correlated with yield. In other words, EM rates        premia increases. For further analysis on the behavior of EM
trade as risky assets when funding stress is high (when our       Rates Curves, please see Emerging Markets Quantitative
Funding Stress Index rises above 20).                             Quarterly, Sep 2010.
This is likely because a high level of funding stress reflects
capital constraints and risk aversion in the system. These
conditions mean that investors face an increasing need (and
desire) to be in cash or highly liquid and defensive assets
such as UST. This combination of risk aversion and liquidity
constraints drives an increase in EM yields. As such, we look
for the stress index to help inform our investment
recommendations in local rates space in EM.




                                                                                                                              9
                                                                             MORGAN STANLEY RESEARCH

                                                                             March, 2012
                                                                             EM Local Markets Guide




Appendix: Statistical Analysis                                                Exhibit 18
                                                                              Rolling Z-Scores for Brazil and US Rates
Having visually observed a change in the trading behavior of                    300
                                                                                                                                         Brazil 5Y Z
EM rates, we used Principal Components Analysis (PCA) to                        200                                                      US 5Y Z
investigate the possible drivers of the change in behavior for
EM rates.                                                                       100

                                                                                  0
We split the data into separate time periods – pre-financial
crisis, post-financial crisis and the whole set from 2007 to                   -100
2012 and conducted PCA on all sections.
                                                                               -200
PCA analysis identifies factors that drive the patterns in a
                                                                               -300
data set – in this case EM rates and US Treasury (UST)
yields. PCA does not, however, identify what those factors                     -400
are, just their relative importance in the co-movement of                         Dec-05      Dec-06       Dec-07      Dec-08   Dec-09    Dec-10
assets.                                                                       Source: Bloomberg, Morgan Stanley Research.
                                                                              We used rolling 6-month Z-scores to reveal more clearly the
From 2006 to 2008, four factors were needed to explain over
                                                                              change in trading behavior of EM rates. As discussed
90% of the variation. However, from 2009 to 2012 this has
                                                                              previously in EM Quantitative Strategy Update: EM Rates –
been replaced by two factors. This change implies that
                                                                              EM or Rates?, Jan 5, 2012, the 6-month rolling Z-scores
idiosyncrasy has been reduced and the trading behavior of
                                                                              normalize the data and reduce the noise.
EM and DM rates has become more similar.
                                                                              As can be seen in the example of Brazil below, there is a
PCA also suggests there had been a significant change in
                                                                              clear increase in co-movement between the two Z-scores
how the first Principal Component (PC1) affected the EM and
                                                                              since 2009. Please see Exhibit 23 to 27 for the rolling EM
UST rates market.
                                                                              Z-scores vs UST for other countries sampled.
Exhibit 17                                                                    An even clearer measure is achieved by taking a simple
PCA for 2006-2008 and 2009-2012                                               difference between two Z-scores. In EM Quantitative
             2006-2008                                  2009-2012             Strategy Update: EM Rates – EM or Rates?, Jan 5, 2012, we
 10Y YLD          PC 1         PC 2           10Y YLD      PC 1     PC 2      proposed that -150 to 150 can be used as the region of co-
ZAR                 0.42         0.51        ZAR            0.44     -0.12    movement for two series which on average are positively
UST                -0.38         0.14        UST            0.44     -0.15    correlated.
TRY                 0.35        -0.36        TRY            0.38     -0.59
POL                 0.36         0.64        POL            0.38      0.50    If the difference in Z-scores multiplied by 100 is within
MXN                 0.44        -0.20        MXN            0.45     -0.14    [-150,150], they are moving together. If the difference is less
BRL                 0.48        -0.37        BRL            0.36      0.59    than -150 or more than 150, it suggests a possibility of a
Source: Bloomberg, Morgan Stanley Research
                                                                              regime shift.
The sensitivity of US rates to PC1 is the same in the post-
financial crisis data set as it is for EM rates, but opposite for             The difference in Z-scores confirmed this change in trading
pre-financial crisis data set (Exhibit 17). Whatever it is that               pattern and illustrated this change much more clearly. All the
drives global rates (PC1) had a profound change in late 2008                  EM markets sampled showed an increase in co-movement
in how it affect EM rates vis-a-vis US rates. This same                       with UST since 2009 (see Exhibit 19, Exhibit 20, and Exhibit
change can be seen in correlation matrix (Exhibit 51 and 53).                 27 to Exhibit 29)
In light of this change highlighted by PCA, we then                           Using the example of Mexican rates (Exhibit 19), since 2009
conducted regression and Z-score analysis to dig deeper into                  the difference in Z-scores spends most of the time (more
the possible drivers of the change in rates’ trading behavior,                than 80%, to be precise) between values of -150 and 150. A
and what PC1 might be.                                                        similar pattern can be seen for all other EM countries
                                                                              sampled.




                                                                                                                                                       10
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




In Brazil for example, the difference between Brazil and US                            We chose these three separate periods to investigate what
Z-scores starts to decline and become less volatile after                              drove the change in regime for EM rates, but also to
2009, spending the majority of time in this range.                                     determine if the drivers changed in significance before and
                                                                                       after this ‘inflection point’ in 2009.
After a brief period of again trading more like risky assets in
August 2011, Brazilian rates have again tended to trade                                Exhibit 30 to Exhibit 44 show a summary of the regressions
along more developed lines (please see Exhibit 20).                                    for the five sample countries across these different date
                                                                                       ranges.
Exhibit 19
Z-Score Difference Between Mexico and US Rates                                         As we initially expected, the size of the local debt market was
  750
                                                                                       a powerful driver, significant for most countries across all the
                                                                                       date ranges. The change in funding stress (measured by the
  600
                                                                                       level of our index) was also a powerful driver, however much
  450
                                                                                       more important since 2009.
  300
                                                                                       That said, the data seem to suggest that the importance of
  150
                                                                                       any specific driver depends on the time period analyzed.
    0
                                                                                       Before 2009, the size of the local debt market seems to be a
 -150                                                                                  powerful variable and is statistically significant for Brazil,
 -300                                                                                  Turkey, Poland and South Africa.
 -450
                                                                                       After the inflection point in 2009, our Funding Stress Index
 -600                                                                                  becomes a more important driver across the board –
    Nov-06     Sep-07       Jul-08     May-09      Mar-10     Jan-11      Nov-11
                                                                                       significant for all countries. Furthermore, the dummy variable
Source: Bloomberg, Morgan Stanley Research.
                                                                                       – essentially a binary factor that turns ‘on’ when the funding
Exhibit 20
                                                                                       environment is challenging and off when benign – was more
Z-Score Difference Between Brazil and US Rates                                         significant after 2009 than at any time before.
  750
                                                                                       To ascertain sensitivity (as opposed to significance) of the
  600
                                                                                       regressor (3m change in co-movement) to the separate
  450
                                                                                       factors used in the regression, we repeated the regression
  300
                                                                                       using the Z-Scores for Funding Stress, Fund Flows, Size of
  150                                                                                  Local Debt markets and the Dummy Variable.
    0

 -150
                                                                                       Across the board (for all countries and time periods) the
                                                                                       initial level of EM rates was the variable with the highest
 -300
                                                                                       sensitivity. Funding stress and local debt markets were
 -450
                                                                                       around the same 0.40 level of sensitivity when analyzing the
 -600
                                                                                       whole data set, but as with significance, the regressor was
    Nov-06       Sep-07       Jul-08      May-09     Mar-10      Jan-11      Nov-11
                                                                                       more sensitive to local markets pre-2009. Funding stress
Source: Bloomberg, Morgan Stanley Research.
                                                                                       was low across the board (around 0.15).
The rolling Z-scores show a definite change in how EM rates
traded after the end of the 2008/2009 financial crisis. To test                        These results suggest that the size of the local bond market
for possible drivers of this structural change, we regressed                           has been an important driver in the evolution of EM rates
the size of the local debt market, fund flows and our EM                               performance from a risky asset to a more defensive asset –
Funding Stress Index (EMFSI) against the six-month change                              at least early on in the development of these markets.
in the difference between EM and US Z-scores – i.e., the 3m
                                                                                       However after the financial crisis in 2008/2009, or perhaps
change in co-movement.
                                                                                       even because of it, funding stress has become a more
We split the sample data into three time periods – from Nov                            important determinant of co-movement between EM and DM
2006 to Dec 2011, Nov 2006 to Dec 2008 and Feb 2009 to                                 rates. This makes intuitive sense as well. In a world in which
December 2011.                                                                         global markets are well-functioning, growing local markets
                                                                                       and a new source of more stable inflows into these markets


                                                                                                                                                   11
                                                                      MORGAN STANLEY RESEARCH

                                                                      March, 2012
                                                                      EM Local Markets Guide




had a material impact in creating a more traditional                   Exhibit 21
(defensive) asset class.                                               Z-Score Difference and EM Funding Stress Index
                                                                       (6m)
Following the financial woes of 2008/2009, this trajectory
                                                                        300                                                                                    45
toward defensive performance for EM rates was inevitably
                                                                        200
put into question as heightened funding risks showed the
                                                                        100                                                                                    40
potential weakness inherent in EM rates trading in spite of
                                                                           0
the progress to date.                                                                                                                                          35
                                                                        -100
With funding risks again increasing since August 2011 on the            -200
                                                                                                                                                               30
back of ongoing European travails, EM rates are no longer               -300

consistently trading as defensive assets and are again                  -400                                                                                   25
increasingly linked to global risk appetite.                            -500
                                                                        -600                                                                                   20
Finally – and contrary to what we had originally expected –                   1-      15-    29-     12-     26-    10-    24-       7-    21-     5-    19-
                                                                             Aug      Aug    Aug     Sep     Sep    Oct    Oct      Nov    Nov    Dec    Dec
fund flows, likely due to data limitations, had a limited impact                 Mexico             Poland           South Africa           EMFSI (RHS)
on explaining the change of regime in EM rates.                        Source: Bloomberg, Morgan Stanley Research
                                                                       Exhibit 22

Appendix: The Evolving Role of                                         Z-Score Difference and EM Funding Stress Index
                                                                       (18m)
Funding Stress                                                          300                                                                                    45
                                                                        200
                                                                                                                                                               40
Across the whole data set, funding stress was not as                    100
powerful as domestic debt market size as a predictor for the               0                                                                                   35

regime change in EM rates.                                              -100
                                                                                                                                                               30
                                                                        -200
Prior to the financial crisis in 2008, it had very little impact on     -300                                                                                   25
whether EM rates traded as EM or DM. However as we point                -400
                                                                                                                                                               20
out in EM Quantitative Strategy: EM Rates – EM or Rates?,               -500
Jan 5, 2012, funding stress is the most important predictor of          -600                                                                                   15
                                                                               Jul-   Oct-   Jan-    Apr-    Jul-   Oct-   Jan-     Apr-   Jul-   Oct-
whether EM rates trade as risky assets or DM rates in the                      09      09     10      10     10     10      11       11    11      11
period of time since 2009.                                                       Mexico             Poland           South Africa           EMFSI (RHS)

                                                                       Source. Bloomberg, Morgan Stanley Research
Indeed we have shown that our EM Funding Stress Index                  We present the statistical evidence in Exhibit 30 to Exhibit
has predictive power on 3m ahead EMFX and EM credit                    44, where we use weekly data since November 10, 2006, to
spreads (EM Profile: Monitoring Funding Markets – EM                   regress 3m ahead change in our Z-score measure on its
Funding Stress Index, Dec 7, 2011). In particular, the                 current value, the past 3m change in EMFSI, and a dummy
relationship was negative so that increased funding stress             indicating if the current EMFSI is above 20.
today reduced future EM asset returns. Moreover, a dummy
variable on whether the current EMFSI was above 20,                    The hypothesis that the elevated level of EMFSI has
(indicating elevated funding market stress) was also highly            statistically significant predictive power at the 5% level is
significant.                                                           confirmed in three countries: Brazil, Mexico and Poland. In
                                                                       addition, it is statistically significant at the 10% level in all
Exhibit 21 displays the sample countries’ Z-score differences          countries. These results suggest that during the last two
against EMFSI since August 1, 2011. The co-movement                    years, EM rates have been trading more and more like
between the four variables is at least visually striking. In           defensive assets. However, if the funding stress is elevated
addition, in Exhibit 22 we show the same four variables since          enough, EM rates return to trading more like risky assets.
July 2009.                                                             Our EM Funding Stress Index, at least recently, seems to be
The increase of EMFSI above the earlier mentioned level of             an important driver determining if a breakdown in this new
20 seems to predict an increase in our Z-score measure. In             regime should occur.
other words, it seems to predict EM rates underperforming
relative to US rates or EM rates trading as risk not as rates.

                                                                                                                                                           12
                                                                                       MORGAN STANLEY RESEARCH

                                                                                       March, 2012
                                                                                       EM Local Markets Guide




Appendix
Exhibit 23                                                                              Exhibit 26
Rolling Z Scores for Mexico and US Rates                                                Poland Rates Trading Behaviour
  400                                                                                     300
                                                                Mexico 5Y Z
                                                                                                                                                         Poland 5Y Z
  300                                                           US 5Y Z                   200                                                            US 5Y Z

  200
                                                                                          100
  100
                                                                                            0
    0
                                                                                         -100
 -100
                                                                                         -200
 -200

 -300                                                                                    -300

 -400                                                                                    -400
    Dec-05      Dec-06       Dec-07      Dec-08        Dec-09      Dec-10                   Jan-05 Oct-05   Jul-06   Apr-07 Jan-08 Oct-08   Jul-09   Apr-10 Jan-11 Oct-11

Source: Bloomberg, Morgan Stanley Research.                                             Source: Bloomberg, Morgan Stanley Research.
Exhibit 24                                                                              Exhibit 27
Rolling Z Scores for Turkey and US Rates                                                Z-Score Difference Between Turkey and US Rates
                                                                                          750
    500
                                                                   Turkey 5Y Z
                                                                                          600
    400                                                            US 5Y Z
                                                                                          450
    300

    200
                                                                                          300

    100
                                                                                          150

        0
                                                                                            0

   -100                                                                                  -150

   -200                                                                                  -300

   -300                                                                                  -450

   -400                                                                                  -600
      Jun-06   Mar-07    Dec-07    Sep-08     Jun-09    Mar-10     Dec-10     Sep-11        Nov-06      Sep-07       Jul-08     May-09      Mar-10      Jan-11     Nov-11
Source: Bloomberg, Morgan Stanley Research.                                             Source: Bloomberg, Morgan Research
Exhibit 25                                                                              Exhibit 28
South Africa Rates Trading Behaviour                                                    Z-Score Difference Between S.Africa and US Rates
                                                                                          750
    400
                                                                  S Africa 5Y Z
                                                                                          600
    300                                                           US 5Y Z
                                                                                          450
    200
                                                                                          300
    100
                                                                                          150
        0                                                                                   0
   -100                                                                                  -150

   -200                                                                                  -300

   -300                                                                                  -450
                                                                                         -600
   -400
      Jan-05 Oct-05 Jul-06 Apr-07 Jan-08 Oct-08 Jul-09 Apr-10 Jan-11 Oct-11                 Nov-06      Sep-07       Jul-08      May-09     Mar-10      Jan-11     Nov-11

Source: Bloomberg, Morgan Stanley Research                                              Source: Bloomberg, Morgan Stanley Research




                                                                                                                                                                       13
                                                                                   MORGAN STANLEY RESEARCH

                                                                                   March, 2012
                                                                                   EM Local Markets Guide




Exhibit 29                                                                          Exhibit 34
Z-Score Difference Between Poland and US Rates                                      Poland: Regressing z(t-3)-z(t). R2 = 0.49
  750                                                                               Variable                           Coefficient Std Error t-statistic
                                                                                    Constant                               238.56      48.64       4.91
  600                                                                               Z (t)                                    -0.94      0.06     -14.81
  450                                                                               EMFSI (t)-EMFSI (t-3m)                    1.74      0.51       3.41
                                                                                    Dummy (t)                               22.07      17.48       1.26
  300
                                                                                    Local Debt (t-3m)                        -1.22      0.27      -4.48
  150                                                                               Fund Flows (t-3m)                         0.27      0.29       0.96
                                                                                    Source: Morgan Stanley research.
    0
 -150                                                                               Regression Results 2007–2008
 -300
                                                                                    Exhibit 35
 -450
                                                                                    Brazil: Regressing z(t+3m)-z(t). R2 =0.54
 -600
                                                                                    Variable                           Coefficient Std Error   t-statistic
    Nov-06      Sep-07       Jul-08     May-09      Mar-10       Jan-11   Nov-11
                                                                                    Constant                             -1026.82    288.89         -3.55
Source: Bloomberg. Morgan Stanley Research                                          Z (t)                                    -1.01      0.15        -6.83
                                                                                    EMFSI (t)-EMFSI (t-3m)                    9.82      3.42         2.87
Regression Results 2007 – 2012                                                      Dummy (t)                              -73.36      99.56        -0.74
                                                                                    Local Debt (t-3m)                         1.74      0.49         3.58
                                                                                    Fund Flows (t-3m)                        -0.01      0.33        -0.02
Exhibit 30                                                                          Source: Morgan Stanley Research
Brazil: Regressing z(t+3m)-z(t). R2 = 0.49                                          Exhibit 36
Variable                                     Coefficient Std Error t-statistic      Mexico: Regressing z(t-3)-z(t). R2 = 0.62
Constant                                        -125.84    103.37       -1.22
                                                                                    Variable                           Coefficient Std Error   t-statistic
Z (t)                                              -1.02      0.08     -13.19
                                                                                    Constant                             -1319.06    335.22         -3.93
EMFSI (t)-EMFSI (t-3m)                              1.35      0.95       1.42
                                                                                    Z (t)                                    -1.20      0.14        -8.44
Dummy (t)                                         87.65      34.17       2.57
                                                                                    EMFSI (t)-EMFSI (t-3m)                  10.00       3.80         2.63
Local Debt (t-3m)                                   0.18      0.12       1.51
                                                                                    Dummy (t)                              -90.12      69.08        -1.30
Fund Flows (t-3m)                                   0.14      0.16       0.89       Local Debt (t-3m)                         7.51      1.89         3.98
Source: Morgan Stanley Research
                                                                                    Fund Flows (t-3m)                         0.33      0.62         0.53
Exhibit 31                                                                          Source: Morgan Stanley Research
Mexico: Regressing z(t-3)-z(t). R2 =0.54                                            Exhibit 37
Variable                                     Coefficient Std Error t-statistic      Turkey: Regressing z(t-3)-z(t). R2 = 0.67
Constant                                        -192.86      95.49      -2.02       Variable                           Coefficient Std Error   t-statistic
Z (t)                                              -1.19      0.08     -14.97       Constant                             -3206.69    719.02         -4.46
EMFSI (t)-EMFSI (t-3m)                              2.86      0.91       3.14       Z (t)                                    -1.01      0.12        -8.37
Dummy (t)                                         72.82      28.95       2.52       EMFSI (t)-EMFSI (t-3m)                    3.81      2.73         1.40
Local Debt (t-3m)                                   0.84      0.39       2.15       Dummy (t)                              -85.87    115.95         -0.74
Fund Flows (t-3m)                                  -0.23      0.20      -1.19       Local Debt (t-3m)                       15.72       3.75         4.19
Source: Morgan Stanley Research                                                     Fund Flows (t-3m)                        -0.23      0.58        -0.40
Exhibit 32                                                                          Source: Morgan Stanley Research.

Turkey: Regressing z(t-3)-z(t). R2 = 0.53                                           Exhibit 38

Variable                                     Coefficient Std Error t-statistic      South Africa: Regressing z(t-3)-z(t). R2 = 0.66
Constant                                       -1118.83    196.92       -5.68       Variable                           Coefficient Std Error   t-statistic
Z (t)                                              -1.07      0.05     -20.01       Constant                             -1185.12    202.54         -5.85
EMFSI (t)-EMFSI (t-3m)                              0.58      0.87       0.67       Z (t)                                    -0.90      0.09       -10.59
Dummy (t)                                        123.16      28.10       4.38       EMFSI (t)-EMFSI (t-3m)                    1.59      1.77         0.90
Local Debt (t-3m)                                   4.95      0.91       5.47       Dummy (t)                              -25.86      26.54        -0.97
Fund Flows (t-3m)                                  -0.43      0.36      -1.20       Local Debt (t-3m)                       19.50       3.17         6.15
Source: Morgan Stanley Research                                                     Fund Flows (t-3m)                        -3.61      2.14        -1.68
                                                                                    Source: Morgan Stanley Research.
Exhibit 33
                                                             2
South Africa: Regressing z(t-3)-z(t). R = 0.45
Variable                                     Coefficient Std Error t-statistic
Constant                                         -36.95      36.38      -1.02
Z (t)                                              -0.87      0.06     -14.54
EMFSI (t)-EMFSI (t-3m)                              1.14      0.49       2.34
Dummy (t)                                         32.39      22.04       1.47
Local Debt (t-3m)                                   0.78      0.33       2.35
Fund Flows (t-3m)                                  -1.01      0.36      -2.80
Source: Morgan Stanley Research.


                                                                                                                                                     14
                                                                           MORGAN STANLEY RESEARCH

                                                                           March, 2012
                                                                           EM Local Markets Guide




Exhibit 39                                                                  Exhibit 44
Poland: Regressing z(t-3)-z(t). R2 = 0.59                                   Poland: Regressing z(t-3)-z(t). R2 = 0.62
Variable                           Coefficient Std Error     t-statistic    Variable                                     Coefficient   Std Error   t-statistic
Constant                               923.56    162.03            5.70     Constant                                          13.93        71.53         0.19
Z (t)                                    -1.00      0.13          -7.97     Z (t)                                              -1.22        0.06       -19.86
EMFSI (t)-EMFSI (t-3m)                    0.96      2.61           0.37     EMFSI (t)-EMFSI (t-3m)                              0.96        0.42         2.26
Dummy (t)                              189.91      65.49           2.90     Dummy (t)                                         40.80        24.98         1.63
Local Debt (t-3m)                        -6.04      1.19          -5.09     Local Debt (t-3m)                                  -0.10        0.33        -0.29
                                                                            Fund Flows (t-3m)                                   0.34        0.28         1.22
Fund Flows (t-3m)                        -0.64      1.06          -0.60
                                                                            Source: Morgan Stanley Research
Source: Morgan Stanley Research.


Regression Results 2009-2012                                                Exhibit 45
                                                                            BRL Selloffs and 5Y Rates Behaviour
Exhibit 40
                                                                                      Begin                End                  Loss           5Y Swap
Brazil: Regressing z(t+3m)-z(t). R2 = 0.60
                                                                                 10-Aug-05           24-Aug-05                -7.59%               4.4%
Variable                           Coefficient   Std Error   t-statistic
Constant                                  4.31       85.06         0.05          11-Nov-05           19-Dec-05                -9.77%              -2.9%
Z (t)                                    -1.25        0.06       -20.68            3-Mar-06            8-Mar-06               -3.30%              -8.1%
EMFSI (t)-EMFSI (t-3m)                    0.53        0.68         0.78
Dummy (t)                              -50.97        37.96        -1.34
                                                                                 16-Mar-06           28-Mar-06                -5.92%              -4.0%
Local Debt (t-3m)                         0.03        0.10         0.36           5-May-06           23-May-06               -13.73%              -3.5%
Fund Flows (t-3m)                         0.23        0.13         1.75
Source: Morgan Stanley Research
                                                                                 26-May-06           30-May-06                -3.15%              -4.5%
                                                                                   23-Jul-07         16-Aug-07               -13.02%              -8.5%
Exhibit 41                                                                       14-Nov-07           27-Nov-07                -5.83%              -0.8%
Mexico: Regressing z(t-3)-z(t). R2 = 0.70                                        14-Jan-08           21-Jan-08                -5.46%              -1.7%
Variable                           Coefficient   Std Error   t-statistic         18-Mar-08           20-Mar-08                -2.57%              -0.9%
Constant                              -473.53        87.96        -5.38          21-Aug-08           21-Nov-08               -45.43%              -9.0%
Z (t)                                    -1.25        0.07       -16.73
EMFSI (t)-EMFSI (t-3m)                    1.65        0.65         2.52
                                                                                 26-Nov-08             8-Dec-08              -12.48%             12.0%
Dummy (t)                               59.52        27.05         2.20            6-Jan-09            2-Mar-09              -12.20%               3.8%
Local Debt (t-3m)                         1.93        0.35         5.59
Fund Flows (t-3m)                        -0.25        0.12        -2.06
                                                                                 11-Jun-09           22-Jun-09                -5.55%               0.8%
Source: Morgan Stanley Research                                                  13-Aug-09           17-Aug-09                -3.10%               0.7%
                                                                                  15-Oct-09           20-Oct-09               -3.13%              -2.0%
Exhibit 42
                                                                                  23-Oct-09           28-Oct-09               -3.57%               0.3%
Turkey: Regressing z(t-3)-z(t). R2 = 0.57                                          4-Jan-10          29-Jan-10                -9.75%               0.9%
Variable                           Coefficient   Std Error   t-statistic
Constant                              -161.19      262.14         -0.61
                                                                                   2-Feb-10            4-Feb-10               -2.56%               0.3%
Z (t)                                    -0.99        0.06       -16.91           3-May-10           20-May-10                -8.92%               1.0%
EMFSI (t)-EMFSI (t-3m)                    3.39        1.27         2.66            2-Jun-10            7-Jun-10               -3.40%               0.9%
Dummy (t)                              -64.15        43.41        -1.48
Local Debt (t-3m)                         0.83        1.20         0.69            26-Jul-11         22-Sep-11               -21.79%               5.1%
Fund Flows (t-3m)                        -1.31        0.46        -2.83           28-Oct-11          24-Nov-11               -12.84%               2.6%
Source: Morgan Stanley Research.
                                                                                   5-Dec-11          14-Dec-11                -5.19%               0.3%
                                                                            Source: Bloomberg, Morgan Stanley Research
Exhibit 43
South Africa: Regressing z(t-3)-z(t). R2 = 0.57
Variable                           Coefficient   Std Error   t-statistic
Constant                              -191.06        52.30        -3.65
Z (t)                                    -1.10        0.08       -13.64
EMFSI (t)-EMFSI (t-3m)                    0.95        0.59         1.62
Dummy (t)                               76.26        30.34         2.51
Local Debt (t-3m)                         1.78        0.41         4.40
Fund Flows (t-3m)                        -0.32        0.36        -0.88
Source: Morgan Stanley Research




                                                                                                                                                         15
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Exhibit 46                                                          Exhibit 48
TRY Selloffs and 5Y Rates Behaviour                                 KRW Selloffs and 5Y Rates Behaviour
         Begin                 End               Loss   5Y Swap              Begin                  End               Loss    5Y Swap
      3-Mar-06             8-Mar-06            -3.85%      -1.3%          9-Aug-07            17-Aug-07             -2.96%       0.94%
      4-May-06            23-Jun-06           -26.80%     -32.9%         28-Feb-08             8-May-08            -11.57%       0.46%
     20-Sep-06            22-Sep-06            -3.20%      -4.5%          28-Jul-08           28-Oct-08            -39.66%       6.79%
     26-Feb-07             1-Mar-07            -4.00%      -2.0%         30-Oct-08            24-Nov-08            -19.56%       0.37%
     27-Apr-07             1-May-07            -3.25%      -4.3%         30-Dec-08             2-Mar-09            -22.49%      -0.02%
      24-Jul-07           16-Aug-07           -11.56%     -10.2%          3-Jun-09             13-Jul-09            -6.63%      -0.15%
     14-Jan-08            21-Jan-08            -5.34%      -2.6%         15-Oct-09            22-Oct-09             -2.97%      -0.46%
      4-Feb-08            11-Feb-08            -5.34%      -2.0%         25-Nov-09            27-Nov-09             -1.90%       0.72%
     28-Feb-08            31-Mar-08           -11.51%      -5.3%         14-Jan-10             1-Feb-10             -4.28%       0.25%
      4-Aug-08            22-Oct-08           -40.45%      -3.9%          4-Feb-10             8-Feb-10             -1.82%       0.05%
      4-Nov-08            19-Nov-08           -15.73%      -2.9%         26-Apr-10            10-Jun-10            -12.92%       1.77%
     18-Dec-08             9-Mar-09           -18.44%       6.8%         21-Jun-10               2-Jul-10           -4.78%       0.80%
     26-Apr-10             6-May-10            -7.91%       0.1%          9-Aug-10            12-Aug-10             -2.24%       0.83%
      28-Jul-11            8-Aug-11            -6.11%       7.0%         11-Nov-10            26-Nov-10             -4.65%       0.41%
Source: Bloomberg, Morgan Stanley Research.
                                                                          8-Feb-11            11-Feb-11             -2.14%       0.14%
Exhibit 47                                                                27-Jul-11             3-Oct-11           -13.29%       3.17%
MXN Selloffs and 5Y Rates Behaviour                                       4-Nov-11            25-Nov-11             -4.74%       0.63%
                                                                          7-Dec-11            19-Dec-11             -4.27%       0.25%
          Begin                End               Loss   5Y Swap     Source: Bloomberg, Morgan Stanley Research.
     10-May-06           13-Jun-06             -5.66%      -2.8%
       28-Jul-06           1-Aug-06            -1.61%      -0.8%    Exhibit 49
       6-Nov-07          12-Nov-07             -2.31%      -1.0%    ZAR Selloffs and 5Y Rates Behaviour
       4-Aug-08           22-Oct-08           -35.41%      -8.5%             Begin                End                Loss    5Y Swap
       4-Nov-08          20-Nov-08            -10.93%      -2.1%          9-Aug-06            3-Oct-06            -15.94%        0.8%
     16-Dec-08             9-Mar-09           -17.98%       0.5%          2-Jan-07            9-Jan-07             -5.66%       -0.4%
     19-May-09             9-Jun-09            -5.00%      -2.9%        23-Feb-07             5-Mar-07             -6.03%       -0.8%
        1-Jul-09           14-Jul-09           -4.70%       0.6%          23-Jul-07         16-Aug-07              -9.27%       -0.7%
     21-Aug-09             1-Oct-09            -7.09%       0.3%          8-Nov-07          27-Nov-07              -7.30%       -1.6%
       3-Dec-09            8-Dec-09            -2.25%      -0.2%        10-Dec-07           20-Dec-07              -5.89%       -0.7%
     19-Jan-10             8-Feb-10            -4.52%       1.0%        14-Jan-08           21-Mar-08             -19.90%       -1.2%
      26-Apr-10          20-May-10             -8.06%       0.2%        16-May-08           12-Jun-08              -8.71%       -2.6%
     18-Jun-10              2-Jul-10           -4.50%       0.8%          1-Aug-08           22-Oct-08            -49.61%        1.6%
       13-Jul-10           16-Jul-10           -2.06%       0.9%          5-Jan-09            5-Mar-09            -14.17%       -0.3%
       4-Aug-10          31-Aug-10             -5.34%       1.8%        30-Jun-09             14-Jul-09            -6.93%       -0.6%
     19-Nov-10           23-Nov-10             -1.77%      -0.4%          31-Jul-09           6-Aug-09             -3.87%       -0.3%
     26-Jan-11           28-Jan-11             -1.74%      -0.1%         14-Oct-09            2-Nov-09             -9.71%        0.1%
     14-Mar-11           16-Mar-11             -2.13%       0.3%        14-Dec-09           21-Dec-09              -4.46%        0.2%
        7-Jul-11         22-Sep-11            -20.39%       4.3%          2-Feb-10            5-Feb-10             -4.13%       -0.3%
      28-Oct-11          25-Nov-11             -9.21%      -2.2%         26-Apr-10          20-May-10              -8.18%       -1.5%
Source: Bloomberg, Morgan Stanley Research                                3-Jan-11            3-Feb-11             -9.26%       -4.4%
.                                                                       14-Mar-11           17-Mar-11              -4.23%       -0.4%
                                                                         29-Apr-11          13-May-11              -6.70%       -0.5%
                                                                           7-Jul-11         22-Sep-11             -24.47%        4.4%
                                                                         14-Oct-11           20-Oct-11             -4.46%        0.1%
                                                                         27-Oct-11          23-Nov-11             -10.85%       -2.0%
                                                                          7-Dec-11          14-Dec-11              -5.24%       -1.5%
                                                                    Source: Bloomberg, Morgan Stanley Research.




                                                                                                                                    16
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Exhibit 50                                                         Exhibit 51
PLN Selloffs and 5Y Rates Behaviour                                Correlation Matrix 2009-2012
          Begin                End              Loss   5Y Swap                                                10 YEAR YIELD
       2-Mar-06          13-Mar-06            -4.13%      -1.1%                          ZAR         UST      TRY   POL     MXN      BRL
                                                                   ZAR                           1
     17-Mar-06           23-Mar-06            -3.25%      -0.6%
                                                                   UST                       -0.44     1.00
     11-May-06           23-Jun-06            -9.24%      -3.6%
                                                                   TRY                        0.35    -0.26     1.00
       2-Jan-07          11-Jan-07            -4.16%      -0.1%    POL                        0.87    -0.28     0.16   1.00
       8-Aug-07          16-Aug-07            -4.78%      -0.6%    MXN                        0.41    -0.42     0.45   0.41   1.00
     10-Dec-07           17-Dec-07            -4.03%      -0.9%    BRL                        0.42    -0.70     0.64   0.28   0.81     1.00
     14-Jan-08           21-Jan-08            -5.08%       0.5%    Source: Morgan Stanley Research

      22-Apr-08           1-May-08            -4.81%      -0.1%
       21-Jul-08           23-Jul-08          -2.33%      -0.5%
       28-Jul-08          27-Oct-08          -42.60%      -0.8%
       4-Nov-08          20-Nov-08           -13.11%       2.7%
     17-Dec-08           17-Feb-09           -32.09%      -3.7%
     25-May-09           28-May-09            -3.09%      -0.3%
       2-Jun-09          15-Jun-09            -4.84%      -0.2%
      21-Oct-09           28-Oct-09           -5.05%      -0.9%
     25-Nov-09           21-Dec-09            -8.25%      -0.4%
     14-Jan-10             8-Feb-10           -7.96%       0.3%
     16-Feb-10           18-Feb-10            -2.50%       0.0%
     16-Mar-10             4-Jun-10          -22.51%       2.2%
     18-Jun-10           29-Jun-10            -4.91%       0.3%
      14-Oct-10           19-Oct-10           -4.34%      -0.6%
      25-Oct-10           27-Oct-10           -2.27%      -0.2%
       4-Nov-10          29-Nov-10           -12.77%      -1.0%
       3-Dec-10            7-Dec-10           -2.36%      -0.1%
     26-Jan-11           28-Jan-11            -2.42%      -0.6%
       9-Feb-11          14-Feb-11            -2.85%       0.2%
     14-Mar-11           16-Mar-11            -2.47%       0.7%
      2-May-11           23-May-11            -6.06%       1.2%
       7-Jun-11          16-Jun-11            -4.54%       0.2%
        4-Jul-11           3-Oct-11          -22.12%       3.4%
      27-Oct-11            6-Jan-12          -15.97%       0.2%
Source: Bloomberg, Morgan Stanley Research




                                                                                                                                           17
                                                                       MORGAN STANLEY RESEARCH

                                                                       March, 2012
                                                                       EM Local Markets Guide




2012 Government Financing Outlook
Pieter Van Der Schaft, Kritika Kashyap, Mihail Bozinov, Vitali Meschoulam
                        th
First published January 30 , 2012: 2012 Government Financing Outlook

   Asia ex-Japan: Bond issuance expectations vary across              Exhibit 52
    the region. Among the major issuers, we expect net                 2012 Net Domestic Borrowing as % of GDP
    issuance to decline in Korea and the Philippines, remain            6.0
    roughly stable in Malaysia and Indonesia, and increase in
    India, Taiwan and Thailand.                                         5.0

                                                                        4.0
   CEEMEA: Net domestic bond issuance forecasts for
                                                                        3.0
    2012 diverge in CEEMEA as well. Issuance should
    decline sharply in Czech Republic and slightly in Poland            2.0
    and South Africa, remain unchanged in Hungary and
                                                                        1.0
    increase in Turkey and Russia. Some issuers such as
    Poland (external) and South Africa (domestic and                    0.0
    external) have opted to draw on their cash reserves                       TH MY     IN   SA CZ MEX SG HU RU TW KO PD PH              ID   TU HK
    rather than increase issuance.
                                                                                                    2012 Net domestic borrow ing / GDP

                                                                       Source: Morgan Stanley Research Estimates, IMF.
    In all countries, however, we think that issuance forecasts
    will likely be challenged if growth turns out to be lower
                                                                       Exhibit 53
    than government projections. Most bond markets have
    had a good start in 2012, but we think that a sustainable          2012 Net Issuance as % of Outstanding Bonds*
    rally will require a durable improvement in euro-zone               6.0

    sovereign and funding markets, and in some cases, the
                                                                        5.0
    return of international investors.
                                                                        4.0

   LatAm: Expectations of net domestic bond issuance are               3.0
    fairly similar in both Mexico and Brazil. Both countries are
    seeking to increase the maturity profile of their debt,             2.0

    emphasizing the issuance of longer-term fixed and                   1.0
    interest-rate-linked securities.
                                                                        0.0
                                                                              IN    TH MY SA BR CZ MX SG HU RU TW KO PD PH               ID TU HK
    Issuance in Brazil should be flat to 2011 (specific
    guidance has not yet been given); however, Mexico’s                                             2012 Net domestic borrow ing / GDP
    2012 plan outlines a slight increase in borrowing by               Source: Morgan Stanley Research Estimates. *Market value
    MXN40bn. We expect that pension growth in the region
    should continue to provide demand for local debt but see
    upwards pressure on issuance guidance if growth and
    commodity exports slow.




                                                                                                                                                 18
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




2012 Government Financing Outlook
We present our 2012 Government Financing Outlook,                 Exhibit 54
covering 16 countries across three EM regions.                    Domestic Bond Issuance by EM Region, bn
                                                                                              2011                                  2012
Domestic bond issuance forecasts for 2012 diverge within                                   Gross        Net        Gross      Chg           Net     Chg
                                                                  Asia ex-Japan
and across regions, mainly due to differences in growth            Hong Kong                  16         -1           11 -33%                1     -250%
prospects and fiscal strategies.                                   India                   5,100      4,360        6,046 19%             5,140       18%
                                                                   Indonesia             211,000 126,500         240,000 14%           134,300        6%
                                                                   Korea                  77,300    30,600        79,800   3%           25,000      -18%
In Asia, issuance declines in Korea and the Philippines, as        Malaysia                   94         49           93 -1%                45       -9%
those countries continue to implement their fiscal                 Philippines               546        261          530 -3%               190      -27%
                                                                   Singapore                  16          4           21 36%                 8       93%
consolidation strategies, and increases in India, on the back      Taiwan                    565        225          686 21%               288       28%
                                                                   Thailand                  450        378          525 17%               475       26%
of a slowdown in growth, and in Taiwan and Thailand due to        CEEMEA
higher government expenditure.                                     Czech Republic            181         77           168 -8%                 34   -55%
                                                                   Hungary                 1,765        574         1,230 -30%               566    -1%
                                                                   Poland                     98         23           106   9%                20   -11%
In CEEMEA, issuance drops sharply in the Czech Republic            Russia                     na        939         1,809    na            1,209    29%
and slightly in Poland as those countries work on improving        South Africa              155        140           151 -3%                120   -14%
                                                                   Turkey                    114         17           102 -11%                20    20%
their fiscal position, but rises in Turkey, where GDP growth is   LatAm
                                                                   Mexico                    547        349           358 -35%              158    -55%
expected to slowdown sharply, and in Russia, where the             Brazil                    576        172           475 -18%              130    -24%
government will again attempt to boost expenditure and run        Source: Morgan Stanley Research Estimates.

a deficit.                                                        Exhibit 55
                                                                  2012 Net Domestic Borrowing as % of GDP
Issuance is mixed in LatAm as well. In Mexico, a sharp
                                                                    6.0
decline in net bond issuance is almost entirely offset by a
large increase in bill issuance.                                    5.0

The largest borrowers relative to the size of the economy can       4.0
be found in Asia, with Thailand, India, and Malaysia, where
                                                                    3.0
domestic borrowing is around 5-6% of GDP. South Africa is
also running a relatively large deficit, with net domestic          2.0

borrowing at 4.4% of GDP. In most other countries, net              1.0
borrowing is between 1.5% and 2.5% of GDP.
                                                                    0.0
The bond markets that should see the largest increase in                  IN   TH MY SA BR CZ MX SG HU RU TW KO PD PH                       ID TU HK
size in 2012 are Russia (36%) and Indonesia (27%), as the
                                                                                               2012 Net domestic borrow ing / GDP
markets there are relatively small to begin with (less than
                                                                  Source: Morgan Stanley Research Estimates.
10% of GDP). The domestic bond markets of Thailand
                                                                  Exhibit 56
(21%), India (20%), and South Africa (14%) should also
expand considerably in 2012 on the back of large deficits. In
                                                                  2012 Net Issuance as % of Outstanding Bonds*
most countries, bond markets will increase by less than 10%.        40

                                                                    35
                                                                    30
                                                                    25
                                                                    20
                                                                    15
                                                                    10
                                                                     5
                                                                     0
                                                                          RU ID   IN TH SA SG MY MX PH HU KO BR TW TU PD CZ HK

                                                                                  2012 Net domestic bond issuance / Total domestic bond market

                                                                  Source: Morgan Stanley Research Estimates. *Market value.



                                                                                                                                                     19
                                                                     MORGAN STANLEY RESEARCH

                                                                     March, 2012
                                                                     EM Local Markets Guide




Asia ex-Japan
Note: This section has been updated since the publication of          India: Fiscal Slippage Still Likely in FY 2013
our 2012 Government Financing Outlook to reflect the
                                                                         Gross Market Borrowings for FY 2013 are budgeted
release of India and Thailand’s 2012 Budget
                                                                          at INR 5.69tn compared with INR 5.1tn in F12 RE.

Hong Kong: Issuance to be muted in 2012
                                                                      The FY 2013 budget has targeted a cut in central
   Gross EFN issuance declines from HKD 16bn in 2011                 government’s fiscal deficit to 5.1% (from BE of 5.9% of GDP
    to HKD 11bn in 2012 and net EFB issuance from                     in F2012). This deficit reduction target seems optimistic,
    2.8bn to 1.2bn due to lower interest.                             given the slowdown in growth and no measures to
                                                                      encourage private investment in the next fiscal. Our
The outstanding amount of EFB / EFN has historically                  economist, Chetan Ahya (see India Economics: Budget
increased in line with the interest rate paid on the paper,           F2013: Sustaining Higher Expenditure With Rise in Tax
although HKMA also increased EFB issuance sharply during              Burden, March 19, 2012) estimates the fiscal deficit for
2009 to expand the pool of risk-free paper as collateral for          F2013 to be higher than budgeted at 5.6% of GDP due to
RTGS and stock exchange clearing and settlement                       higher than budgeted subsidy burden and lower than
purposes.                                                             budgeted non-tax revenues. Gross Market Borrowings for FY
                                                                      2013 are budgeted at INR 5.69tn compared with INR 5.1tn in
For 2012, gross issuance should reduce slightly compared to
                                                                      F12 RE. However, according to our deficit forecast of 5.6%,
2011 on account of lower interest paid on EFB/EFN paper
                                                                      we expect market borrowings to be higher at INR 6tn. The
and as HKMA’s Aggregate Balance has remained broadly
                                                                      remaining deficit gap will be filled through further short-term
stable since May 2010. In turn, we expect net EFB issuance
                                                                      borrowings: we project short term borrowings of INR 500bn,
to decline to only HKD 1.2bn in 2012 from HKD 2.8bn in
                                                                      as against INR 346bn budget estimate and a further
2011, while gross EFN issuance will also be lower on
                                                                      drawdown in cash reserves by INR 50bn.
account of lower redemption flows.
                                                                      With the banks still suffering from a negative funding gap and
Most EFB/EFN funding is short-term, i.e., through 91-day,
                                                                      given the limited pick-up in bonds by FIIs, the government
182-day and 364-day EFBs as: a) it enables HKMA to earn
                                                                      will, in our opinion, continue to rely heavily on the RBI for
positive carry on its FX reserves (HKMA’s backing reserves
                                                                      help to absorb the projected GOI issuance during FY13. Our
are invested in US Treasuries), and b) as it allows HKMA
                                                                      analysis estimates 15% of the issuance will be funded by the
greater flexibility to vary the outstanding amount of Exchange
                                                                      central bank to bridge the demand shortfalls for central and
Fund instruments in line with fluctuations in the monetary
                                                                      state government bonds.
base.
                                                                      We expect these supply pressures to weigh on GoI yields
The budget for 2012-13 was released on 1st February, 2012.
                                                                      taking yields higher to 8.5% to 8.75%.
Exhibit 57
                                                                      Exhibit 58
HKD: EFN/EFB Issuance to Reduce Slightly
                                                                      Net Borrowing to Increase in FY 2013
HKD bn                                                 2011   2012
                                                                      INRbn                      2012 (RE)       2012 (MS)       2013 (BE)       2013 (MS)
EFN                                                                   Fiscal Deficit                     5220          5526            5136            5690
 Changes in EFN                                        -0.6    0.9    - as % GDP                           5.9           6.2             5.1             5.6
 Gross Issuance                                        16.4   11.0    Redem ptions
 Redemptions                                           17.0   10.1    - Market Loans                      736           736             906             906
EFB                                                                   Gross Funding Needs                5956          6262            6042            6596
 Changes in EFBs                                        2.8    1.2
                                                                      Funding
Total EFN/EFB Issuance                                  2.2    2.0    - Gross Market Loans               5100          5100            5696            6046
Source: Morgan Stanley Research Estimates, Bloomberg
                                                                      - Other Borrow ings                1103          1162             346             500
                                                                      - Draw dow n in cash               -247                0               0           50
                                                                      Total                              5956          6262            6042            6596
                                                                      Source: RBI, Morgan Stanley Research Estimates




                                                                                                                                                         20
                                                                                       MORGAN STANLEY RESEARCH

                                                                                       March, 2012
                                                                                       EM Local Markets Guide




Indonesia: Slashing Deficit Projections                                                Korea: Continued Fiscal Consolidation
    Net SBN issuance increases from IDR 127trn in 2011                                    Net issuance drops from KRW 31trn in 2011 to KRW
     to IDR 135trn in 2012, as decline in budget deficit is                                 25trn in 2012, as the budget deficit continues to
     offset by decline in non-debt financing.                                               shrink. Gross issuance increases from KRW 77trn to
                                                                                            KRW 80trn though due to a jump in redemptions.
In 2011, the Ministry of Finance decided to suspend the
issuance program at 96.88% of the net issuance target (at                              Korea’s 2012 budget continues to improve its fiscal balance
IDR 121.05trn), after considering the projected budget                                 with a reduction in the fiscal deficit from 2% of GDP last year
realization and cash balances at the end of 2011.                                      down to 1% for 2012. Government debt is also expected to
                                                                                       decrease 2.3 percentage points to 32.8 percent to GDP,
For 2012, the draft budget puts the fiscal deficit target at
                                                                                       down from 35.1 percent in 2011. This consolidation is in line
1.5% of GDP, which is significantly lower than the 2.1%
                                                                                       with the government’s aim to achieve fiscal balance by 2013.
deficit that was projected for 2011. This is explained by the
strong GDP growth expectations for 2012, with the budget                               Redemptions in 2012 will be higher than in the previous year
estimate being 6.7% GDP growth (Morgan Stanley                                         at KRW 43trn and we estimate that buybacks by the MoSF
projections are lower at 5.8%), providing stronger revenue                             will be around KRW 12trn, mainly in short term tenors.
support to the budget.
                                                                                       Given these estimates, gross issuance for 2012 is slightly
The deficit of IDR 124trn will be financed mainly through                              higher than previous year at KRW 79.8trn.
domestic SBN issuance targeted at net IDR 134.59trn or
                                                                                       Exhibit 60
1.7% of GDP. Conversely, offshore borrowing may be
                                                                                       KRW: Fiscal Deficit to Reduce to 1% of GDP
reduced slightly from the previous year to IDR 54.3trn to
lower the government’s foreign currency exposure.                                       KRW trn                                   2011          2012
                                                                                        Fiscal Balance                             24.6          14.3
Exhibit 59                                                                               % of GDP                                 2.0%          1.0%
IDR: Slashing Deficit and Borrowing Projections
IDR trn                                           2011 (Revised) 2012 BE                Government Debt                           437.0         448.2
Budget deficit                                             150.9    124.0                % of GDP                                35.2%         32.8%
 %GDP                                                       2.1%    1.5%

    Revenues                                                   1169.9         1311.4     Gross Issuance                            77.3          79.8
    Expenditures                                               1320.8         1435.4     Redemptions                               29.2          42.8
                                                                                         Buybacks*                                 17.5          12.0
Gross Borrowing Rqt                                             294.6          286.0
                                                                                        Net Issuance                               30.6          25.0
                                                                                       * Morgan Stanley Estimate
Debt Issuance                                                   269.1          295.6   Source: Morgan Stanley Research, MoSF
  Domestic Issuance                                             211.2          240.3
  Offshore Issuance                                              56.4           54.3
  Domestic Loan                                                   1.5            1.0
Domestic Redemptions                                             84.5          105.7
(including buybacks)
Offshore Redemptions                                              47.2          47.3
Net Bond Issuance
  Domestic                                                      126.7          134.6
  Offshore                                                       -2.7           -1.9
Other Repayments                                                 11.8            9.1
Non Debt                                                         25.5           -9.5
Changes in Cash                                                   0.0            0.0
Financing                                                       294.6          286.1
Source: Indonesia Debt Management Office, Morgan Stanley Research Estimates




                                                                                                                                                    21
                                                                                MORGAN STANLEY RESEARCH

                                                                                March, 2012
                                                                                EM Local Markets Guide




Malaysia: Fiscal Consolidation to Remain Gradual                                Philippines: Aiming for a 2% of GDP Deficit in 2013
                                                                                 Total gross debt issuance declines to PHP 704bn in
   Total gross issuance should reach MYR 96bn in 2012,                            2012 from PHP 738bn in 2011 on smaller deficit.
    roughly unchanged from 2011, as we expect only a
    modest deficit reduction.                                                   The budget deficit target for Philippines for 2012 has been
                                                                                laid down at 4.6% of GDP, with the aim to reduce the fiscal
The fiscal deficit target for 2012 is budgeted at 4.7% of GDP                   deficit to 2% of GDP by 2013. The budget deficit of PHP
or MYR 43bn, lower than the 5.4% deficit target set for last                    286bn and outstanding debt will be financed mainly through
year. In turn, this continues the gradual fiscal consolidation                  borrowings. Of total debt funding, PHP 529.5bn will be from
trend, which should see gross issuance (domestic + external                     the domestic market and PHP 174.8bn from offshore
debt) decline to MYR 91.3bn from MYR 96.6bn in 2011.                            borrowings. The government has decided to reduce the
However, we believe that the budgeted deficit estimate is too                   foreign-to-domestic borrowing mix to 25:75 from the 2010
optimistic. We expect growth to moderate to 4% yoy for 2012                     ratio of 34:66 in order to further to reduce the economy’s
and the fiscal deficit to reduce only modestly to MYR 46bn,                     vulnerability to external shocks. Total debt issuance for 2012
i.e., 5% of GDP, on lower revenue outcomes. Accordingly,                        will be 6.4% of GDP (PHP 704.3bn), which is PHP 33.9bn or
we estimate gross issuance for next year to be MYR 96bn                         4.6% less than in 2011.
(assuming issuance of MYR 55bn of MGS, MYR 35bn for GII                         The government also announced that they will seek to further
and the rest in International bonds and Sukuks), slightly                       improve the debt profile by extending maturities for current
higher than the budgeted MYR 91bn.                                              debt and by re-denominating external debt by increasing the
The issuance calendar for this year is evenly distributed over                  issuance of peso-denominated bonds in the international
the four quarters; while the average maturity of the issuance                   market. In its first year in office, the Department of Finance
will be similar to previous years, at 8.8 years.                                has successfully swapped USD 2.3bn of the existing debt
                                                                                externally and PHP 171.9bn domestically.
Exhibit 61
MYR: Gradual Fiscal Consolidation to Continue                                   Exhibit 62
                                                                                PHP: On Track for a 2% Fiscal Deficit by 2013
MYR bn                                          2011    2012 BE*   2012 MSE*     PHP bn                                                        2011        2012
Budget Deficit                                   46.0       43.0         46.0    Budget deficit                                               298.0       286.3
 % of GDP                                       5.4%       4.7%         5.0%      %GDP                                                         3.0%       2.6%
                                                                                 Revenue                                                     1410.4      1563.6
Gross Issuance                                   96.6       91.3         96.0
 MGS                                             57.5       53.0         55.0    Expenditure                                                 1708.4      1849.9
 GII                                             32.5       32.0         35.0
                                                                                 Gross Borrowing Rqt                                          726.8       691.8
 International Bonds                              3.0        3.0          3.0
 Sukuks                                           3.0        3.0          3.0    Debt Issuance                                                738.2       704.3
Redemptions                                      50.6       48.3         48.3
                                                                                  Domestic issuance                                           546.3       529.5
 MGS                                             37.5       45.6         45.6
                                                                                  Offshore Issuance                                           191.8       174.8
 GII                                              7.6        2.7          2.7
 International Bonds                              5.5        5.5          5.5
                                                                                 Domestic Redemptions                                         285.1       339.6
Net Issuance                                     46.0       43.0         47.7    Offshore Redemptions                                         143.7        65.9
*BE = Budget Estimate, MSE = Morgan Stanley Estimate
Source: FAST BNM, Morgan Stanley Research Estimates                              Net Financing
                                                                                  Domestic                                                    261.2       189.9
                                                                                  Offshore                                                     48.1       108.9

                                                                                 Non Debt                                                          0.0      0.0
                                                                                 Change in Cash
                                                                                  Budget                                                        9.3        12.8
                                                                                  On/off-budget                                                -6.5        -6.2
                                                                                 Total                                                        300.0       286.0
                                                                                 Financing                                                    738.2       704.3
                                                                                Source: Bureau of Treasury Philippines, Morgan Stanley Estimates




                                                                                                                                                                 22
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Singapore: SGD/Bill Issuance may be higher in 2012                Taiwan: Issuance to Rise Sharply in 2012

   We expect both bond and bill issuance will be higher             Gross issuance increases from TWD 565bn in 2011 to
    in 2012, with overall net issuance up from SGD 21bn               TWD 686bn in 2012 due to a rise in the deficit and
    in 2011 to SGD 31bn in 2012.                                      redemptions.

The MAS only releases an auction schedule with the maturity       The central government’s fiscal deficit for 2012 is projected
breakdown. We expect SGS issuance in 2012 to increase to          at TWD 225bn (including TWD 94bn of debt repayments).
SGD 21bn, about SGD 5bn higher than issuance in 2011.             The fiscal deficit will be part funded by a TWD15bn
                                                                  drawdown in cash reserves and TWD 210bn of net issuance.
There will likely be a further increase in Tbill issuance, both
in SGD bills and of MAS bills, particularly in the event of       For 2012, gross issuance will be a record TWD 686bn, up
further SGD inflows. We estimate gross SGD bill issuance to       TWD 103bn from 2011. However, notwithstanding this sharp
rise to SGD 210bn, up from SGD 191bn issued last year,            pick-up in gross TWGB issuance, we expect the TWGB
while net issuance of MAS bills may rise by up to SGD 5.4bn       curve to remain well supported, on strong onshore demand
to SGD 20.4bn for 2012.                                           from banks and life insurance companies.

We estimate the size of SGS issuance to be between
SGD1.5 – 2bn for most re-openings, in excess of SGD 2.5bn         Exhibit 64
for new issues with a maturity less than 5yrs, and between        TWD: Issuance to Rise Sharply in 2012
SGD 1- 1.5bn for longer maturities. With these assumptions,        TWD bn                                       2011    2012
we reach an estimate of SGD 8.1bn in net issuance, which is        Total Revenues                              1,854   2,033
almost double the net issuance amount for 2011.                     Annual Revenues                            1,646   1,729
                                                                    Bond Issuance                                205     288
                                                                    Appropriation from                             3      15
Exhibit 63                                                          Previous Year's Surplus
SGD: SGD/Bill Issuance May be Higher in 2012
SGD bn                                       2011    2012          Total Expenditures                          1,854   2,033
SGS                                                                 Annual Expenditures                        1,788   1,939
 Net Issuance                                  4.2    8.1           Debt Repayment                                66      94
 Gross Issuance                               15.7   21.3
 Redemptions                                  11.5   13.2           Gross Issuance                              565     686
                                                                    Redemptions                                 340     398
SGD bills                                                           Cash Balance                                  0       0
 Net Issuance                                  1.5    2.8          Net Issuance                                 225     288
                                                                  Source: Morgan Stanley Research Estimates.
 Gross Issuance                              191.2    210
 MAS Bills (Net)                              15.0   20.4
Total SGS Issuance (Net)                      20.7   31.3
Source: Morgan Stanley Research Estimates.




                                                                                                                               23
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Thailand: Supply Pressures to Remain High

     Total gross issuance jumps from THB 378bn in FY11
      to THB 450bn in FY12 on higher deficit.

On Feb 23, the cabinet approved the preliminary budget for
Fiscal 2013. The government plans to bring the country's
fiscal deficit down to THB 300bn from the current THB 400bn
for FY2012. The government expects that in FY2013, while
expenditures will increase by Bt55.2, revenues will be up by
Bt120 billion. An emergency decree, issued on Jan 26th,
allows the government to shift the interest payment on FIDF
bonds (up to THB 65bn each year) to the Bank of Thailand.
The decree also authorises an extra borrowing of THB 350bn
for flood relief projects, increasing public debt from 40.3% of
GDP to 47%. The Finance Ministry plans to borrow this
amount from the money market in lots of Bt30 billion to Bt60
billion by issuing promissory notes, in keeping with the
progress of the projects. This short-term loan might then be
converted into government bonds, to save on interest cost,
which may add to the supply in FY2013.

In FY2012, supply pressures will remain high in 3Q and 4Q
2012, given that only THB 193.5bn of the total budgeted
issuance of THB 525bn for FY2012 has been issued in the
first two quarters. The issuance will remain skewed towards
shorter maturity with maximum issuance in the 5y tenor.

Exhibit 65
Issuance to Rise Sharply in FY2012
                     Budgeted                          Q3-Q4
Tenor                Issuance Q1 IssuanceQ2 Issuance Issuance
3Y                              50          0      16       34
5Y                            100           15     31       54
7Y                              65          9      23       33
10Y                             60          0      15       45
15Y                             35          0      10       25
20Y                             35          0      16       19
30Y                             30          0      7.5    22.5
50Y                             25          0      10       15

FRN                             50          8      18       24
TGB Linkers                     45          0      15       30
Amortized Bond                  30          0       0       30
Total                         525           32   161.5   331.5
Source: Morgan Stanley Research Estimates




                                                                                            24
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




CEEMEA
Hungary: Net domestic issuance unchanged as                                            Total net FX borrowing is also little changed on current
redemptions decline, higher net external issuance                                      plans, but net external bond issuance increases from EUR
   Net domestic bond issuance should be roughly                                       2bn to EUR 2.7bn to compensate for higher redemptions on
    unchanged at HUF 566bn, while net external issuance                                multilateral loans. Hungary has SDR 3.2bn (EUR 3.8bn) of
    jumps to EUR 2.7bn. Redemptions down HUF 434bn.                                    principal repayments due to the IMF in 2012, starting with the
                                                                                       first repayment of SDR 527m (EUR 630m) on 10 February.
The net financing requirement for 2012 is expected to reach
                                                                                       Note though that Hungary may sign a new IMF arrangement
HUF 674bn, on a central government cash deficit of HUF
                                                                                       this year, which will likely affect the AKK’s existing plans for
576bn. Total net borrowing should come at HUF 643bn, HUF
                                                                                       external issuance.
98bn higher than 2011.
Exhibit 66                                                                             Hungary has a relatively low average maturity of domestic
Hungary Net Financing Requirement, HUF bn                                              debt at 3.6 years and high domestic rollover requirement for
                                               2011                    2012            2012 at 7.5% of GDP.
                                    Planned     Outturn    Chg     Planned vs 2011
CG cash deficit                          689         na     na          576     na     Exhibit 68
% of GDP                              2.40%          na     na       2.00%      na
Net pre-financing of EU transfers        171         na     na           98     na     Hungary Redemption Profile in 2012, HUF bn*
Net financing requirement                860         na     na          674     na
                                                                                         700
Net Tbill issuance                      199         -74    -273           11     85
Gross issuance                        1,778       1,544    -234        1,555     11
                                                                                         600
Redemptions                           1,579       1,618      39        1,544    -74
Net retail debt issuance                 36          37       1           36     -1
Gross issuance                          366         338     -28          366     28      500
Redemptions                             330         301     -29          330     29
Net domestic bond issuance              382         574     192          566     -8      400
Gross issuance                        1,360       1,765     405        1,230   -535
Redemptions                             978       1,191     213          664   -527      300
Net fx borrowing                        156           8    -148           30     22
Gross issuance                        1,344       1,278     -66        1,437    159      200
 Bonds                                1,120       1,090     -30        1,198    108
 IFIs loans                             224          na      na          240     na      100
Redemptions                           1,188       1,270      82        1,408    138
 Bonds                                  560         540     -20          400   -140
 IMF loans                                0          na      na          975     na
                                                                                          -
 IFIs loans                             628          na      na           33     na               Jan Feb Mar Apr May Jun                Jul Aug Sep Oct Nov Dec
Total net borrowing                     773         545    -228          643     98
                                                                                                      Domestic Bonds           Tblls      External Bonds            IMF
Total domestic redemptions          2,887         3,110     223        2,538   -572    Source: Morgan Stanley, AKK, IMF, EC, Bloomberg. * Excluding retail bonds.
Total fx redemptions                1,188         1,270      82        1,408    138
Total redemptions                   4,075         4,380     305        3,946   -434    Non-resident holdings of HGBs increased sharply in 2011 by
Source: Morgan Stanley Research, AKK.
                                                                                       HUF 1,200bn to HUF 3,800bn or almost 50% of the total.
The AKK expects net domestic bond issuance of HUF
                                                                                       Holdings peaked in September, however, and have declined
566bn, roughly unchanged from 2011, with both gross
                                                                                       slightly since then. Hungary was recently downgraded to
issuance and redemptions falling sharply. Gross issuance
                                                                                       non-investment grade by all three major ratings agencies.
should reach HUF 1,230bn, down HUF 535bn. Issuance will
be concentrated in 3-year and 5-year maturities,                                       HGBs underperformed most CEEMEA markets in 2011 but
representing 47% and 32% respectively of total. Bond                                   have rallied sharply at the start of 2012. The sustainability of
redemptions drop sharply by HUF 527bn to HUF 664bn.                                    the recent rally will hinge on the successful conclusion of
                                                                                       talks with the IMF and EC on a new lending arrangement.
Exhibit 67
                                                                                       More volatility is likely in the near term, though, as
HGB Expected Bond Issuance in 2012 by Maturity                                         negotiations with lenders might prove tricky and eurozone
                                               HUF bn     % of total
                                                                                       sovereign bond markets could deteriorate again. We do,
3y                                                580          47%
5y                                                390          32%                     however, expect the government to eventually sign an IMF
10y                                               200          16%                     deal by end-Q1 or early Q2 at the latest, and the central
15y                                                60            5%
                                                                                       bank to start cutting rates from the middle of the year, which
Total                                           1,230         100%
Source: Morgan Stanley Research, AKK.                                                  should open the door for a further rally in HGBs.




                                                                                                                                                                          25
                                                               MORGAN STANLEY RESEARCH

                                                               March, 2012
                                                               EM Local Markets Guide




                                                               rollover requirement next year at 6.5% of GDP and a below-
Poland: Small decline in net domestic issuance,
                                                               average maturity of domestic debt of 4.2 years.
large drop in net external issuance
                                                               On the other hand, net TBill issuance should increase in
    Poland’s net borrowing requirement for 2012 is
                                                               2012 by about PLN 19bn to PLN 5.2bn. The MoF repaid
     expected to drop by PLN 5bn to PLN 50bn. Net
                                                               about PLN 14bn of TBills in 2011 and benefited from a
     domestic bond issuance declines by PLN 2.6bn to
                                                               consolidation of the liquid assets of state agencies and local
     PLN 20.4bn. Net external issuance drops sharply
                                                               governments, which boosted the MoF’s cash position by PLN
     from EUR 4bn to EUR 1bn as the MoF has decided to
                                                               23bn. The consolidation is not expected to produce further
     draw on its FX cash holdings to fund the higher
                                                               benefits in 2012, which means the MoF needs to up TBill
     redemptions.
                                                               sales to meet its borrowing needs.
Poland’s net borrowing requirement is expected to decline
                                                               Net FX financing is also expected to drop by about PLN 2bn
slightly next year to PLN 45bn from PLN 50bn in 2011.
                                                               to PLN 17.6bn. Within that, net external bond issuance drops
Borrowing is lower despite a large downward revision of the
                                                               sharply from PLN 16.1bn (~EUR 4bn) to PLN 4.4bn (about
GDP growth forecast in the new Budget from 4% to 2.5%, as
                                                               EUR 1bn), as the MoF has decided to draw on its FX cash
the government continues to implement its fiscal
                                                               holdings to finance the higher bond redemptions in 2012. FX
consolidation plans. Although the domestic deficit increases
                                                               financing represents around 40% of total net borrowing by
by about PLN 5bn, a smaller deficit in EU funds means that
                                                               Poland, by far the highest of any CEEMEA borrower.
the overall deficit in 2012 is lower by about PLN 3bn. Also,
the MoF announced that it had completed the borrowing          Exhibit 70
requirement for 2011 by October and that it has been pre-      Poland Domestic Debt Redemption Profile in 2012,
financing for 2012 since then.                                 PLN bn
Exhibit 69                                                        30
Poland Borrowing Requirement, PLN bn                              25
                                               2011    2012
Overal deficit                                 42.6    39.5       20
 Treasury deficit                              30.2     35.0
                                                                  15
 Deficit in EU funds                           12.4      4.5
                                                                  10
Net borrowing requirement                      50.4     45.1
                                                                   5
Tbills                                         -13.7     5.2
                                                                   0
Net domestic bond issuance                      22.8    20.4           Jan   Feb   Mar   Apr      May   Jun   Jul   Aug   Sep    Oct    Nov   Dec
 Gross issuance                                 97.6   106.0
  Floating                                      24.0    11.5                             TBills                                 Bonds
  Fixed                                         71.5    89.6
                                                               Source: Morgan Stanley, Bloomberg.
  Index-linked                                   2.1     4.8
 Redemptions                                    74.8    85.6   Foreigners’ holdings of Polish bonds increased from PLN
Other domestic securities                       -0.8     2.6   125bn in 2010 to PLN 150bn at the end of November 2011.
Net fx financing                                19.5    17.6   Holdings peaked in September, however, and have declined
 Net bond issuance                              16.1     4.4
  Gross issuance                                20.0    18.2   slightly since then.
  Redemptions                                    3.8    13.8
 Loans                                           6.6     6.4
                                                               Polish bonds posted a better than average return in 2011
 FX cash position and other                     -3.2     6.8   and outperformed their CEE peers, and like most local
Liquidity management and other credits          22.6    -0.7   markets have rallied at the start of 2012. However, Eurozone
Total financing                                 50.4    45.1
Source: Morgan Stanley, Ministry of Finance.
                                                               conditions are likely to continue to weigh on investor
Net domestic bond issuance is set to decline by PLN 2.6bn      sentiment in the near term, and a sustained rally in Polish
to PLN 20.4bn, while gross issuance jumps by PLN 8.4bn to      bonds will have to wait for the return of international
PLN 106bn due to higher redemptions. Fixed-rate and index-     investors. Some support for the bond market could come if
linked issuance is expected to be higher in 2012 while         monetary policy is eased from Q2, as our economists expect.
floating issuance should be lower. Total redemptions in 2012
are PLN 103bn. Poland has a relatively high domestic


                                                                                                                                                26
                                                                         MORGAN STANLEY RESEARCH

                                                                         March, 2012
                                                                         EM Local Markets Guide




Note: This section has been updated since the publication of             Exhibit 72
our 2012 Government Financing Outlook to reflect the                     SAGB Issuance by Maturity and Type, ZAR bn
release of South Africa’s 2012 Budget.                                    Fixed Rate        Maturity            2010/11                2011/12
                                                                                              5-8                21         21%         21        24%
                                                                                             8-15                50         49%         38        43%
South Africa: Issuance lower despite higher deficit                                         Over 15y             32         31%         30        33%
                                                                                             Total              103                     89
    The borrowing requirement for 2012/13 is expected to
     jump to ZAR 169bn from ZAR 153bn in 2011/12. Net                     CPI                5-15                23         53%          8        24%
     issuance declines though due to a large drawdown in                                    Over 15y             20         47%         25        76%
     cash balances.                                                                          Total               43                     33
                                                                         Source: Morgan Stanley, Bloomberg.

The National Treasury published the 2012 Budget on 22                    To create benchmark bonds and smooth the maturity
February with updated economic and borrowing forecasts for               structure, the Treasury plans to introduce five new bonds in
2012/13 and future years. The borrowing requirement is now               2012/13, two of which will have a maturity split over three
expected to jump by ZAR 16bn to ZAR 169bn.                               years, as follows:

Exhibit 71                                                               Exhibit 73
South Africa Borrowing Requirement, ZAR bn                               New Domestic Bonds, 2012/13
                                             2011/12   2012/13   Chg                 Fixed-rate                            Inflation-linked
National budget balance                        156.6     170.0   13.4           Code          Maturity                  Code             Maturity
% of GDP                                        -5.2      -5.2    0.0           R2023        28-Feb-23                  RI2025          31-Jan-25

Net borrowing requirement                      152.7     168.8   16.1                             28-Feb-47             RI2038          31-Jan-38
                                                                                R2048             28-Feb-48
Short-term issuance                             20.8      22.0     1.2
                                                                                                  28-Feb-49                             31-Dec-49
Net domestic bond issuance                     139.9     120.0   -19.9
                                                                                                                        RI2050          31-Dec-50
 Gross issuance                                155.4     151.4    -4.0
 Redemptions                                   -15.5     -31.4   -15.9                                                                  31-Dec-51
                                                                         Source: Morgan Stanley, Bloomberg.
Net foreign borrowing                            9.5      -7.5   -17.0
 Gross issuance                                 12.0       4.0    -8.0   Net purchases by foreigners in 2011 were 46bn, down from
 Arms procurement                                1.0       0.2    -0.8   56bn in 2010. Cumulative flows peaked in August, though,
 Redemptions                                    -3.5     -11.7    -8.2   and are slightly down since then. Moody’s and Fitch
Change in cash                                 -17.6      34.3    51.9
                                                                         downgraded South Africa’s rating outlook to negative from
Financing                                      152.7     168.8    16.1
Source: Morgan Stanley, National Treasury.                               stable in recent months.
While TBill issuance is expected to increase by ZAR 1bn, net
                                                                         Exhibit 74
domestic bond issuance and net foreign borrowing drop
                                                                         SA Domestic Redemption Profile in 2012, ZARbn
sharply by ZAR 20bn and ZAR 17bn, respectively. The
Treasury is able to lower issuance despite the higher deficit              40

due to a large drawdown on its cash balance of ZAR 34bn, a                 35
sharp reversal of last year’s increase of ZAR 17.6bn. The                  30
change in cash finances not only the higher deficit but also
                                                                           25
the increase in domestic and foreign redemptions of ZAR
                                                                           20
16bn and ZAR 8bn. Gross domestic and external issuance
declines therefore to ZAR 151bn and ZAR 4bn.                               15

                                                                           10
In terms of maturity, in 2011/12 issuance of fixed rate bonds
                                                                            5
remained concentrated in the belly or 8-15y part of the curve,
which accounts for about 40-50% of all issuance. CPI linked                 0
                                                                                Jan Feb     Mar       Apr May   Jun   Jul   Aug Sep   Oct    Nov Dec
issuance on the other hand has been shifted noticeably
towards longer-dated maturities, in particular the 2033 R202.                                 Bills                                   Bonds

                                                                         Source: Morgan Stanley




                                                                                                                                                   27
                                                                                           MORGAN STANLEY RESEARCH

                                                                                           March, 2012
                                                                                           EM Local Markets Guide




Turkey: Net issuance higher in 2012                                                        Net foreign borrowing is again expected to be negative in
                                                                                           2012, with TRY 9.5bn of gross financing coming from a
    Borrowing is expected to increase in 2012 to TRY
                                                                                           combination of Eurobond issuance and loans from
     18bn as GDP growth slows down. Net domestic bond
                                                                                           multilateral institutions.
     issuance should jump to TRY 20.3bn from TRY
     16.9bn, with redemptions declining to TRY 82bn.                                       Domestic bond redemptions should decline next year to TRY
                                                                                           82bn. Turkey has the lowest average maturity of domestic
The improvement in Turkey’s fiscal position continued in
                                                                                           debt among CEEMEA borrowers at 2.5 years and a relatively
2011, with the central government budget balance declining
                                                                                           high domestic rollover requirement for 2012 at 6% of GDP.
to 1.7% of GDP from 3.6% in 2010 and below earlier
projections of 2.8%. Net borrowing almost halved to TRY                                    Exhibit 77
15bn, while total financing was roughly unchanged due to an                                Turkey Domestic Debt Redemption Profile in 2012,
increase in other financing. However, borrowing is expected                                TRY bn
to increase next year by 23% to about TRY 18bn, as GDP
                                                                                              20
growth cools down from 7.5% to 4.0%.
                                                                                              18

Exhibit 75                                                                                    16

Turkey Borrowing Requirement, TRY bn                                                          14

                                                          2011                                12
                                         2010     Program       Est    Chg         2012       10
CG Budget balance                        -39.6       -33.5    -22.2    11.3        -21.1
% of GDP                                   3.6          2.8      1.7       -1.1      1.5       8

                                                                                               6
Total financing                          47.9          42.4     41.7   -0.7        47.6
Net borrowing                            27.0          21.3     14.5   -6.8        17.8        4

                                                                                               2
Net domestic bond issuance                22.8          19.8    16.9 -2.9          20.3
                                                                                               0
 Gross issuance                          159.0         119.1   114.0 -5.1         101.9
                                                                                                   Jan   Feb   Mar    Apr   May   Jun   Jul   Aug   Sep   Oct   Nov   Dec
 Redemptions                             136.2          99.3    97.1 -2.2          81.6
Net foreign borrowing                       4.2          1.5     -2.4 -3.9         -2.5
 Gross borrowing                          14.9          12.5      9.2 -3.3          9.5                                             Bonds
 Redemptions                              10.7          11.0    11.6     0.6       12.0    Source: Morgan Stanley, Bloomberg.
  Eurobonds                                                                         4.1
  IMF                                                                               3.6    Non-resident holdings of TRY bonds increased from TRY
  Other                                                                             4.2    50bn at the end of 2010 to TRY 71bn at end-2011. Holdings
Other financing                           20.9          21.1    27.2     6.1       29.8
Source: Morgan Stanley, Undersecretariat of Treasury: 2012 Treasury Financing Program.     peaked in August, however, and declined through November
The Treasury expects net domestic bond issuance to                                         but recovered sharply in December.
increase to TRY 20.3bn in 2012 from TRY 16.9bn in 2011.
                                                                                           Turkish bonds were the worst performing in CEEMEA in
Zero-coupon bonds accounted for the majority of issuance in
                                                                                           2011, with massive bouts of volatility as the central bank
2011, but the Treasury has stated that its preference for
                                                                                           changed monetary policy dramatically, initially easing then
2012 will be to issue more fixed-rate instruments at
                                                                                           tightening sharply to support the currency. The effective
maturities over 12 months. In order to improve liquidity,
                                                                                           policy rate was increased by about 600bp from the beginning
issuance will be concentrated in benchmark maturities, with
                                                                                           of October to December.
2y issued monthly, and 5y and 10y in months of high
redemptions.                                                                               Bonds have rallied sharply this year as the CBT cut the
                                                                                           effective rate and recent auctions have been extremely
Exhibit 76
                                                                                           strong. We expect yields to stabilize in the near term as the
TRY Bond Issuance in 2011 by Type, as % of Total*
                                                                                           central bank evaluates the effectiveness of its measures and
Zero coupon                                                                       44.6
                                                                                           the durability of the recent appreciation of the TRY. However,
Fixed coupon                                                                      36.0
                                                                                           a sustained rally in Turkish bonds will require a durable
CPI-linked                                                                         9.9
Floating rate                                                                      8.7     improvement in the external environment.
Revenue indexed                                                                    0.8
Source: Morgan Stanley, Undersecretariat of Treasury. *Through Oct-2011.




                                                                                                                                                                        28
                                                                                           MORGAN STANLEY RESEARCH

                                                                                           March, 2012
                                                                                           EM Local Markets Guide




Czech Republic: Net borrowing continues to decline                                         Russia: Bond issuance increases again
      Net bond issuance declines sharply from CZK 77bn                                         Net domestic issuance is expected to rise from RUB
       in 2011 to CZK 34bn in 2012 as the government fiscal                                      939bn in 2011 to RUB 1,209bn in 2012 as the deficit is
       position improves.                                                                        again projected to increase.

The Ministry of Finance expects the net borrowing                                          The Ministry of Finance expects a borrowing requirement for
requirement to decline from CZK 143bn in 2011 to CZK                                       2012 of RUB 1,381bn, up from RUB 937bn in 2011, and net
105bn in 2012, as the government continues to reduce the                                   domestic borrowing of RUB 1,209bn, up from RUB 939bn.
primary fiscal balance. Net domestic bond issuance is likely                               Net external borrowing is unchanged. The projections are
to fall from CZK 77bn to CZK 34bn.                                                         based on a deficit of RUB 877bn or 1.5% of GDP.

In 2011, the MoF borrowed CZK 45bn more than initially                                     Domestic issuance in 2011 was lower than planned by RUB
planned, and as a result increased its cash reserve by CZK                                 400bn, as the Russian government ran a balanced budget
5bn instead of drawing it down by CZK 36bn. Borrowing was                                  instead of the planned deficit of 3.6% of GDP. Issuance was
higher than expected across all market segments, but the                                   elevated in the first half of the year, but was then reduced
biggest contribution of CZK 20.4bn came from government                                    sharply as the MinFin had built up a large cash surplus.
savings bonds, launched for the first time in 2011 and sold
                                                                                           Exhibit 79
through retail channels to individuals.
                                                                                           Russia 2012-14 Budget and Borrowing Program,
Exhibit 78                                                                                 RUB bn
Czech Republic Net Borrowing Requirement, CZK                                                                                            2011

bn                                                                                                                               Plan    Outturn       Chg         2012    2013
                                                                                            Budget deficit                       1,814          0    -1,814         877    1,025
                                                      2011                  2012
                                    2010        Plan Outturn Chg           Plan vs 2011     % of GDP                               3.6       -0.5                   1.5     1.6
State budget deficit                156.4      135.0     142.8 7.8        105.0    -37.8    Reserve fund                            na       -911                  -512     -571
% of GDP                               4.1        3.5       3.5  0.0         2.7    -0.8    National Welfare fund                   na          3                     8      10
    Primary balance                 120.6       66.5       97.6 31.1       32.1    -65.5
    Net expenditure on state debt    35.8       68.5      45.1 -23.4       72.9     27.8
                                                                                            Borrowing requirement                1,814        907     -907        1,381    1,586
Net borrowing requirement           156.4      135.0    142.8    7.8      105.0    -37.8
                                                                                            Total financing                      1,814        937     -877        1,381    1,586
Net borrowing                       168.5      106.3    151.0 44.7      107.2      -43.8    Net borrowing                        1,387      1,077     -309        1,346    1,330
 Net money market issuance           25.1       30.0     49.3 19.3       50.0        0.7
                                                                                               Net domestic issuance             1,341        939     -402        1,209    1,186
 Net bond issuance                  134.1       71.9     77.1 5.2        34.4      -42.7
  Gross bond issuance               217.1      184.0    181.2 -2.8      167.5      -13.7         Gross bond isssuance            1,710                            1,809    1,803
   Domestic bonds                                       180.3      64.5-177.5                    Bond redemptions                 369                               600     617
   External bonds                                         0.9            0-73                  Net external borrowing*              46        138       92          137     144
  Bond redemptions*                   83.0     112.1    104.1 -8.0      133.1       29.0
                                                                                                 Gross external borrowing         242                               210     213
 Gross savings bond issuance           0.0       0.0     20.4 20.4       20.0       -0.4
 Net EIB loans                         9.3       4.4      4.2 -0.2        2.8       -1.4         External redemptions             197                                73      69
                                                                                            Privatisation                         298         174     -124          300     380
Assets operations                     -11.8       28.7     -7.8 -36.5       -2.2     5.6
 Financial assets operations           -3.0        -2.1    -2.5 -0.4        -2.2     0.3   Other                                   130       -315       -444        -265    -124
 On-lending                            -1.7        -4.9     0.0 4.9          0.0     0.0   Source: Morgan Stanley, Ministry of Finance. *95% of which is bond issuance.
 Financial reserve                     -7.1       35.7     -5.3 -41.0        0.0     5.3
                                                                                           Domestic borrowing in the capital markets, rather than the
Source: Morgan Stanley, Ministry of Finance. * Including buybacks and switches.
Net borrowing in 2012 is expected to decline by CZK 44bn to                                sovereign funds, will continue to be the main source of
CZK 107bn. Net bond issuance drops by CZK 43bn to CZK                                      financing the federal budget deficit in the medium-term. Most
34bn, and gross issuance by CZK 14bn to CZK 168bn.                                         oil and gas budget revenues will cover budget expenditures
Floating rate notes will comprise at least 30% of domestic                                 and only a small proportion (3.3% in 2012) will be allocated
issuance. Net money market issuance on the other hand, is                                  to the Reserve Fund and none to the National Wealth Fund.
projected to remain at similar levels of around CZK 50bn.                                  The average maturity of OFZ issuance in 2011 was about 5
Savings bond sales will continue in 2012 at similar levels and                             years and is expected to remain at a similar level in 2012.
will become an integral part of debt management planning.                                  The Ministry expects to issue fixed-rate only OFZs, and to
External issuance was just CZK 0.9bn in 2011 and is                                        continue issuing savings bonds (GSOs) to meet demand
expected to remain low next year as well. Based on market                                  from domestic institutional pension investors.
conditions, the Ministry will aim to issue one syndicated
government bond on the foreign market.



                                                                                                                                                                            29
                                                                                          MORGAN STANLEY RESEARCH

                                                                                          March, 2012
                                                                                          EM Local Markets Guide




LatAm
Mexico: Stable Funding Outlook                                                            Exhibit 81
                                                                                          Mexico Bond Issuance by Maturity and Type
    Overall net issuance in 2012 is roughly unchanged at
     MXN 360bn, with a large decline in net bond issuance                                  Fixed Rate                  Maturity                                      2010/11                      2011/12
     from MXN 349bn to MXN 158bn and large increase in                                                                 Cetes                                                       14%                     16%
     net Bill issuance from MXN 47bn to MXN 202bn.                                                                     Bondes D                                                     7%                         7%
                                                                                                                       Fixed Rate Bonds                                            55%                     53%
In 2012 the government fiscal plan is a continuation of its
2011 plan. This plan entails some modest fiscal stimulus
                                                                                           CPI                         Udibonos                                                    20%                     20%
while preserving the sustainability of public finances.
                                                                                                                       Udibonos Udi                                                 4%                      4%
Excluding PEMEX’s investment, Mexico will run a fiscal                                    Source: Morgan Stanley, Bloomberg. Banxico – 2012 figures based on Q1/2012 projection.
deficit equivalent to 0.4% of GDP in 2012. The government
approved budgetary revenues of MXN 3,310.5bn and                                          Average Maturity Increasing
financing of MXN 396.9bn for 2012, leading to total
resources available for the public sector of MXN 3,706.9bn,                               The share of long term fixed-rate Government securities
MXN 147.7bn more than the amount approved for 2011                                        increased from 60.5% at the end of 2010 to 60.8% at the end
(4.1% in real terms).                                                                     of 2011. The average maturity of domestic debt increased by
                                                                                          0.29 years (105 days), changing from 7.20 to 7.6 years in the
Exhibit 80                                                                                same period (vs. the US for example with an average
Authorized Borrowing for 2011 to Remain Stable                                            maturity of 5.7 years). This was the result of the adjustments
                                    2010               2011E               2012E          in the issuance program of government securities, the
Issuance                                                                                  issuance of long term securities and the syndicated
 Bonds                                 538,181             547,200             358,215
 Bills                               1,416,869           1,211,282           1,345,000    issuances performed during January-November of the
                                                                                          current year.
Gross Issuance                       1,955,050           1,758,482           1,703,215
                                                                                          Exhibit 82
Redemptions                                                                               Mexico Debt Redemption Profile in 2011
Bonds                                  264,129            197,848              200,000
Bills                               1,477,836           1,164,386           1,143,100
                                                                                            180

Net Issuance                           213,085            396,249             360,115       160
Source: Morgan Stanley, Banxico, Mexico Ministry of Finance – 2012 figures extrapolated
from Q1 2012 issuance plan.                                                                 140
The 2012 budget lifts the ceiling for government borrowing                                  120                                                                                    Bills
higher by MXN 40bn to MXN 435bn. This increase in the                                       100
                                                                                                                                                                                   Bonds
borrowing ceiling is due to the upward revision of public                                    80
deficit resulting from increased federal government spending.                                60
An external net debt issuance of MXN 7.0bn was approved.
                                                                                             40
This increased borrowing means that as a percentage of
                                                                                             20
GDP, net debt in Mexico will increase by 0.8% in 2012 to
33.9%. We expect the government will attempt to frontload                                     0
                                                                                                  Dec-11

                                                                                                           Jan-12

                                                                                                                    Feb-12

                                                                                                                             Mar-12



                                                                                                                                               May-12

                                                                                                                                                        Jun-12

                                                                                                                                                                 Jul-12




                                                                                                                                                                                             Oct-12

                                                                                                                                                                                                      Nov-12

                                                                                                                                                                                                               Dec-12
                                                                                                                                                                                    Sep-12
                                                                                                                                      Apr-12




                                                                                                                                                                          Aug-12




issuance in the first quarter of the year as much as it can in
order to minimize potential volatility arising form the July
                                                                                          Source: Morgan Stanley Research, Bloomberg
2012 presidential elections.




                                                                                                                                                                                                                        30
                                                                                   MORGAN STANLEY RESEARCH

                                                                                   March, 2012
                                                                                   EM Local Markets Guide




Brazil: Likely to Maintain its Issuance Strategy                                   Exhibit 84
                                                                                   Type of Bonds Issued in Percent in 2011
Note: This section has been updated since the publication of
our 2012 Government Financing Outlook to reflect the                                                              Bond Type               Percent Issued
release of Brazil’s 2012 Budget                                                                                          LFT                       4.4%
                                                                                                                         LTN                      30.1%
    The Brazil financing plan was released in March 2012,                                                           NTN-B                        28.3%
     they expect a gross issuance of R$464bn and the                                                                 NTN-F                        37.2%
     continued extension of the debt maturity profile. .                           Source: Morgan Stanley Research
                                                                                   While the government would like to continue deepening the
In 2011 Brazil improved its debt profile, increasing the share                     fixed income market, fear of further pressure on the
of inflation linked and fixed rate securities to 65.5% of                          exchange rate due to the search for carry may slow the rate
Federal Public Debt. They also increased the average                               at which it is willing to let this happen. As such, assuming
maturity of their debt to 3.6 years, the highest value since                       that liquidity remains high in 2012 as DM markets keep rates
2002. New Issuances had an average maturity if 4.3 years                           low, we would expect the government to look for ways to
compared to 3.6years average maturity for outstanding debt.                        reduce interest in domestic debt instruments as much as
In its 2012 debt plan the Government outlines its plans to                         their budget deficit will allow.
replace floating-rate securities by fixed-rate and inflation-                      Exhibit 85
linked instruments, to increase the average maturity of the                        Brazil Domestic Maturity Profile in 2012 (R$bn)
outstanding debt and smooth the maturity profile, with                                140                                                        132.7
special attention given to short-term maturities. The Brazilian                             113.6
                                                                                      120
authorities also seek to develop the yield curve on both
                                                                                      100
domestic and external markets and grow the liquidity of
                                                                                       80
federal government securities on the secondary market.
                                                                                       60
                                                                                                                                                         45.1
As a result of a successful strategy adopted by the National                           40
                                                                                                              34.8                                              39.0

Treasury throughout recent years, the outstanding EFPD                                 20
                                                                                                                         23.5
                                                                                                                                                                                 11.6
                                                                                                      8.0                         7.7                                      7.4
(External Federal Public Debt) has been systematically                                                                                    1.1
                                                                                        0
shrinking in the past years, reaching US$ 44.4bn in 2011 (in
                                                                                                                                12




                                                                                                                                                    2
                                                                                           12




                                                                                                                                         12
                                                                                                     12


                                                                                                             12




                                                                                                                                                              12




                                                                                                                                                                                 2


                                                                                                                                                                                 2
                                                                                                                                                                       2
                                                                                                                                                  12
                                                                                                                     12




                                                                                                                                                 l-1




                                                                                                                                                                               -1


                                                                                                                                                                               -1
                                                                                                                                                                     -1
contrast with US$ 55.0bn, in 2010) and representing only
                                                                                                                                                            p-
                                                                                                                              -
                                                                                         n-




                                                                                                                                       n-
                                                                                                  b-


                                                                                                              -




                                                                                                                                                g-
                                                                                                                    r-




                                                                                                                                                                            ov


                                                                                                                                                                            ec
                                                                                                                           ay
                                                                                                           ar




                                                                                                                                                                   ct
                                                                                                                                               Ju
                                                                                       Ja




                                                                                                                                     Ju
                                                                                                Fe




                                                                                                                  Ap




                                                                                                                                                          Se
                                                                                                                                              Au




                                                                                                                                                                   O
                                                                                                          M




                                                                                                                                                                           N


                                                                                                                                                                           D
                                                                                                                          M




                                                                                            Fixed Rate             Exchange Rate                Floating Rate      Inflation Linked
4.5% of total Federal Public Debt (FPD).
                                                                                   Source: Morgan Stanley Research, Brazil National Treasury
Exhibit 83
FBP Assumed Borrowing 2011 vs. Actual Issuance
                                                                                   Even though the central bank continues to ease monetary
        Gross Borrowing                                        2011      2012
         Requirements                 2011 Forecast           Actual   Forecast    policy, we have seen an offset to some extent by the
              R$bn                        365.6                 489    498-598bn   government’s commitment to maintaining a primary surplus
Source: Brazil Ministry of Finance, Morgan Stanley Research
                                                                                   of 3.2% of GDP. But in spite of our expectation of a slightly
Simulations by the Brazilian authorities indicate a final 2012                     lower public sector deficit this year (-2% in 2012 versus -
volume of FPD stock between R$1.95tr and R$2.05 trillion,                          2.7% in 2011), issuance is set to grow at around 10% in
compared to R$1.87tr in 2011. These limits imply an                                2012 vs. 2011.
issuance in 2012 of between R$ 497bn and 597bn, covering
a principle and interest payment R$ 413bn and an increase
in the public debt stock of between R$84 and 184bn.

We expect to see foreign holdings of Brazilian debt continue
to increase, presently at around 10% as percent of total
outstanding (or closer to 35-37 as a percentage of NTN-Fs
and LTNs).




                                                                                                                                                                                        31
               MORGAN STANLEY RESEARCH

               March, 2012
               EM Local Markets Guide




Asia Markets




                                         32
                                                                                                                                                  MORGAN STANLEY RESEARCH

                                                                                                                                                  March, 2012
                                                                                                                                                  EM Local Markets Guide




China
Market Overview                                                                                                                                   Of these, the Interbank market is the most dominant with
                                                                                                                                                  90% of the total amount outstanding and transactions on this
Over the past three decades, China has grown at an average
                                                                                                                                                  market. This market is a quote-driven OTC market with
rate of about 9.7% per year to become the world’s second
                                                                                                                                                  institutional investors as the main participants. The interbank
largest economy in 2010. Currently China’s GDP stands at
                                                                                                                                                  market is supervised by the People’s Bank of China (PBoC)
USD 5.9 trillion , making it highly influential in the global
                                                                                                                                                  and is also the market for all Open Market Operations
marketplace. As the economy grows and foreign investors
                                                                                                                                                  (OMOs) with CPs, MTNs and Central Bank Bill trading only in
get increasing access to the local currency markets, the
                                                                                                                                                  this market.
Chinese market is expected to develop both in terms of
market size and complexity.                                                                                                                       The Exchange market (on Shanghai and Shenzhen stock
                                                                                                                                                  exchanges) is supervised by China Securities Depository &
The Chinese debt capital market has developed significantly
                                                                                                                                                  Clearing Corporation Ltd. (CSRC) and is the market for
over the past decade into a greater than RMB 20 trillion
                                                                                                                                                  small-scale and individual investors. Treasury securities
(~USD 3 trillion) market, the largest among the emerging
                                                                                                                                                  account for most of the trading in this market. Commercial
economies of Asia. Increasingly more sophisticated fixed
                                                                                                                                                  banks do not trade on the exchange.
income instruments are also being introduced.
                                                                                                                                                  The Bank counter market mainly involves individual investors
The bond market in China is dominated by Government
                                                                                                                                                  and there are very few securities, such as government bonds
issued bonds. Moody’s upgraded China’s sovereign debt
                                                                                                                                                  being traded on this market.
ratings for foreign and local currency bonds to Aa3 from A1
in November 2010 (S&P: AA-, Fitch AA- for long-term local                                                                                         The China Government Securities Depository Trust &
currency debt). The country’s strong external payments                                                                                            Clearing Co. (CGSDTC) is the general custodian and the
position and capital controls has helped it remain resilient in                                                                                   central depository for the entire fixed income market. It is
the face of the recent economic crisis.                                                                                                           responsible for the registration and settlement of all bonds.

Exhibit 86
                                                                                                                                                  Monetary Policy Framework
Rising Trend of Domestic Debt Outstanding
                                                                                                                                                  The PBoC states its monetary policy target is to “maintain
    3500            Size of Chinese LCY Bond Market (in USD Bn)
                                                                                                                                                  stability in the value of currency and thereby promote
    3000                                                                                                                                          economic growth”. 8
    2500
                                                                                                                                                  PBoC manages monetary policy through both the long-term
    2000                                                                                                                                          1 year benchmark lending and deposit rates and the short-
    1500
                                                                                                                                                  term rediscount rate. PBoC also conducts open market
                                                                                                                                                  operations through repos and the issuance of PBoC bills.
    1000
                                                                                                                                                  The central bank is committed to price and currency stability
     500                                                                                                                                          as its target, before the growth target.
       0
                                                                                                                                                  In March 2012, China’s main economic planning agency set
           Jan-98

                    Jan-99

                             Jan-00

                                      Jan-01

                                               Jan-02

                                                        Jan-03

                                                                 Jan-04

                                                                          Jan-05

                                                                                   Jan-06

                                                                                            Jan-07

                                                                                                     Jan-08

                                                                                                              Jan-09

                                                                                                                       Jan-10

                                                                                                                                Jan-11

                                                                                                                                         Jan-12




                                                                                                                                                  the 2012 growth target as 7.5% and inflation target as 4%.
                                  Government Bonds                                   Corporate Bonds
                                                                                                                                                  Monetary Policy Instruments
Source: Asian Bonds Online
                                                                                                                                                  PBoC lending and deposit rates: 1 year benchmark rates
Structure of the Chinese Bond Market                                                                                                              that PBoC charges as lending and deposit rates for banks or
                                                                                                                                                  other depository institutions (Bloomberg: CHLR12M Index for
The Chinese bond market is structured into the Interbank
                                                                                                                                                  1yr lending rate, CNDR1Y Index for 1yr deposit rate).
market, the Exchange and the Bank counters.




    Reference: IMF Data and Statistics                                                                                                            * Reference: PBoC website


                                                                                                                                                                                                               33
                                                                 MORGAN STANLEY RESEARCH

                                                                 March, 2012
                                                                 EM Local Markets Guide




Rediscount rate: It is the benchmark central bank lending         traded, but can be used as collateral or sold back to banks
rate charged to banks for short-term borrowing. The               before maturity.
rediscount policy primarily aims at influencing the commercial
                                                                  PRC Certified Treasury Bond Certificates are a type of
paper market. (Bloomberg: CNDSC Index).
                                                                  savings bonds which are sold by banks and post office
Reserve requirements: The PBoC uses a differentiated              savings counters. These bonds are not traded but can be
required reserve ratio system whereby it can set different        sold back to the seller before maturity.
reserve requirements for different financial institutions
                                                                  Other Government bonds include local government bonds/
depending on their size and macro-prudential criteria. The
                                                                  municipal bonds which are book entry bonds issued by
PBoC can also change the interest rates on these reserves
                                                                  state/provincial governments.
as a monetary policy tool. (Bloomberg: CHRRDEP Index,
Reserve interest rates: CHIRRR Index).                            Exhibit 87
                                                                  Investor Profile of Central Government Bonds
Open market operations include repurchase and outright
                                                                                                                        Others
market operations for government bonds; central bank bills           100%

and financial bonds from policy banks. Repurchase                     90%                                               Exchanges

operations include repo and reverse repo for managing                 80%
                                                                                                                        Individuals
monetary liquidity. Since 2003, repos have been used                  70%
                                                                                                                        Funds Institutions
actively as a tool to sterilize the foreign exchange purchases        60%

that became necessary to maintain the RMB de-facto peg.               50%                                               Insurance Institutions

                                                                      40%
PBoC Bill Rate: The PBoC issues bills to withdraw market                                                                NBFIs
                                                                      30%
liquidity and injects liquidity through redemptions. The
                                                                                                                        Credit Cooperative
maturities range from 1 month to 3 years. Due to active               20%
                                                                                                                        Banks
                                                                                                                        Commercial Banks
trading in PBoC bills, their yield cut-off makes an important         10%

benchmark for interest rates in the money market. The daily            0%
                                                                                                                        Special Members
                                                                            97     99   01     03   05   07   09   11
quotes on bill yields are taken from different banks, and the
                                                                  Source: Asian Bonds Online
average is calculated as the 3m PBoC bill yield fixing.
(Bloomberg: CNBI3MO Index for 3 month PBOC Bill Yield).
                                                                  Government Bond Auctions
USD/CNY fixing: Used as an instrument for managing
                                                                  Treasury bonds and bills are issued by the Ministry of
liquidity, capital controls are implemented through a fixed
                                                                  Finance (MOF). The MOF announces the issuing schedule
exchange rate regime which enables China to have an
                                                                  for its treasury bonds at the beginning of the year. Weekly
autonomous monetary policy.
                                                                  Dutch-style auctions are held for issuance of bonds of
                                                                  maturities from 1-7 yrs as well as for longer maturities (10y,
Bond Markets                                                      15y, 20y, 30y, 50y)
Government Bonds                                                  Government securities are sold in the primary market in two
Chinese government bonds (CGBs) or Treasury bonds, are            ways:
issued by the Ministry of Finance (MoF) for government debt                     Mandatory allocation to Government banks for
financing. The interest earned on government bonds is tax                        bonds to be distributed to individuals;
exempt.                                                                         Bonds auctioned to underwriting syndicates and
Book-entry bonds make up more than 98% of the total                              primary dealers.
government bonds outstanding. These are issued and traded         Corporate Bonds
mainly in the interbank and exchange markets targeting
institutional investors.                                          Local state-owned enterprises (SOEs) are the major issuers
                                                                  of corporate debt in China. Most corporate bond issuance by
Electronic savings bonds mainly target the individual             SOEs has to be approved by National Development Reform
investor. These provide a platform for household savings to       Commission (NDRC) and require a bank guarantee.
be translated into National funds. These bonds are not            However, starting 2007 the Chinese Security Regulatory
                                                                  Commission (CSRC) allowed listed corporations, satisfying
                                                                                                                                             34
                                                                MORGAN STANLEY RESEARCH

                                                                March, 2012
                                                                EM Local Markets Guide




criteria of minimum net asset levels of CNY 1bn and record       Investors
profitability, to issue bonds to the public on the exchanges,
                                                                 The investors in the Chinese local currency bond market
without bank guarantees.
                                                                 include commercial banks, insurance companies, mutual
Corporate CPs and MTNs                                           funds, credit cooperatives, securities companies, national
Commercial paper can be issued by non- financial                 social security funds, pension funds, non-bank financial
institutions which are members of NAFMII (National               institutions, QFIIs and individual investors.
Association of Financial Market Institutional Investors).        Derivatives
Corporate CPs and MTNs can be issued without NDRC
approval or bank guarantees, and are traded in the interbank     The fixed income derivatives market in China is currently
market. CP and MTN markets are among the most active             limited to currency and interest rate futures and swaps. Only
Chinese fixed income markets with commercial banks and           CBRC approved financial institutions can deal in derivatives
mutual funds as the main investors. Maturities vary between      transactions. Bond forward transactions are limited to central
3 to 30 years.                                                   government bonds, central bank bonds, financial bonds, or
                                                                 other bonds authorized by PBoC.
Financial bonds
                                                                 Onshore IRS
Policy bank bonds represent more than 85% of the financial
bond market. These are issued by the three policy banks of       The CNY IRS market has 3 benchmark rate swap curves
China, which are owned by the central government –               with 3 different underlying fixings:
namely, China Development Bank, Export-Import Bank of                                                       
                                                                          -      7 day repo fixing
China and Agricultural Development Bank of China. These
bonds are auctioned by the China Government Securities                    -      3mth SHIBOR rate
Depository Trust and Clearing Company (CGSDTC) and are                    -      1yr PBoC deposit rate
actively traded in the interbank market.
                                                                 The most liquid among these is the 7-day repo swap
Asset Backed Securities                                          accounting for over 60% of the overall turnover in the IRS
The Chinese ABS market is still in the early stages of           market; although the PBoC has been encouraging banks to
development. This market started in December 2005 when           transact more in Shibor IRS. However, it should be noted
the first MBS bond was launched by China Construction            that since bank deposit and lending rates are largely set
Bank and later listed on the interbank market. The               administratively in China, both the Shibor and 7-day repo
securitized market grew through 2007-08 with new issuance        fixings have limited resemblance to balance sheet risks
reaching RMB 60 billion. These instruments can be traded in      faced by banks and corporates (for the latter, the 1yr deposit
the OTC and exchange markets, but are currently very             rate would be the most appropriate rate). Also, since banks
illiquid.                                                        generally carry bonds on a hold to maturity basis, there is
                                                                 limited hedging of bond trading books using swaps, though
Panda Bonds                                                      this is slowly changing.
These are RMB denominated bonds that can be issued by            Foreign investors are not allowed to participate in the
international development institutions in the Chinese capital    onshore swap market.
market. This market is very small and illiquid with only four
issues outstanding.                                              For terms 1y and above: Quarterly fixed and floating rates,
                                                                 Day convention: Act/365, Value date: T+1
CNH Bonds
                                                                 For terms below 1y: simple money market yield is used.
The evolution of the offshore RMB market has led to the
development of an offshore Dim Sum (CNH) bond market
denominated in CNH. These bonds can be issued by
                                                                 
                                                                     The 7- day repo rate is the benchmark rate for the interbank market making it the most
financial institutions and corporate entities outside China,          widely used funding instrument. It is used as the fixing rate for both onshore and offshore
with the proceeds kept offshore. (for details on the CNH              swaps. (Bloomberg: CNRR007 Index).

bond market see: Asia Credit Strategy: CNH Bonds - The                Shanghai Interbank Offer Rate (SHIBOR) is the weighted average of the lending rates
                                                                      offered by a group of domestic banks with high credit rating (currently 16 commercial
Birth of a Global Chinese Corporate Market, May 6, 2011).             banks are included in the quotation group). (Bloomberg: SHIFON Index for Overnight
                                                                      SHIBOR).
                                                                                                                                                             35
                                                                 MORGAN STANLEY RESEARCH

                                                                 March, 2012
                                                                 EM Local Markets Guide




Cross Currency Swaps: The State Administration of                 funds and insurance funds that are required to invest for a
Foreign Exchange (SAFE) has allowed a number of                   minimum period of 3 months.
designated commercial banks to trade RMB cross currency
                                                                  QFII’s are required to apply with the State Administration on
swaps through the China Foreign Exchange Trade System
                                                                  Foreign Exchange to purchase foreign exchange for
(CFETS), with their clients, starting from 1 March 2011. The
                                                                  repatriation of principal. The amount of the application
applicable interest rates under a RMB-FX cross currency
                                                                  cannot exceed 20% of the total principal, with each
swap must be agreed between the bank and its client within
                                                                  application at least one month apart.
the parameters of deposit and lending rates set by the PBoC.
Meanwhile, banks in Hong Kong have already started                A RMB QFII pilot scheme launched in December 2011
offering cross-currency swaps, with RMB - fixed and US            allows offshore RMB funds to be invested in the domestic
dollar floating legs for tenors up to five years, with one and    securities market. The pilot scheme allows for an initial quota
two years as the most liquid. As China's capital account          of RMB 20Bn, requires that at least 80% of funds be invested
remains largely closed, the onshore and offshore rates            in fixed income products, and is initially open to HK-
quoted for RMB derivatives are governed by different sets of      subsidiaries of China asset management and securities firms
market dynamics.                                                  only.
Non deliverable IRS (NDIRS) and NDCCS: The non                    Eligibility
deliverable contract is traded offshore and net settled in USD
                                                                          Fund management companies
(FX rate for settlement is the PBoC fixing)
                                                                          Insurance companies
NDIRS day count convention: Quarterly fixed rate Act/365 vs
Quarterly 7-day repo Act/365                                              Other institutional investors including pension funds,
                                                                           trust companies, charitable foundations and
NDCCS day count convention: Semi annual Act/365 vs 6m
                                                                           donation foundations
US Libor Act/360.
                                                                  The eligibility criteria for the above investor types include at
Credit Default Swaps: The Chinese CDS market was
                                                                  least five years of experience in asset management,
launched in November 2010 as a ‘management tool’ for
                                                                  managing a minimum of USD 5 billion of securities assets in
credit risk hedging purposes. There is a limit on the leverage
                                                                  the most recent accounting year.
amount used in these swaps. Only a few major banks are
allowed to trade in these credit derivatives.                             Securities companies with at least 30 years of
                                                                           experience in securities operation, no less than
Qualified Foreign Institutional Investor                                   USD1 billion of paid-in capital, managing a
                                                                           minimum USD10 billion of securities assets in the
Foreign Individual investors are restricted from trading in
                                                                           most recent accounting year
local currency bonds.
                                                                          Commercial banks which are ranked among the
Foreign Institutional investors that are registered with the
                                                                           Top 100 in the world by total assets, managing no
PRC's qualified foreign institutional investor (QFII) system
                                                                           less than USD10 billion of assets
are allowed to invest in local currency bonds listed on the
Shanghai and Shenzhen stock exchanges. A QFII must open           Instruments available for FIIs
a special yuan account for trading in the securities and bond
                                                                  The following fixed income instruments are available for QFII
markets. This special account is subject to an investment
                                                                  investment:
quota issued by the State Administration of Foreign
Exchange (SAFE) that ranges from USD 50 million to USD                    Treasury, corporate and financial bonds, CPs and
800 million equivalent.                                                   MTNs listed in the exchange market

QFII investments in the securities market should have a                   Units of mutual funds
minimum period of one year. The lockup period is shorter for
                                                                          Other financial instruments approved by CSRC
some medium- and long-term QFII funds such as pension



    Reference: Asia Bonds Online
                                                                                                                                36
                                                                                MORGAN STANLEY RESEARCH

                                                                                March, 2012
                                                                                EM Local Markets Guide




                 
Taxation
Tax laws are always complex, and clients should seek
appropriate tax advice. However, as a broad guide, our
reading is that the following basic tax rules apply.

              Onshore bonds are subject to taxes as stated by
               the Chinese State Administration of Taxation (SAT)
               in January 2009.

              Interest income from government and most financial
               bonds is tax exempt. In general, interest earned on
               corporate bonds is subject to a 20% withholding tax,
               though it is subject to different tax rates for
               individual and institutional investors.

              Capital gains are generally subject to an income tax
               of 25% and a business tax of 5%.

              Foreign investors covered by double tax treaty may
               derive reduced capital gains tax.

Useful Websites
      People’s Bank of China
      www.pbc.gov.cn
      China Securities Regulatory Commission (CSRC)
      www.csrc.gov.cn
      China Banking Regulatory Commission
      www.cbrc.gov.cn
      State Administration of Foreign Exchange (SAFE)
      www.safe.gov.cn
      China Bond
      www.chinabond.com.cn





    * Morgan Stanley is not a tax adviser. Consult a professional adviser for
      further guidance.
                                                                                                          37
                                                                         MORGAN STANLEY RESEARCH

                                                                         March, 2012
                                                                         EM Local Markets Guide




India
Market Overview                                                              with the success ratio of T-Bills, underwriting commitment,
                                                                             and bidding commitment prescribed by RBI.
In the last two decades, the Indian economy has transformed
from a closed and controlled economy to a more open,                        Market making in government securities: Primary dealers
liberalized and one of the fastest-growing economies in the                  offer two-way prices in government securities. They carry
world. The gross domestic product (GDP) of India has been                    out their transactions through NDS-OM, OTC or through
growing steadily (around 7-8%). India’s GDP in 2010-11 was                   the NSE or BSE exchanges.
approximately INR52 trillion (US$1.7 trillion). With continuing
                                                                         NBFCs are financial institutions that provide non-banking
increase in access for foreign investors to the local currency
                                                                         services. They do not hold a banking license. NBFCs typically
bond markets, we expect to see more inflows in the high-
                                                                         participate in loans and credit facilities, retirement planning,
yielding fixed income market.
                                                                         money markets, underwriting, and merger activities.
The Indian bond markets have matured significantly in the
                                                                         Other than these institutions, mutual funds and insurance
past decade. There has been a marked increase in the depth
                                                                         companies are also major players in the bond markets. Mutual
and width of the secondary bond market, both in terms of size
                                                                         funds generally invest up to 80-90% in short-term bonds.
and turnover. Also, more sophisticated fixed income
                                                                         Insurance companies generally buy long-term bonds ranging
instruments are being introduced over time. Foreign investors
                                                                         from 10-20 years maturity. Provident funds typically do the
hold approximately 2.6% of the government bonds
                                                                         same and hold the bonds until maturity.
outstanding, while their holdings in equity markets are much
higher at 18% as of March 2011.
                                                                         Monetary Policy Framework
The Indian bond market is dominated by government
                                                                         The basic functions of the Reserve Bank of India are to:
securities, though corporate and financial institutions also
issue a large number of bonds regularly. The Indian                              regulate the issue of currency notes
government local currency debt rating was upgraded by
                                                                                 keep reserves to provide monetary stability
Moody’s from Ba2 to Ba1/Positive (in 2010), owing to the
decline in the fiscal deficit and strong GDP growth prospects.                   centralize cash reserves of commercial banks
Indian foreign currency bond ratings stand at Baa3/Stable.
                                                                                 manage the credit system of the country
(Fitch: BBB-, S&P: BBB-u).
                                                                         The RBI has the twin objectives of growth and price stability
However, investment risks such as high inflation and
                                                                         but the relative emphasis between the two tends to vary
moderate secondary market trading liquidity remain.
                                                                         depending upon the broader economic outlook. Unlike some
                                                                         other central banks, the RBI does not have an explicit inflation
Financial Structure
                                                                         target framework.
India’s central bank is called the Reserve Bank of India (RBI).
The main players of the finance industry include banks,                  Monetary Policy Instruments
Primary Dealers (PDs) and Non-banking Financial
                                                                         Repo Rate: The repo rate is the rate at which the RBI injects
Corporations (NBFCs), Insurance companies, Mutual funds
                                                                         liquidity or lends overnight money to the banks. This
and Provident/ Pension funds.
                                                                         instrument is the operational rate used for injections of
A primary dealer (PD) ‡‡‡ is a market maker for government               liquidity.
securities. They have two key roles:
                                                                         Reverse Repo Rate: The reverse repo rate is the rate at
   Support to primary market: PDs are required to support               which the RBI absorbs liquidity or at which banks can lend
    auctions of government dated securities and Treasury Bills           excess cash to the RBI overnight. This instrument is used to
                                                                         absorb liquidity.

‡‡‡
    List of primary dealers in India:
                                                                         LAF Corridor: The LAF corridor is the difference between the
http://www.rbi.org.in/scripts/AboutUsDisplay.aspx?pg=PrimaryDealer.htm   repo and the reverse repo rates. The LAF corridor is used to

                                                                                                                                      38
                                                                        MORGAN STANLEY RESEARCH

                                                                       March, 2012
                                                                       EM Local Markets Guide




reduce volatility in the overnight funding market. If the demand       major participants in this market are nationalized banks,
for money is lower than supply, then the average overnight             private banks, foreign banks, co-operative banks, financial
funding rate for banks is typically closer to the reverse repo;        Institutions, Insurance Companies, Mutual Funds, Primary
similarly, if demand is higher than supply, then it is closer to       Dealers, Bank cum Primary Dealers, NBFC, Corporate,
the repo rate.                                                         Provident/ Pension Funds.

The LAF corridor was reduced by 50bp to 100bp in 2010,                 Liquidity Adjustment Facility (LAF): The LAF allows banks
which has reduced the volatility in money markets.                     to borrow or lend to the RBI overnight at the repo or reverse
                                                                       repo rate, respectively, to help with short-term cash
Cash Reserve Ratio (CRR): The cash reserve ratio is the
                                                                       management. Banks use eligible securities (such as G-secs)
proportion of deposits banks are required to hold in the form
                                                                       as collateral for the repo agreement. The facility is used to
of cash with the RBI. The RBI uses the CRR to manage
                                                                       take out excess liquidity or infuse liquidity during shortfalls.
liquidity in the banking industry.
                                                                       The effective rate can be repo rate or reverse repo rate
Statutory Liquidity Ratio (SLR): The statutory liquidity ratio         depending on the liquidity.
is the minimum percentage of net demand and time liabilities           Eligibility: All commercial banks and PDs having a current
(NDTL) that banks must maintain in the form of gold, cash or           account and SGL account with the RBI.
other approved securities such as government bonds. The                Minimum bid size: INR50 million and in multiples of INR50
RBI uses the SLR to regulate credit growth in India.                   million.

Marginal Standing Facility (MSF): The marginal standing                Mumbai Interbank Offer Rate (MIBOR): MIBOR is the rate at
facility is collateralized borrowing, designed to curb volatility in   which local banks can lend and borrow from each other
the overnight lending rates. The MSF allows banks to borrow            overnight, unsecuritized. MIBOR is determined daily through a
overnight from RBI up to an additional 1% of their NDTL at             poll of local participants such as banks and PDs. It is a
100 basis points above the repo rate. This facility would be           benchmark rate for most interest rate swaps, forward rate
used by banks as a last resort to raise money given the high           agreements, floating rate debentures, and term deposits.
interest rate associated with it.                                      Under tight liquidity, MIBOR is generally 10-15 bps above the
                                                                       repo rate; similarly, under easy liquidity, MIBOR is typically
Money Markets                                                          10-15 bps above the reverse repo rate. During times of
                                                                       extreme funding squeezes, such as fiscal year-end, MIBOR
Call/Notice Money Market: Under the call money market,
                                                                       can fix >100 bps above the repo rate.
banks can borrow from or lend to each other overnight on an
uncollateralized basis, and under the notice money market,             Exhibit 88
banks can borrow/lend for a period of between 2 days and 14            LAF Corridor and O/N MIBOR
days, also on an uncollateralized basis.                                 %                                                                INR bn
                                                                        18                                                                 2000
Banks use this money market for the following purposes:
                                                                        16                                                                1500
• To fill gaps or temporary mismatches in funds;                        14                                                                1000
                                                                        12
• To meet the CRR & SLR mandatory requirements as                                                                                         500
                                                                        10
stipulated by the RBI;                                                                                                                    0
                                                                         8
• To meet sudden demand for funds arising out of large                   6
                                                                                                                                          -500
outflows.                                                                4                                                                -1000

Thus, call money is generally used to balance banks’ short-              2                                                                -1500

term liquidity positions. Participants in the call/notice money          0                                                               -2000
market currently include mainly banks and PDs.                           Feb-07        Feb-08        Feb-09   Feb-10   Feb-11       Feb-12
                                                                                      O/N MIBOR (21D Avg)              Reverse Repo -Yield
Collateralized Borrowing and Lending Obligation (CBLO):                               Repo - Yield                     Net Liq Injection (RHS)
A CBLO is a fully collateralized and secured instrument for
                                                                       Source: Bloomberg
borrowing/lending money, and was developed by the Clearing
Corporation of India Ltd. (CCIL). The CCIL is the counter-
party on all CBLO trades and guarantees settlement. The

                                                                                                                                          39
                                                                    MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Bond Markets                                                       Government of India. These are discounted securities and
                                                                   thus are issued at a discount to face value. The return to the
The Indian bond markets operate from 9am to 5pm. Some of
                                                                   investor is the difference between the maturity value and
the major securities traded in this market are described below.
                                                                   issue price. Treasury Bills are issued for the following tenors:
Government Bonds                                                   91 days, 182 days and 364 days. The issuance calendar is
                                                                   typically pre-determined at the start of each half of the fiscal
Types of Government Securities:
                                                                   year. The key features of these securities are:
Dated Security (Fixed rate): Dated securities are fixed                     Highly liquid money market instrument
maturity and fixed coupon securities and usually pay a semi-                Transparency
annual coupon. These are identified by their date of maturity               Simplified settlement
and coupon. The key features of dated securities are:              Average Daily Volume: INR 7-10 billion
       Issued at face value                                       Outstanding Amount: INR1.66 trillion
       Coupon is fixed at the time of issuance and is             Cash Management Bill: Cash Management Bills are short-
        constant until redemption                                  term instruments issued by the RBI to cover temporary
       Maturity date is fixed                                     mismatches in the government’s cash flows. CMBs have the
       Redeemed at par on the maturity date                       generic character of T-Bills but are issued for maturities of
Average Daily Volume: INR100-120 billion                           less than 91 days. Like T-Bills, they are also issued at a
Outstanding Amount: INR20.1 trillion                               discount and redeemed at face value at maturity. The tenure,
                                                                   notified amount and date of issue of the CMBs depend upon
Floating Rate Bond: Floating rate bonds have a variable
                                                                   the temporary cash requirement of the government. The
coupon rate, usually a fixed margin over a benchmark rate.
                                                                   announcement of their auction is made by RBI through a
There may be a cap and a floor rate attached, thereby fixing a
                                                                   press release, which is typically issued one day prior to the
maximum and/or minimum coupon. The key features of these
                                                                   date of auction.
securities are:
                                                                   Outstanding Amount: No CMBs outstanding as of Dec 2011
       Issued at face value
       Coupon is a fixed margin over a predefined                 Exhibit 89
        benchmark rate                                             Commercial Banks and Insurance Companies Have
       Maturity date is fixed                                     Been the Biggest Owners of Central Government
       Coupon is paid semi-annually                               Bonds
       Redeemed at par on the maturity date                           100%                                                  Others
Average Daily Volume: Illiquid                                                                                               RBI
Outstanding Amount: INR493.5 billion                                    80%                                                  Provident Funds
                                                                                                                             FIIs
Oil Bond: Oil bonds are issued by the Government of India to            60%                                                  Corporates
fund the losses of the public sector oil companies who sell                                                                  Financial Institutions
petroleum products at a loss due to the price regulation of             40%                                                  Co-Operative Banks
energy products. The key features of these securities are                                                                    Mutual Funds
                                                                        20%
similar to fixed rate bonds. Only the nomenclature is different.                                                             Primary Dealers
                                                                                                                             Insurance Companies
Average Daily Volume: Illiquid (INR0-500 million)                        0%
                                                                                                                             Commercial Banks
                                                                                Mar-07


                                                                                         Mar-08


                                                                                                  Mar-09


                                                                                                           Mar-10


                                                                                                                    Mar-11




Outstanding Amount: INR1441.8 billion

Fertilizer Bond: Fertilizer bonds are similar to oil bonds,
                                                                   Source: Morgan Stanley Research
issued by the government to fertilizer companies as
compensation towards fertilizer subsidy.
Average Daily Volume: Illiquid
Outstanding Amount: INR275 billion

Treasury Bills: Treasury Bills are money market instruments
used to finance the short-term requirements of the
                                                                                                                                                      40
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Investors                                                                  are arranged in descending order and the successful
                                                                           bidders are those who have bid at or above the cut-
The major investors in Indian Government bonds over the
                                                                           off price.
past decade have been commercial banks and Life Insurance
Corporation of India (LIC) (as shown in Exhibit 89). Apart from   Depending upon the method of allocation to successful
these, other investors have been the EPF Scheme, mutual           bidders, auctions can be classified as Uniform Price based or
funds and the RBI. The Clearing Corporation of India Ltd.         Multiple Price based.
(CCIL) releases a daily net buying/selling of the Central
                                                                          Uniform Price auction: All the successful bidders
Government bonds broken by the type of investors i.e.,
                                                                           are required to pay for the allotted quantity of
Foreign Banks, Public Banks, Private Banks, Mutual Funds,
                                                                           securities at the same rate, i.e., at the auction cut-off
Primary Dealers and Others. (Bloomberg ticker: CCIL 10)
                                                                           rate, irrespective of the rate quoted by them. This
                                                                           type of auction is typical for Government bonds.
Government Bond Auctions
                                                                          Multiple Price auction: All the successful bidders
Indian government bonds are issued through auctions
                                                                           are required to pay for the allotted quantity of
conducted on the electronic platform called the NDS-Auction
                                                                           securities at the respective price / yield at which they
Platform. Participants in these auctions are:
                                                                           have bid. This type of auction is typical for T-Bills and
        Commercial Banks                                                  CMBS.
        Scheduled Urban Co-Operative Banks
        Primary Dealers                                          Corporate Bonds
        Insurance Companies                                      The corporate bond market has been in existence in India for
        Provident Funds                                          a long time. The size of the public issue segment of the
        Individuals                                              corporate bond market in India has remained quite
                                                                  insignificant. The lack of market infrastructure coupled with
Only those who have a fund account and a securities account
                                                                  low issuance is the reason for low liquidity.
(SGL) with the RBI are allowed to participate. All non-NDS
members participate through SCB or PDs by opening a Gilt          The market for long term corporate debt has two large
account with them. A Gilt Account is a dematerialized account     segments:
maintained by a scheduled commercial bank or Primary
                                                                          Bonds issued by public sector units, including public
Dealer for its constituent.
                                                                           financial institutions;
A notification and a press release giving exact particulars of            Bonds issued by the private corporate sector.
the securities, viz., name, amount, type of issue and
procedure of auction, are issued by the Government of India       Some of types of corporate bonds available in the Corporate
about a week prior to the actual date of the auction.             Bond List are:
Government bonds trade in the When Issued (WI) segment                    Convertible bonds
after the auction announcement until the close of the auction             Non convertible debentures
day. However, the volume in WI is poor.                                   Commercial paper
An auction may either be yield based or price based.                      Pass Through certificate
                                                                          Certificates of deposit
        Yield Based Auction: A yield based auction is
                                                                          Structured bonds
         generally conducted when a new Government
         security is issued. Investors bid in yield terms up to   Nonconvertible debentures: Non Convertible debentures
         two decimal places. Bids are arranged in ascending       are secured / unsecured bonds that cannot be converted to
         order and the cut-off yield is arrived at the yield      company equity or stock. These bonds predominantly carry
         corresponding to the notified amount of the auction.     fixed interest rate. However, some corporates do issue
                                                                  floating rate bonds, depending on investor appetite. The
        Price Based Auction: A price based auction is
                                                                  common benchmarks for floating rate bonds are MIBOR, OIS,
         conducted when Government of India re-issues
                                                                  T-bill and government bond yields. Most of the bonds are
         securities issued earlier. Bidders quote in terms of
                                                                  issued through the private placement route. Secondary
         price per Rs.100 of face value of the security. Bids
                                                                                                                                 41
                                                                     MORGAN STANLEY RESEARCH

                                                                     March, 2012
                                                                     EM Local Markets Guide




market for AAA rated bonds is fairly liquid. However, for lower-     Equity-linked structured notes are popular with high net-worth
rated corporate bonds, the liquidity is significantly lower.         individuals, but a secondary market in these instruments is
                                                                     almost non-existent.
Commercial Paper: Commercial Paper (CP) is an unsecured
                                                                     Average Daily Volume: Very Illiquid
money market instrument issued in the form of a promissory
note. CP, as a privately placed instrument, was introduced in        Investors in Corporate Bonds
India with a view to enabling highly rated corporate borrowers
                                                                     Mutual Funds are the most active investors in the shorter end
to diversify their sources of short-term funding.
                                                                     of the corporate curve. In the longer end, insurance
Average Daily Volume: INR30-50 billion **
                                                                     companies, provident funds and trusts are the major
Certificate of Deposit: CDs are generally issued by                  investors. Banks and primary dealers arrange the primary
commercial banks. Like commercial papers, CDs are also               issue and also trade actively in the market, providing liquidity.
discounted instruments with a specific maturity date. Usually        FIIs are active in tenors up to three years.
issued up to one year, these instruments are rated on a short-
                                                                     Interest Rate Risk Management
term debt rating scale of P1 / P1+. In general, the market
trading lot size is Rs. 5 cr, though smaller lot sizes can also be   With the phased deregulation of interest rates in India and the
issued. The volume of CDs is much higher than that for CPs           operational flexibility given to banks in the pricing assets and
since banks use CDs as a source for raising institutional            liabilities, a number of interest rate hedging derivatives have
deposits.                                                            developed and are being used in the Indian fixed income
Average Daily Volume: INR5-10 billion **   §§§                       markets. Some of these are listed below:

Convertible Bonds: A convertible bond gives the holder an            Interest Rate Swaps:
option to exchange the bond for a predetermined number of            Fixed for Floating Interest Rate Swaps
shares of the issuing company. The market for onshore INR-
                                                                        Overnight Index Swap (OIS): OIS is the most liquid
denominated convertible bonds is highly illiquid with very rare
                                                                         interest rate swap in which the underlying benchmark is
issuances. However, the USD-denominated foreign currency
                                                                         the overnight call money rate. The floating rate is MIBOR.
convertible bonds traded offshore are more liquid and are                The floating rate is reset daily and compounded. The
frequently tapped by Indian issuers.                                     settlement frequency is semi-annual for tenors greater
Average Daily Volume: Illiquid                                           than 1y and settlement takes place at the end of the swap
Pass-Through Certificate (PTC): A PTC is a securitized                   for tenors of 1y or lesser. Although OIS swaps are quoted
product issued by a SPV against the assets or mortgages                  out to ten years, the maximum liquidity is for a tenor of up
held by it. Asset-backed securities issued on the back of                to five years. Only local investors can trade onshore OIS;
                                                                         foreign investors can trade ND-OIS (as discussed below).
vehicle finance and personal finance are more popular among
                                                                         Average Daily Volume: INR80-100 billion
investors than the mortgage-backed securities. However, the
volume in this market has reduced over the last couple of               ND-OIS: NDOIS is a non-deliverable overnight index
years due to more stringent RBI regulations.                             swap net settled in USD. Offshore investors can trade
Average Daily Volume: Illiquid                                           NDOIS to hedge interest rate risk. The floating leg is
                                                                         MIBOR. NDOIS typically trades -5 to +10bp relative to
Structured Bonds: Structured Bonds are a combination of                  onshore OIS.
two elements:                                                            Average Daily Volume: as liquid as onshore OIS
    A fixed income security with maturity ranging from a few           MIFOR Swap: A MIFOR swap is a fixed/floating interest
     months to a few years, with coupon payable periodically.            rate swap with Reuters 6m MIFOR as the floating leg.
    A contract determining the coupon or value of capital at            Mumbai Inter-Bank Forward Offer Rate (MIFOR) is the
     maturity in accordance with price trends of one or more             Indian interest rate benchmark derived from USD Libor and
     financial parameter.                                                USD/INR Forward Premia. Counterparties settle the net
                                                                         proceeds between the fixed and floating leg on an agreed
                                                                         principal amount of the swap every six months in INR.
                                                                         MIFOR swaps are mainly used to hedge the local currency
                                                                         interest rate risk component of Cross-Currency Swaps.
**Values estimated by Morgan Stanley
                                                                         Average Daily Volume: Illiquid.
                                                                                                                                    42
                                                                        MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




   ND-MIFOR: FIIs and other offshore investors cannot              Foreign Institutional Investor
    trade in the onshore MIFOR market so a non-deliverable
                                                                    Registration Process 
    MIFOR (ND-MIFOR) net settled in USD has developed. It
    generally trades 10-20bp around onshore MIFOR, but is           The following entities/institutions incorporated outside India
    very illiquid.                                                  are eligible to register as an FII:
    Average Daily Volume: Very Illiquid
                                                                             Pension Funds
   INBMK Swap: An INBMK swap is a fixed INR coupon                          Mutual Funds
    versus floating 1y BMK, where BMK is the yield on the                    Insurance / Reinsurance Companies
    benchmark 1y government bond, as per the Reuters                         Investment Trusts
    fixing for Indian Government Bonds (0#INANCMTBMK=).                      Banks
    BMK swaps are typically used by corporates to convert                    Foreign Government Agencies or a Foreign Central
    fixed rate issuance to floating. It usually trades 30-40bp                Bank
    above or below the theoretical value, depending on the                   Sovereign Wealth Funds
    overall hedging demand at that time.                                     University Funds
    Average Daily Volume: Illiquid                                           Endowments
                                                                             Foundations
   Cross-Currency Swap: The onshore Cross-Currency
                                                                             Charitable Trusts / Charitable Societies
    Swap is a fixed/floating cross-currency swap with 6m US
                                                                             International / Multilateral Organization / Agencies
    Libor as the floating leg. The fixed receiver pays 6m USD
    Libor (Act/360) in exchange for INR at the fixed swap rate      Further, the following entities established or incorporated
    (Act/365). The net proceeds are converted into USD at           outside India and proposing to make investments in India on
    the then spot rate. At expiry, counterparties exchange          behalf of broad-based funds and its proprietary funds are also
    principals at the spot FX rate at the maturity date and the     eligible to be registered as an FII:
    net proceeds are settled in INR. The CCS is quoted off
                                                                             Asset Management Companies
    the MIFOR curve and the rates are usually close to
                                                                             Investment Manager or Advisor
    MIFOR rates.
                                                                             Institutional Portfolio Manager
   NDS: Offshore investors can trade in the non-deliverable                 A Trustee of a Trust
    Cross Currency Swap (NDS) which is net settled in USD.
    Like the Cross currency swap, the NDS is also is also           Eligibility: As per Regulation 6 of SEBI (FII)
    fixed (Act/365)/floating (Act/360) with 6m US Libor as the      Regulations,1995, the following considerations are taken into
    floating leg. The flows in NDS are very volatile, though        account for grant of registration as an FII:
    average daily volume is around USD 30-50 million**.                      The applicant’s track record, professional
Interest Rate Futures: The market for interest rate futures                   competence, financial soundness, experience,
has not picked up in India. Currently, futures are permitted on               general reputation of fairness and integrity.
the 10-year government bonds, with physical settlement.                      Whether the applicant is regulated by an appropriate
These are traded on the National Stock Exchange (NSE), but                    foreign regulatory authority.
the market in IR futures is very illiquid at present. The RBI has
recently permitted contracts on the 91-day treasury bills which              Whether the applicant has been granted permission
are likely to be launched by NSE in July 2011. This contract                  under the provisions of the Foreign Exchange
would be cash settled. FIIs are permitted to trade in futures                 Management Act, 1999 from the Reserve Bank of
within the FII limits for holding government bonds. The RBI is                India for making investments in India as a FII.
also planning to introduce cash settled futures on 2-year and       Documents Required
5-year government bonds.
                                                                             “Form A” as prescribed in SEBI (FII) Regulations,
                                                                              1995.


                                                                    
                                                                     Reference: SEBI website

                                                                                                                                     43
                                                                                    MORGAN STANLEY RESEARCH

                                                                                   March, 2012
                                                                                   EM Local Markets Guide




          Certified copy of relevant clauses (clauses permitting                  (iii) Long-term capital gain is taxed at 10%
           the stated activities) of Memorandum of Association,
                                                                                   (iv) Foreign investors based in a country covered by double-
           Article of Association or Article of Incorporation.
                                                                                   tax agreement with India may obtain a beneficial treatment on
          Audited financial statement and annual report for the
                                                                                   interest withholding/capital gains tax, depending on the DTA.
           last one year (period covered should not be less than
           twelve months)                                                          Exhibit 90
                                                                                   Standard Withholding Taxes across AxJ Countries
Registration Fees: The registration fee is US$5000.
                                                                                                              Standard Withholding Taxes
Validity: The FII registration is permanent unless suspended
                                                                                   India                                       20%
or cancelled by SEBI.
                                                                                   Indonesia                                   20%
Instruments Available for FIIs                                                     Korea                                       14%
          Securities in the primary and secondary markets                         Malaysia                                    15%
           including shares, debentures and warrants of                            Phillipines                                 20%
           companies listed, unlisted or to be listed on a
                                                                                   Thailand                                    15%
           recognized stock exchange in India
                                                                                   Singapore                                   15%
          Dated Government Securities and Treasury Bills                          Source: Morgan Stanley Research
          Commercial Papers
          Units of domestic Mutual Fund schemes                                   Useful Websites
           A new product popular with FIIs is fixed maturity plan                  Reserve Bank of India
           (FMP), a debt mutual fund product, which is like a                      http://www.rbi.org.in/
           fixed deposit but with a lower tax rate. FIIs can invest
                                                                                   Ministry of Finance
           in FMPs as long as the mutual fund does not invest
                                                                                   http://www.finmin.nic.in/
           in CDs.
                                                                                   Securities and Exchange Board of India
FIIs can allocate their investment between equity and debt
                                                                                   http://www.sebi.gov.in/
instruments. However, a debt-dedicated entity can also
register as a 100 % debt FII, which is permitted a higher                          The Clearing Corporation of India (CCIL)
investment ceiling in debt securities only.                                        http://www.ccilindia.com/
Limitations: The government has increased the limit on FII                         Fixed Income Money Market and Derivatives Association of
investment in government securities from US$10bn to US$15                          India (FIMMDA)
billion. The incremental US$5 billion can be invested in bonds                     http://www.fimmda.org/
without any maturity restrictions. The limit for FII investment in
corporate bonds is US$45 billion. Of this, US$25 billion is for
bonds issued by companies in the infrastructure sector with a
residual maturity of more than five years and the remaining
US$ 20 billion can be invested in corporate bonds with no
restrictions.

Taxation
Tax laws are always complex, and clients should seek
appropriate tax advice. However, as a broad guide, our
reading is that the following basic tax rules apply. 

(i) Withholding tax of 20% on interest payments
                  ****
(ii) Short-term          capital gain is taxed at 30%
                                                                                   ****
                                                                                        Short term and long term, in case of investments in listed debt securities

 Morgan Stanley is not a tax adviser. Consult a professional adviser for further   and units of mutual fund, refers to a holding period of less than 12 months and
guidance.                                                                          greater than 12 months, respectively.
                                                                                                                                                                44
                                                                                                MORGAN STANLEY RESEARCH

                                                                                                March, 2012
                                                                                                EM Local Markets Guide




Indonesia
Market Overview                                                                                  Clearing and Settlement: The settlement of securities is
                                                                                                 governed by the Capital Market and Financial Institution
Indonesian bonds are the largest high yielding market in the
                                                                                                 Supervisory Board (BAPEPAM-LK). Bank Indonesia (BI)
Asia-ex-Japan region with strong foreign investor
                                                                                                 uses the Scripless Securities Settlement System (BI-SSSS)
participation. The size of the tradable local currency
                                                                                                 for state debt securities, open market operations, and to
government bonds outstanding is IDR 735Tn (as of Jan
                                                                                                 connect to BI’s Real Time Gross Settlement System.
2012), of which foreign investors hold IDR 236Tn or 32% of
                                                                                                 Transactions are conducted via delivery versus payment.
bonds.
                                                                                                 Corporate bond clearing and settlement is conducted by PT
Indonesian bonds carry high weightings (5-10%) in most                                           Kustodian Sentral Efek Indonesia (KSEI) through the Fixed
Emerging Market and Asia-ex-Japan government bond                                                Income Trading System (FITS).
indices. Indonesia’s sovereign debt rating has been
                                                                                                 Exhibit 91
upgraded by rating agencies in 2011-12 given its resilient
                                                                                                 Local Currency vs Foreign Currency Bond Markets
growth and declining public debt ratios. Indonesia’s long term
foreign currency sovereign debt ratings are BB+ by S&P,                                            120
                                                                                                                          L/C Govt Bonds
BBB- by Fitch and Baa2 (positive Outlook) by Moody’s.                                                                     L/C Corp Bonds
                                                                                                   100
With a positive outlook by the rating agencies and                                                                        F/C Bonds
                                                                                                     80
Indonesia’s nascent status as the next BRIIC-country, we
believe that positive momentum towards Indonesia will                                                60
continue and expect more portfolio allocations from bond
funds in coming years. That said, investment risks include                                           40
high historical inflation and FX volatility as well as moderate
                                                                                                     20
secondary market trading liquidity.

Fixed Income Markets                                                                                  0
                                                                                                          2000

                                                                                                                 2001

                                                                                                                        2002

                                                                                                                               2003

                                                                                                                                      2004

                                                                                                                                             2005

                                                                                                                                                    2006

                                                                                                                                                           2007

                                                                                                                                                                  2008

                                                                                                                                                                         2009

                                                                                                                                                                                2010

                                                                                                                                                                                        2011
The government, which is the dominant issuer of bonds in
                                                                                                 Source: Asianbondsonline, Asian Development Bank
Indonesia, authorizes the Ministry of Finance (MOF) to issue
Treasury instruments, while Bank Indonesia issues Sertifikat                                     Market Ownership
Bank Indonesia (SBI), or Central Bank Bills, for liquidity
sterilization purposes. The main instruments in Rupiah                                           Banks are the major holders of government bonds, with
denominated government debt market are Fixed Rate bonds                                          ownership of 37% of all government bonds. Other investors
(FR series), Zero Coupon bonds (ZC), Treasury Bills known                                        include asset-pooling industries (e.g., mutual funds, pension
as Surat Perbendaharaan Negara (SPN), and Variable Rate                                          funds, and insurance companies), and private investors.
bonds (VR), both in conventional form and instruments                                            Breakdown of government bond ownership can be found at
subject to Sharia-law.                                                                           Indonesia Debt Management Office website
                                                                                                 (http://dmo.or.id/en).
Corporate sector issuers comprise a small percentage of the
bond market, representing less than 15% of total bonds                                           Foreign investors in government bonds mainly consist of real
outstanding as of December 2011. Corporate debt                                                  money investors, while hedge funds typically buy SBIs and
instruments include bonds issued by both private and state                                       short-term Treasury paper for FX appreciation purposes.
owned companies. Corporate issues are also offered in the                                        Since 2009, there have been enormous inflows into the
form of conventional bonds or shari'a-compliant                                                  government bond market, raising total foreign ownership to
instruments. 12 Most corporate bonds are listed on the                                           31% of the total debt outstanding.
Indonesia Stock Exchange (IDX).




12
     Shari’a securities are structured as revenue-sharing instruments that conform to shari’a
     principles by not recognizing interest payments. These securities—including drafts,
     shari’a bonds, and shari’a mutual fund certificates—are usually traded in money and
     capital markets.


                                                                                                                                                                                       45
                                                                                   MORGAN STANLEY RESEARCH

                                                                                   March, 2012
                                                                                   EM Local Markets Guide




Exhibit 92                                                                          FX
Investor Profile of Indonesian Government bonds
                                                                                    IDR is a non-convertible currency, and is traded by offshore
 100%                                                         Sharia Bank           investors using non-deliverable forwards. The IDR used to
   90%                                                        Bank of Indonesia     be one of the most volatile currencies in Asia. However, BI
   80%                                                        Others                has successfully smoothed the volatility of spot FX since
   70%                                                        Security              2009 through active FX interventions. The lower level of IDR
   60%                                                        Pension Fund          volatility provides foreign investors with more confidence to
   50%                                                        Foreign Holders       trade IDR NDFs too. Onshore-offshore yield differentials also
   40%                                                        Insurance             provide decent arbitrage opportunities, which investors and
   30%
                                                              State Recap Bank
                                                                                    corporates use to establish cheap FX hedges and as a
   20%                                                                              source of portfolio alpha.
                                                              Private Recap Bank
   10%
                                                              Non Recap Bank        Derivatives
    0%
                                                              Regional Bank
                                                                                    The IRS market is very illiquid and the onshore and offshore
    03

    04

    05

    06

    07

    08



    10
    09



    11




                                                              Mutual Fund
                                                                                    CCS is used instead, as a proxy by banks for hedging their
  20

  20

  20

  20

  20

  20

  20

  20

  20




Source: Asianbondsonline, Asian Development Bank                                    liabilities. Average daily volume for CCS is IDR 10-15 mn,
                                                                                    although, flows can be abrupt, with daily trade volumes
Auctions
                                                                                    increasing to IDR 50-60mn on the back of notes issuance or
T-Bill and T-Bond auctions are usually held twice a month (in
                                                                                    corporate flows. Settlement in the derivatives market is on a
tenors varying from 1y-30y) in accordance with an issuance
calendar announced in advance. The issuance sizes for T-                            cash basis. The Indonesian Clearing and Guarantee
bonds and T-bills are announced a week in advance. New                              Corporation (KPEI) acts as the counter party for settling and
bond issues and tap re-openings are both conducted through                          liquidating an open position upon contract maturity
Dutch auctions. On several occasions, the government has                            Monetary Policy Framework
held switches (also called ‘bond exchanges’) to extend the
bond maturity or reduce debt-service costs. Similarly, the                          Bank Indonesia’s (BI) main policy objective is to establish
ministry of finance has conducted debt-buybacks on several                          and maintain rupiah stability. This objective incorporates a
occasions. 9, 12 month SBIs are auctioned once a month for                          stable inflation rate and exchange rate stability against other
liquidity sterilization purposes. Bank Indonesia phased out                         foreign currencies. Under the Bank Indonesia Act, the
issuance of tenor less than 9mth for SBIs, during 2010-2011.                        government sets inflation targets for a period of 3 years. The
                                                                                    inflation targets for 2010-2012 are 5.0%, 5.0% and 4.5%
Exhibit 93                                                                          respectively with ±1% deviation.
Historical Foreign Ownership of Government
                                                                                    BI uses various instruments to achieve its policy objectives,
Bonds
                                                                                    notably open market operations (OMOs), changes in bank
  300
                                                                                    reserve requirements and the setting of BI’s policy rate.
  250                                                                               Through OMOs, BI drains liquidity by issuing SBI sterilization
                                                                                    paper and through term deposits, while it fine-tunes money
  200
                                                                                    market liquidity through twice-daily liquidity operations (FRK).
  150                                                                               BI steers the O/N rate within a corridor around BI’s policy
                                                                                    rate by offering a deposit facility at its FASBI rate (set at BI
  100
                                                                                    policy rate minus 200bp in Jan 2012) and by providing
   50                                                                               liquidity at its repo rate (BI policy rate plus 100bp). BI
                                                                                    occasionally changes the level and composition of bank
     -
         2007           2008          2009         2010         2011                reserve requirements – primarily with the aim of influencing
                Foreign Holding of Bonds             Foreign Holding of SBI         the volume and cost of bank credit, but occasionally also for
Source: Morgan Stanley Research estimates
                                                                                    macro-prudential reasons.




                                                                                                                                                46
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Taxation
The Center for Government Bond Management at the
Ministry of Finance (MOF) regulates the taxation of bonds.

Interest and capital gains on bonds coursed through the
Indonesia Stock Exchange (IDX) are subject to a single, final
withholding tax of 20%. Interest and capital gains from bond
transactions not reported to the IDX are subject to general
income taxes (30% maximum) after a preliminary withholding
tax of 15% is deducted. Certain entities specifically
mentioned in country-specific double tax treaties may be
subject to partial or full exemption of interest withholding tax.
Interest paid by a domestic taxpayer to a resident is subject
to a 15% withholding tax.

Useful Links
Bank of Indonesia
http://www.bi.go.id/web/en
Indonesia Debt Management Office
http://dmo.or.id/en
Indonesia Stock Exchange
http://www.idx.co.id/
BPS-Statistics Indonesia
http://dds.bps.go.id/eng/
Morgan Stanley AXJ Rates Strategy Bloomberg Page: IRAX




                                                                                              47
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Korea
Market Overview                                                   the market in recent years, as the BoK uses MSBs to
                                                                  “sterilize” FX market intervention.
Korea’s bond market is the second largest in Asia- ex- Japan
region, trailing only the Chinese markets. Korean FX and local    Daily trading volumes in KTBs id USD 20-30bn, in MSBs is
rates markets are also among the most liquid in the region.       USD7-10bn and for interest rate swaps is USD2-3bn.
The average daily turnover in government bonds is around
                                                                  Settlement: Trading in the secondary market can be done
KRW4.6tn or USD4bn. Foreign investors have unrestricted
                                                                  through the Korea Exchange (KRX) or over-the-counter
access to nearly all fixed income instruments in Korea, which
                                                                  (OTC). Bond settlement on the Korea Stock Exchange (KSE)
also makes major players in these markets.
                                                                  is T+1 for local investors and T+2~3 for foreign investors.
Korean cash instruments include Korean Treasury Bonds             Over-the-counter settlement in Korea can be negotiated
(KTBs), Monetary Stabilisation Bonds (MSBs) issued by Bank        between T+1 and T+30. Payments are made to the KRX
of Korea for sterilization purposes, corporate bonds as well as   account opened at the BOK-Wire, and the securities are
a range of other public sector bonds. Korea’s derivatives         settled under the book-entry system of the Korea Securities
markets are also well developed and include 3yr, 5yr and 10yr     Depository.
KTB futures, as well as interest rate and cross currency
                                                                  Exhibit 94
swaps and swaptions. Local banks and securities firms are
                                                                  Issuance Volume of LCY Bond Market (in USD bn)
also actively involved in the derivatives market and often act
as intermediary between foreign banks and local clients.           200
                                                                                Corp
There is good liquidity in 3y KTB futures with an average daily    180
                                                                                Govt
volume of around 100,000 contracts per day, with each              160
contract worth KRW 100mn.                                          140
                                                                   120
Korea has a long term Foreign Currency Sovereign Debt
                                                                   100
rating of A by S&P, A+ by Fitch and A1 by Moody’s.
                                                                    80
Fixed Income Markets                                                60
                                                                    40
The Korean bond market is one of the largest and the most
                                                                    20
liquid bond market in the Asia-ex-Japan region. Currently, all
fixed income instruments are accessible to foreign investors,        0
                                                                         2002


                                                                                  2003


                                                                                                2004


                                                                                                       2005


                                                                                                                 2006


                                                                                                                            2007


                                                                                                                                     2008


                                                                                                                                            2009


                                                                                                                                                      2010


                                                                                                                                                              2011
though subject to withholding tax. As of 3Q11, the size of the
local currency government bond market (including MSBs) was        Source: AsianBondsOnline, Asian Development Bank
USD524bn (48% of GDP) and the size of local currency              Exhibit 95
corporate bond market was USD719bn. In comparison to the          Investors Profile of KTBs and MSBs
USD1229bn outstanding in the local currency bond market,
                                                                     100%
the size of foreign currency bonds issued by Korean issuers                                                                                        Others
                                                                      90%
was approx. USD145bn at end-3Q10.
                                                                      80%
The government bond market consists mainly of treasury                70%                                                                          Contractual
bonds or KTBs, that are issued by the MOSF to fund the                                                                                             Savings
                                                                      60%                                                                          Institutions
budget deficit of the central government.
                                                                      50%                                                                          Banks

The MOSF also issues Inflation-linked treasury bonds (KTBi),          40%

the coupon payments of which are fixed and semi-annual with           30%
                                                                                                                                                   Govt
the principal calculated on the basis of CPI on the reference         20%
rate. There are only three KTBis currently outstanding with a         10%
                                                                                                                                                   Central Bank
total amount of KRW 4.4tn.                                               0%
                                                                        02

                                                                                  03

                                                                                           04

                                                                                                  05

                                                                                                  06

                                                                                                                07

                                                                                                                       08

                                                                                                                              09

                                                                                                                                     10




Monetary Stabilization Bonds (MSBs) are money market
                                                                      20

                                                                                20

                                                                                         20

                                                                                                20

                                                                                                20

                                                                                                              20

                                                                                                                     20

                                                                                                                            20

                                                                                                                                   20




instruments issued by the Bank of Korea (BoK). Sizeable
                                                                  Source: AsianBondsOnline, Asian Development Bank
balance of payments surplus has contributed to the growth of




                                                                                                                                                                  48
                                                                                     MORGAN STANLEY RESEARCH

                                                                                     March, 2012
                                                                                     EM Local Markets Guide




Market Ownership                                                                     normally trades in line with onshore FX forwards (during times
KRW-denominated bonds are mainly owned by domestic                                   of normal liquidity). There is a liquid OTC FX options market
banks, local pension funds and insurance companies as well                           as well.
as Korean Investment Trust Companies and local securities
                                                                                     Monetary Policy Framework
brokerages. Although offshore investors, including both real
money investors and foreign central banks, have increased                            The Bank of Korea’s (BoK) main monetary policy objective is
their ownership of KRW-denominated bonds, the majority of                            price stability. The current Bank of Korea Act stipulates that
these bonds are Monetary Stabilisation Bonds issued by the                           the BoK sets an inflation target in consultation with the
Bank of Korea. Offshore investors also actively trade the 3yr                        government. The inflation target for 2010-2012 is 3%
KTB future.                                                                          plus/minus 1 percentage point around this target, in terms of
                                                                                     the 12-month rate of change in the consumer price index.
Exhibit 96
Historical Foreign Ownership of Government Bonds                                     The BoK’s Monetary Policy Committee meets during the
   %                                                                  KRW tn
                                                                                     second week of every month to set the Bank of Korea Base
 12                                                                            600
                    % of Total                                                       Rate. The 7-day Repo Rate is set as the Central Bank Policy
 10                 Foreign Holdings                                           500
                                                                                     Rate (Bloomberg: KOCRD Index).
                    Total
                                                                                     Monetary Policy Instruments
  8                                                                            400
                                                                                     Open Market Operations: The BoK uses Open Market
  6                                                                            300
                                                                                     Operations (OMOs) to manage the overnight call rate so that
  4                                                                            200   it does not deviate too widely from the Bank of Korea Base
                                                                                     Rate. These operations are conducted in three ways: through
  2                                                                            100   securities transactions, through the issuance and repurchase
                                                                                     of Monetary Stabilization Bonds (MSBs), and through
  0                                                                            0
   2002      2003      2004      2005   2006   2007   2008   2009   2010
                                                                                     commercial banks’ deposits made in the Monetary
                                                                                     Stabilization Account (MSA).
Source: CEIC, Morgan Stanley Research
Auctions                                                                             Securities transactions involve purchases and sales of
Treasury Bonds (KTBs) are issued (or tapped) in maturities of                        government securities, securities guaranteed by the
3, 5, 10 and 20 years by the Ministry of Strategy and Finance                        government, MSBs and other types of securities specified by
(MOSF), every Monday of the month, according to pre-                                 the Monetary Policy Committee. Securities transactions take
announced auction calendar released on the MOSF website.                             the form of outright sales and purchases or of repurchase
KTBs are issued via Dutch style yield auction. Only the                              agreements (RPs). RPs are employed as a major instrument
primary dealers (PDs) and preliminary primary dealers (PPDs)                         for the routine adjustment of market liquidity (mostly 7-day
are allowed to participate in the auction, while investors can                       maturities).
purchase these bonds via the 20 designated PDs and 2                                 Lending and Deposit Facilities: The Bank of Korea's regular
PPDs. Investors need to apply for an investment                                      lending facilities currently include (i) Liquidity Adjustment
registration certificate (IRC), issued by the Financial                              Loans,' provided to financial institutions experiencing fund
Supervisory Board and designate a foreign exchange                                   shortage (ii) Aggregate Credit Ceiling Loans', support given
bank, custodian bank, and hold a standing proxy                                      up to a certain ceiling to small and medium-sized enterprises
agreement with a securities firm.                                                    and regions, and (iii) Intraday Overdrafts,' provided to financial
                                                                                     institutions facing temporary shortages of funds for settlement
FX
                                                                                     until the daily funds-transfer business closing. These
The KRW is a non-convertible currency, and is traded by                              loans are implemented in the form of bill rediscounts or
offshore investors using a Non-Deliverable Forward. The                              securities collateral loans. The collateral includes credit
Korean market has the deepest liquidity in the AXJ region with                       securities, government and public bonds and Monetary
close to USD8-10bn of daily trading volume in onshore spot                           Stabilization Bonds.
and USD5-6bn in deliverable forwards. Fairly liquid offshore
                                                                                     Reserve Requirements: The financial institutions subject to
Non-Deliverable Forward (around USD 1bn) and FX-swap
                                                                                     the current reserve requirement regime include commercial
markets support trading of the Korean won offshore, which

                                                                                                                                                   49
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




banks and special banks. These institutions are required to       Taxation
maintain reserves corresponding to their reserve requirement
                                                                  Interest Income
ratios differentiated within a range between 0~7% depending
upon their deposit liability types.                               Interest on bonds issued by Korean companies or
                                            13                    government bodies is generally subject to a 14% WHT rate.
Regulatory Developments
                                                                  An additional resident tax surcharge of 10% on CIT liability is
Tax Revision Bill for 2011: In September the Ministry of          assessed, making the effective tax rate at 15.4%. For
Strategy and Finance introduced its tax revision bill for 2011.   residents of countries with a tax treaty with Korea, reduced
One of the key aspects of the bill is fair taxation: yields       WHT rates could be applied.
derived from issuing FCY bonds domestically will be taxed
                                                                  Capital Gains
just like LCY bonds, and capital gains taxes coming from
financial products, including derivatives, will be legislated.    10% of gross proceeds or 25% of capital gains, whichever is
                                                                  lower. (Combined taxes result in an effective rate of 11% and
FSCMA Revision to Allow Hedge Funds: The Financial
                                                                  27.5%, respectively.) Foreign investors in certain countries
Services Commission (FSC) reported that the government
                                                                  may be exempted from CIT depending on the relevant double
approved on 27 September the proposed revision of the
                                                                  taxation treaty.
Enforcement Decree of the Financial Investment Services and
Capital Markets Act (FSCMA), which will allow for the entry of    Exhibit 97
locally based hedge funds into the Republic of Korea’s capital    Korea: Treaty Interest Withholding Rates
market and allow them to invest in a wider range of asset                                                    Interest Withholding Tax
classes. It also eases the restrictions on derivatives trading     Australia                                           15%
and leverage, and increases the diversity of hedge fund            China                                               10%
                                                                   France                                              10%
investors.
                                                                   Germany                                             10%
Interest Rate Derivatives                                          Italy                                               10%
                                                                   Japan                                               10%
KTB Futures: 3y and 10y KTB futures are actively traded            New Zealand                                         10%
with foreign investors being major players in this market. The     Singapore                                           10%
contract size is KRW 100 million and contract months are           Switzerland                                         10%
                                                                   Thailand                                            10%
March, June, September and December. The final settlement
                                                                   United Kingdom                                      10%
is done on a cash-settlement basis, and the price is               United States                                       12%
determined using the underlying bond basket, typically            Source: Morgan Stanley Global Network Management
containing 3 bonds.                                               Useful Websites
Interest Rate Swaps: For Korea, both the onshore and              Bank of Korea
offshore IRS curves are liquid, and used extensively by           http://eng.bok.or.kr/eng/engMain.action
banks, debt issuers and foreign investors to hedge duration       Korea Ministry of Finance
risk. The 91-day CD rate (Bloomberg: KWCDC Index) is used         http://english.mosf.go.kr/
as the fixing for the floating leg of the swap. Structured note   Financial Supervisory Service
issuers are major players in IRS and FRA markets.                 http://english.fss.or.kr/fsseng/index.jsp
                                                                  KRX Korea Exchange
Cross Currency Swaps: A USD/KRW cross-currency swap
                                                                  eng.krx.co.kr/
(CCS) can be structured out to 10 years, but one to five years
                                                                  Korea Securities Dealers Association
are more actively traded tenors. Exporters’ USD forward
                                                                  http://www.ksda.or.kr/
selling and the central bank’s FX intervention are the main
                                                                  Statistics Korea
drivers in the short end of the cross-currency curve, while
                                                                  http://kostat.go.kr/eng/
external debt issuers and investors generally trade in the long
                                                                  Korea Money Brokerage Corp.
end.
                                                                  http://www.kmbco.com

                                                                  Morgan Stanley AXJ Rates Strategy Bloomberg Page: IRAX

13
     Source: AsianBondsOnline, Asian Development Bank


                                                                                                                                        50
                                                                              MORGAN STANLEY RESEARCH

                                                                              March, 2012
                                                                              EM Local Markets Guide




Malaysia
Market Overview                                                               of managing liquidity in the conventional financial market.
                                                                              These new BNMN issuances can have maturity between one
The Malaysian bond market is one of the more developed
                                                                              to three years and can either discounted or a coupon-bearing
bond markets in the region, with a liberal foreign exchange
                                                                              depending on investors' demand.
regime (with gradually easing and liberalizing foreign
administration rules, and no restrictions on the repatriation of              MTB are short-term securities issued by the Government of
capital, income and profits earned from Malaysia). Investment                 Malaysia to raise short-term funds for Government's working
in bonds is also exempt from withholding and capital gains                    capital. Bills are sold at discount through competitive auction,
tax. Malaysia also has an active derivatives market.                          facilitated by Bank Negara Malaysia, with original maturities of
                                                                              3-month, 6-month, and 1-year.
MYR bonds can only be traded onshore, but can be settled
via Euroclear/ Clearstream. Malaysian government bonds                        Exhibit 99
carry a weighting of 5-10% in Citi and JPM EM bond indices.                   Market Ownership of Malaysian Government Bonds
As of end-Jan 2012, total outstanding size of the government                     100%
bond market stood at MYR280bn or USD 90bn with average                            90%                                                            Foreign Holders
daily turnover value of MYR74bn.                                                  80%
                                                                                                                                                 Financial Institutions
Malaysia’s long term foreign currency sovereign debt rating is                    70%

A- by S&P, A- by Fitch and A3 by Moody’s.                                         60%                                                            Insurance Companies
                                                                                  50%
Exhibit 98                                                                                                                                       Public Sector
                                                                                  40%
Size of Malaysia LCY Bond Market (MYR bn)                                         30%
   300
                                                                                                                                                 Bank Negara Malaysia
                                                                                  20%
                Corp
                Govt                                                              10%                                                            Social Security
   250
                                                                                    0%                                                           Institutions
                                                                                         1996

                                                                                                1998

                                                                                                       2000

                                                                                                              2002

                                                                                                                     2004

                                                                                                                            2006

                                                                                                                                   2008

                                                                                                                                          2010
   200


   150                                                                        Source: Asianbondsonline, Asian Development Bank


   100                                                                        Market Ownership

    50                                                                        Financial institutions are the biggest investors in government
                                                                              bonds, holding 43% of total outstanding MGS and GII as of
     0                                                                        Sept 2011. Foreign investors and Social Security Institutions
     Mar-04   Mar-05   Mar-06    Mar-07   Mar-08   Mar-09   Mar-10   Mar-11
                                                                              hold around 25% each, of the total MGS, GII outstanding.
Source: Asianbondsonline, Asian Development Bank                              Other investors include insurance companies (6%), Bank
                                                                              Negara Malaysia (1%) and the public sector (1%).
Fixed Income Markets
Malaysia's local currency bond market is active for both
conventional Malaysian Government Securities (MGS) and
Islamic bonds (GII). Domestic and foreign investors can buy
and sell conventional and Islamic debt instruments through
the exchange and over-the-counter (OTC) markets. Nearly all
securities are scripless, with securities transferred
electronically via BNM’s Real-Time Electronic Transfer of
Funds and Securities (RENTAS). Transfer instructions are
conducted on a trade-by-trade basis.

BNMN are securities issued by Bank Negara Malaysia
replacing the existing Bank Negara Bills (BNB) for purposes

                                                                                                                                                                      51
                                                                             MORGAN STANLEY RESEARCH

                                                                             March, 2012
                                                                             EM Local Markets Guide




Exhibit 100                                                                  Derivatives
Historical Foreign Ownership of Government Bonds
                                                                             Malaysia has a relatively active derivative market with many
 MYR bn                                                                      instruments being available to investors.
 400           Total Government Bonds Outstanding                       30

 350           Foreign Holdings                                              Bond Futures: The 3-year, 5-year and 10-year MGS
                                                                        25
               Foreign Holdings as % of Total                                contracts carry a hypothetical coupon rate of 6%, and the
 300
                                                                        20   yield is derived from the weighted yields of all eligible MGS in
 250
                                                                             the basket. The weights are announced by the Exchange.
 200                                                                    15   Upon maturity, all bond derivatives are settled in cash terms
 150
                                                                        10
                                                                             using a final settlement value.
 100
                                                                             Interest Rate Futures (3mth KLIBOR futures): underlying
                                                                        5
  50                                                                         instrument is the ringgit interbank time deposit in the
   0                                                                    0    wholesale money market with 3-month maturity on a 360-day
   1996       1998     2000       2002      2004   2006   2008   2010        year. Quarterly cycle contract months are March, June,
Source: Morgan Stanley Research estimates                                    September and December up to five years ahead and two
Auctions                                                                     serial months

The auction process of the MGS is open only to PDs, while                    Interest Rate Swaps: Malaysia 3 month interbank offered
the GII tenders are open for PDs and Islamic PDs, which are                  rate or 3 month KLIBOR (Bloomberg ticker: KLIB3M Index)
then obliged to tender a minimum amount of the issue size in                 makes the floating leg of the swap settled on a quarterly,
order to ensure 100% subscription. Bids submitted during the                 Actual/365 day count basis for the long end. Tenors up to 10
auction could be based on price in case of reopening of                      years are actively traded. As onshore swaps can be used only
existing securities and trading in the secondary market, or                  for hedging, offshore investors access the market though the
based on yield if it is a new issuance. For a new issue, the                 non-deliverable NDIRS swap market.
weighted average of yields accepted is taken as the coupon                   Cross Currency Swaps: These are less liquid than IRS, but
for that MGS or GII issue.                                                   the curve exists up to 10 years. Onshore investment is
Typical issuance size is approximately MYR 3-4bn with 1-2                    restricted to bond hedging.
auctions per month. The exact date and issuance size is                      Monetary Policy Framework
announced a week in advance (at least 5 business days
before the issue date), and the bonds are auctioned through a                Bank Negara Malaysia’s (BNM) primary policy objective is to
variable-rate multiple-price auction format (also known as                   maintain price stability while developing a progressive and
English Auction). The "when-issued" (WI) trading commences,                  resilient financial system. BNM steers its monetary policy
after the tender announcement date, in Fully Automated                       stance through changes in the Overnight Policy Rate (OPR),
System for Issuing/Tendering (FAST).                                         which it decides on every two months. The overnight money
                                                                             market rate is managed within a corridor of ±25bp of the OPR
FX                                                                           - through Open Market Operations. BNM conducts
Spot USD-MYR is OTC-traded with average daily volume of                      sterilization operations by means of BNM money market
about USD1.4bn. Foreigners can buy and sell USD-MYR spot                     tenders, repo tenders, through issuance of BNM notes, FX
intraday without restrictions. However, foreigners are not                   forward transactions, as well as through changes in bank
allowed to carry negative MYR balances overnight. Instead,                   liquidity and statutory reserve ratios.
investors can tap the onshore spot market as an alternative                  Taxation
source of liquidity for hedging spot risk. Foreign investors
cannot access onshore forward markets without a qualifying                   Many of the government and corporate fixed income
underlying security or foreign direct investments in Malaysia.               securities are tax exempt, such as:
The MYR NDF market is well developed with about USD1bn                               all fixed income securities including debentures
daily volumes, predominantly around the fixing.                                       (other than convertible loan stocks) that are so
                                                                                      approved by the Securities Commission;



                                                                                                                                          52
                                                            MORGAN STANLEY RESEARCH

                                                            March, 2012
                                                            EM Local Markets Guide




       securities issued by the Malaysian government in
        Malaysian Ringgit (MYR) including MGS, MTNs, GII,
        MTBs and BNM notes ;

       Islamic securities approved by the Securities
        Commission.



Relevant Websites
Bank Negara Malaysia (BNM)
http://www.bnm.gov.my/
Fully Automated System for Issuing/Tendering (FAST), BNM
https://fast.bnm.gov.my/fastweb/public/MainPage.do
Department of Statistics Malaysia
http://www.statistics.gov.my
Ministry of Finance
http://www.treasury.gov.my/
Morgan Stanley AXJ Rates Strategy Bloomberg Page: IRAX




                                                                                      53
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Philippines
Market Overview                                                    The BTr also issues RTBs that cater to small individual
                                                                   investors. These issues have smaller denominations and
The Philippine domestic bond market consists of short- and
                                                                   more frequent, smaller fixed coupon payments. RTBs are
long-term bonds and is dominated by government securities.
                                                                   offered in 3 and 5 year tenors. In April 2010, the Philippines
Although the size of the Philippine corporate bond market is
                                                                   began issuing multi-currency RTBs (to enable Filipinos to
small, it has grown rapidly over the years.
                                                                   invest in foreign-currency denominated government
Government Bonds form the majority of total debt outstanding       securities). Finally, there is a small corporate debt market
(primarily issued for deficit funding). Government bonds can       whose issues pay floating rate coupons and have tenors
be classified as Treasury Bills, Fixed-Rate Treasury Notes         between 2 and 7 years.
(FXTN), Retail Treasury Bonds RTB) and Multi-Currency
                                                                   Settlement
Retail Treasury Bonds. Commercial paper comprises mostly
floating rate debt instruments issued by leading companies.        Transactions by Government Securities Eligibile Dealers
                                                                   (GSEDs) are cleared through the Registry of Scripless
The government has one of the highest percentages of
                                                                   Securities (RoSS), which is managed by the BTr. Settlement
foreign currency debt outstanding (in the region) as the
                                                                   of government securities occurs at T+1, but an alternate
government tends to tap foreign currency debt for longer
                                                                   settlement date is possible if agreed upon by both parties.
maturities. But, this funding mix has been shifting from foreign
currency to local currency debt over the past few years.           Exhibit 102
                                                                   Government Debt: Shift from Foreign Currency to
The Philippine sovereign has a foreign currency sovereign
                                                                   Local Currency Debt
debt rating of BB by S&P, BB by Fitch and Ba3 by Moody’s.
                                                                   70            Government FCY Debt
Exhibit 101                                                                      (USD Bns)
                                                                   60
Bonds Market: LCY/FCY, CORP/GOVT                                                 Government LCY Debt
                                                                                 (USD Bns)
                                                                   50
                   8.06%                     LCY Govt Bond
                                                                   40

                                                                   30
                                             LCY Corp Bond
                                                                   20
         34.11%
                                                                   10
                                  67.42%     FCY Govt Bond
                                                                     0
                                                                     Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12
                                                                   Source: Morgan Stanley Research estimates
           9.94%                             FCY Corporate
                                             Bonds and
                                             Financial             Market Ownership
                                             Institutions
Source: Morgan Stanley Research estimates                          The largest owners of Philippine government bonds are
                                                                   government-owned or controlled corporations (GOCCs), BSP,
Fixed Income Markets                                               Pag-IBIG housing fund and the two state banks (Land Bank of
The Bureau of the Treasury (BTr) issues zero-coupon                the Philippines and Development Bank of the Philippines).
treasury bills maturing in less than one year every two weeks.     The largest pension funds are the state-owned Government
The three standard maturities are 91-day, 182-day, and 364-        Service Insurance System (GSIS) and the Social Security
day. Treasury bills are the most actively traded instrument in     System (SSS) for public and private employees, respectively.
the Philippine bond market. FXTNs are issued by the BTr in         Insurers and asset management companies are also
standard maturities of 2, 5, 7, 10, 15, and 25 years with          important players. Generally, there are no restrictions on non-
semiannual coupons.                                                residents’ investment in local bonds, money market
                                                                   instruments or other portfolio investments.


                                                                                                                                 54
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Auctions                                                           To increase or reduce liquidity in the financial system, the
                                                                   BSP accepts fixed-term deposits, offers standing facilities,
The Bureau of the Treasury announces monthly bond
                                                                   and requires banking institutions to hold reserves on deposits
offerings on a quarterly basis, including the auction date and
                                                                   and deposit substitutes. BSP also accepts deposits from
bond offer volume. The Bureau of the Treasury also issues
                                                                   banks through Special Deposit Accounts (SDA), which were
public notices of treasury bonds being offered either through
                                                                   introduced in 1998 with access expanded to trust entities in
auction or the tap system. T-Bonds are issued (new or re-
                                                                   2007. SDA’s have grown in popularity in the absence of other
issue) twice a month (maturities of 2, 3, 4, 5, 7, 10, 20 and
                                                                   major high-yielding instruments in the country.
25y with an amount of PHP 7-9bn) under the Dutch system of
competitive issuance. The government also conducts debt            In addition, the BSP provides loans and a standing credit
exchanges as a way to lengthen and redistribute maturities.        facility to help banks meet temporary liquidity needs. Finally,
T-Bills are auctioned twice a month (for three tenors, and a       the BSP can also exercise monetary policy through the
total amount of PHP 6-7bn) through a competitive multiple          reserve requirement channel.
price auction. The Bureau of Treasury can also choose to
                                                                   The BSP’s key policy rates are its repo and reverse repos as
reject the entire auction if bid yields are too high for both T-
                                                                   well as quantum and rates paid on its Special Deposit
Bills as well as T-Bonds.
                                                                   Accounts.
FX
                                                                   Derivatives
The Republic maintains a floating exchange rate system
                                                                   No offshore IRS market exists for the Philippines yet, but CCS
under which market forces determine the exchange rate for
                                                                   does trade offshore, albeit with low liquidity. Benchmark Index
the peso. Bangko Sentral may, however, intervene in the
                                                                   for floating leg is the 3m PHIREF, the 3m interbank reference
market to maintain orderly market conditions and limit sharp
                                                                   rate.
fluctuations in the exchange rate. Payments related to foreign
loans registered with Bangko Sentral and foreign investments       Regulatory Developments 14
approved by or registered with Bangko Sentral may be
                                                                   Bangko Sentral ng Pilipinas imposed stricter rules on hedging
serviced with foreign exchange purchased from authorized
                                                                   instruments, particularly non-deliverable forward (NDFs)
agent banks. Bangko Sentral must approve and register all
                                                                   contracts, to reduce speculation in the foreign exchange
outgoing investments by residents exceeding $6 million per
                                                                   market. The market risk capital charge to be used in the
investor per year if the funds will be sourced from the banking
                                                                   capital adequacy ratio (CAR) computation for banks for the
system. Daily turnover is approximately PHP 1bn for onshore
                                                                   net open position of NDF contracts was be raised from 10% to
while approximately PHP 0.5bn Offshore.
                                                                   15% effective 1 January 2012.
Monetary Policy Framework
                                                                   Taxation
The primary objective of BSP's monetary policy is to promote
                                                                   Interest income from Philippine peso-denominated
low and stable inflation conducive to balanced and
                                                                   government or corporate debt securities is subject to a 20%
sustainable economic growth. BSP adopted its inflation
                                                                   final withholding tax. The withholding tax rate applies to both
targeting framework in January 2002 and is aimed at
                                                                   resident and non-resident investors engaged in trade or
achieving the government’s inflation target as defined in terms
                                                                   business in the country. Government securities are exempt
of the average year-on-year change in the consumer price
                                                                   from capital gains tax, while a 5–10% tax rate is levied on
index (CPI) over the calendar year. The BSP has retained an
                                                                   other debt securities.
inflation target of 3-5% between 2012 and 2014.

The BSP has a number of monetary policy instruments at its
disposal to promote price stability.

The BSP conducts open market operations (OMO) to control
the money supply which in turn helps it achieve its inflation
objectives. The BSP can do this through repurchase/reverse
repurchase agreements, the outright purchases and sales of
government securities, and FX swaps.
                                                                   14
                                                                        Source: AsianBondsOnline, Asian Development Bank


                                                                                                                                55
                                                              MORGAN STANLEY RESEARCH

                                                              March, 2012
                                                              EM Local Markets Guide




Relevant Websites
Bangko Sentral Ng Pilipinas (Central Bank)
http://www.bsp.gov.ph
The Government of the Republic of the Philippines, Investor
Relationship Office
http://www.iro.ph/index.php
Bureau of the Treasury
http://www.treasury.gov.ph/
Department of Budget and Management
http://www.dbm.gov.ph/
National statistics Office
http://www.census.gov.ph/
Morgan Stanley AXJ Rates Strategy Bloomberg Page: IRAX




                                                                                        56
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Singapore
Market Overview                                                   Settlement
Singapore has one of the biggest corporate bond markets in        Bond trading can be done through the Singapore Exchange or
the region, but a relatively small local currency government      over an automatic order matching system called the Bonds
bond market (Singapore Government Securities, SGS). The           Quotation System developed by the exchange. However,
average daily turnover in government bonds at end of 2011         SGS and corporate bonds in the secondary market are mostly
was about USD100bn. As Singapore is a completely open             traded OTC. SGS transactions are cleared on a delivery-
economy with no capital controls, all entities and individuals,   versus-payment (DVP) basis through the MAS Electronic
local and foreign, can participate in the market.                 Payment System (MEPS) in T+1 , while non-government
                                                                  bonds are settled through an electronic clearing system
Singapore’s long term foreign currency sovereign debt rating
                                                                  established by the Central Depository (Pte) Limited, a
is AAA by S&P, AAA by Fitch and Aaa by Moody’s.
                                                                  subsidiary of the SGX.
Fixed Income Markets
                                                                  Exhibit 104
Singapore’s bond market is dominated by government debt,          Size of Local Currency Bond Market
which consists of SGS treasury bills and SGS bonds.
                                                                      250             Corp
However, because the Singapore government has typically
run budget surpluses, it does not need financing from issuing                         Govt
government bonds. Instead SGSs are primarily issued for               200
benchmark maintenance purposes. In addition to SGS, there
is also an active corporate bond market, as well as a growing         150
Islamic debt market, which was introduced in January 2009.

SGS treasury bills are short-term instruments maturing in one         100
year or less, issued at a discount. SGS bonds have tenors
ranging from 2 to 20 years. Corporate bonds are generally               50
issued by corporations, property companies, financial
institutions, and supra-nationals. The amount of outstanding              0
SGS (bills and bonds) at end 2011 was about USD118bn,


                                                                          11
                                                                          08



                                                                          09




                                                                          10
                                                                          07




                                                                           0




                                                                           1
                                                                           6



                                                                           7




                                                                           8



                                                                           9



compared to USD71bn for corporate bonds.                                 -1



                                                                         -1
                                                                         -0




                                                                         -0



                                                                         -0



                                                                         -0
                                                                       n-



                                                                       n-




                                                                       n-



                                                                       n-



                                                                       n-
                                                                      ec



                                                                      ec
                                                                     ec




                                                                     ec



                                                                     ec



                                                                     ec




                                                                    Ju
                                                                    Ju



                                                                    Ju




                                                                    Ju



                                                                    Ju

                                                                    D



                                                                    D
                                                                    D



                                                                    D




                                                                    D



                                                                    D




SGD-denominated bonds include Government Bonds (SGS
                                                                  Source: Asia Bonds Online
Bonds and SGS Bills) and Corporate Bonds (issued by
Corporations, Property Companies, Financial Institutions and      Market Ownership
supra-nationals).
                                                                  The largest investors in the Singapore bond market are fund
Exhibit 103                                                       management companies, such as Government of Singapore
Maturity Profile of Singapore Government Bonds                    Investment Corporation (GIC) and Central Provident Fund
                                                                  (CPF). Private asset management companies, banks, and
                                                                  insurers are also important players in the market. Generally,
                                      1-3 years
                                                                  there are no restrictions on non-residents’ investment in local
          22.66%             32.02%                               bonds, money market instruments, or other portfolio
                                      3-5 years
                                                                  investments.

                                      5-10 years                  Monetary Policy Framework
        27.04%
                            18.28%
                                                                  The main policy objective of the Monetary Authority of
                                      >10 years
                                                                  Singapore (MAS) is to ‘promote sustained, non-inflationary
                                                                  economic growth and a sound and progressive financial
                                                                  centre’. In conducting its monetary policy, MAS manages the
Source: Asia Bonds Online



                                                                                                                              57
                                                                     MORGAN STANLEY RESEARCH

                                                                     March, 2012
                                                                     EM Local Markets Guide




exchange rate through a trade-weighted index (NEER) of               Regulatory Developments 15
Singapore’s main trading partners within an undisclosed
                                                                     MAS released new capital rules for banks in Singapore that
target band. MAS publishes semi-annual Monetary Policy
                                                                     exceed the levels established under the Basel III agreement.
Statements (MPS), generally in April and October, in which it
                                                                     Effective 1 January 2015, MAS will require Singapore-
announces changes in its policy stance though changes in the
                                                                     incorporated banks to have a minimum common equity Tier 1
SGD’s target band in terms of slope, width and/or a re-
                                                                     capital adequacy ratio (CAR) of 6.5%, a Tier 1 CAR of 8.0%,
centering of the midpoint of the band.
                                                                     and a total CAR of 10.0%. In addition, MAS will require
The MAS may intervene in the FX spot or forwards market to           Singapore-incorporated banks to meet the minimum capital
manage the floating NEER within the policy band and prevent          adequacy requirements of Basel III by 1 January 2013, which
excessive fluctuations in the exchange rate.                         is 2 years ahead of the Basel Committee’s 2015 timeline.
                                                                     MAS plans to adopt Basel III’s capital standards to improve
The MAS also conducts money market operations in the form
                                                                     the consistency, transparency, and quality of the capital base,
of repos, FX swaps, borrowing and lending operations.
                                                                     and to strengthen the risk coverage of capital rules for banks.
However, in the context of Singapore’s open economy,
managing the exchange rate as the main tool for monetary             Auctions
policy implies that interest rates and money supply are
                                                                     The Monetary Authority of Singapore issues SGS Bills and
endogenous. Thus, money market operations are primarily for
                                                                     Bonds on behalf of the Singapore Government for maturities
managing liquidity and support orderly trading of the SGD
                                                                     of 2y, 7y, 10y and 15y respectively, with an approximate
within its trading band. The MAS also offers a borrowing and
                                                                     issuance size of SGD 1-2.5bn. The auction format is uniform-
lending facility to primary dealers in order to control interest
                                                                     pricing for both bills and bonds. The auctions take place as
rate volatility. In April 2011, the MAS began issuing 4-week
                                                                     per the issuance calendar announced at the beginning of the
and 8-week tenor bills for market operations purposes.
                                                                     fiscal year, with a bond auction once every month and bill
                                                                     issuance once every week. At most 40% of each auction’s
Exhibit 105                                                          total issuance amount is allotted to non-competitive bids.
MAS S$NEER Index and Policy Band                                     FX
 (S$NEER, Jan 1999 = 100)
 120
                                                                     The Singapore FX market, though deliverable, is split
                                                                     between onshore/offshore that are non-fungible.
                                                                     Onshore/offshore differentiation occurs in the FX swaps
 115
                                                                     market, where offshore players can only B/S USDSGD swaps
                                                                     to onshore banks; as a result, the offshore SGD curve trades
 110
                                                                     at a varying (typically small) premium. Offshore swap liquidity
                                                                     exists through spot FX and basis swaps.
 105
                                                                     The spot market is well developed and sees about $6-7bn
 100                                                                 daily turnover. Offshore investors are free to trade spot with
                                                                     onshore entities. It has been used as a low-beta but liquid
                                                                     USD-Asia FX proxy.
  95
   Jan-99     Jan-01    Jan-03   Jan-05   Jan-07   Jan-09   Jan-11
                                                                     Taxation Policy
                                                                     Interest Income
Source: Bloomberg
Derivatives                                                          Singapore’s qualifying debt securities (QDS) scheme grants
                                                                     concessionary tax treatment to investors of securities covered
Interest Rate Swaps
                                                                     by the scheme. Interest income earned by nonresidents from
Singapore’s IRS market is the second largest in Asia, with an        QDS is exempt from withholding tax. A 10% concessionary
average daily turnover of SGD 78Bn. Typical transaction sizes        tax is granted on interest income earned by financial
are about SGD 20mn, with the fixing based on the 6-month
swap offer rate (SOR). In contrast to a typical LIBOR-type
fixing, the SOR is an FX-implied rate.                               15
                                                                          Source: AsianBondsOnline, Asian Development Bank


                                                                                                                                  58
                                                                 MORGAN STANLEY RESEARCH

                                                                 March, 2012
                                                                 EM Local Markets Guide




institutions. No stamp duties apply for securities. Qualifying
debt securities include:

    (a) Singapore government securities issued during the
        period from 28 February 1998 to 31 December 2013.

    (b) Prescribed bonds, notes, commercial papers and
        certificates of deposits which are arranged in
        accordance with stipulated guidelines during the
        period 28 February 1998 to 31 December 2013.

    (c) Prescribed Islamic securities which are arranged in
        accordance with stipulated guidelines during the
        period 1 January 2005 to 31 December 2013.

Capital Gains

There is no tax payable on capital gains from securities
transactions in Singapore.

Relevant Websites
Monetary authority of Singapore
http://www.mas.gov.sg/
Singapore Government Securities
http://www.sgs.gov.sg/
Morgan Stanley AXJ Rates Strategy Bloomberg Page: IRAX




                                                                                           59
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




Thailand
Market Overview                                                                       Government Debt Instruments
The Thai bond market has grown rapidly since the 1997                                 There are four types of government debt securities in
economic crisis. The government issued bonds in June 1998                             Thailand: (i) Treasury bills; (ii) Government bonds; (iii) Bank of
for financing the budget deficit that resulted from the crisis.                       Thailand (BOT) bonds; and (iv) state-owned enterprise (SOE)
The substantial amount of new government bonds coupled                                bonds.
with the subsequent downtrend in interest rates contributed to
                                                                                      Treasury Bills (T-Bills) are short-term debt instruments
the development of the market as reflected in improving
                                                                                      issued by the Public Debt Management Office (“PDMO”)
turnover ratios. The outstanding value of the total bond
market increased from THB 547 billion in 1996 to THB 7112                             Government Bonds are medium to long-term debt
billion at the end of 2011. The turnover ratio (ratio of value of                     instruments issued by the PDMO, and are classified into Loan
bonds traded over average amount of bonds outstanding) of                             bonds (LB) and Savings Bonds (SB). Loan Bonds, which
the market increased from 0.4 in 2001 to 0.65 in 2011 16.                             make a majority of the market, are issued for financing the
                                                                                      budget deficit and refinancing legacy debt of central bank’s
The majority of bonds are traded over-the-counter (OTC). As
                                                                                      financial restructuring unit, the Financial Institutions
both depository and registrar, the Bank of Thailand (BOT) is
                                                                                      Development Fund (“FIDF”). Savings bonds are mainly used
responsible for the settlement of government bonds. Most
                                                                                      by retail investors, as instruments for household savings
corporate bonds are cleared and settled at a Stock Exchange
                                                                                      purposes, who generally hold these bonds till maturity.
of Thailand (SET) subsidiary— the Thailand Securities
Depository Co. Ltd (TSD). Finally, the Thai Bond Market                               Bank of Thailand (BoT) Bills and Bonds are issued by the
Association (ThaiBMA) acts as a self-regulating organization                          central bank for the purposes of liquidity management and
that functions as an information center, code and standard                            bond market development.
setter, forum for bond market updates, and a frontline for
                                                                                      State Owned Enterprise (SOE) Bonds can be either
market surveillance.
                                                                                      guaranteed by the Ministry of Finance or non-guaranteed.
Thailand’s long term foreign currency sovereign debt rating is                        Guaranteed bonds account for 70% of total and are eligible
BBB+ by S&P, BBB by Fitch and Baa1 by Moody’s.                                        for liquidity reserve requirement.

Exhibit 106                                                                           Auctions
Size of LCY Bond Market (in USD bn)
                                                                                      Government bonds and T-Bills are issued on a weekly basis
 250                                                                                  through competitive price auctions (American auction)
                 Corporate Bonds
                                                                                      organized by Bank of Thailand. Terms and size of the
                 Govt Bonds
 200                                                                                  auctions are announced prior to the auction date.
                                                                                      Approximate issuance size for T-Bonds varies from THB 3-
 150                                                                                  20bn depending on the maturity of the bond and whether the
                                                                                      bond carries a fixed or flexible coupon.
 100
                                                                                      Market Ownership

     50
                                                                                      The institutional investor base, made up of pension funds,
                                                                                      provident funds, insurance companies and mutual funds are
                                                                                      the major investors in Thai Government bonds (including
      0
      Dec-97     Dec-99    Dec-01     Dec-03    Dec-05     Dec-07   Dec-09   Dec-11
                                                                                      Central Bank Bonds) followed by depository institutions and
                                                                                      households/Non- Profit institutions. Retail investors also hold
Source: Asian Bonds Online
                                                                                      a significant 14.5% of domestic debt. Non-residents’
                                                                                      investment has picked up after the 2008 financial crisis, and
                                                                                      currently stands at 11.5% of total government debt
                                                                                      outstanding.

16
     Source: Asian Bonds Online, Asian Development Bank.


                                                                                                                                                    60
                                                                                         MORGAN STANLEY RESEARCH

                                                                                         March, 2012
                                                                                         EM Local Markets Guide




Exhibit 107                                                                              carry a short THB balance on settlement day. Total spot
Investors Profile of Thailand Government Bonds                                           volume each day is about USD1-1.5bn, the vast majority
 100%
                                                                                         being done between local names. Spot FX amongst offshore
                                                             Non-residents
                                                                                         names is predominantly done through voice brokers (approx.
  90%                                                        General Government
                                                                                         USD100m a day).
  80%                                                        Residents
  70%                                                                                    Monetary Policy Framework
                                                             Corporations
  60%
                                                             Contractual Savings Funds   The Bank of Thailand (BoT) is responsible for monetary policy
  50%                                                                                    and exchange rate management. BoT has conducted
                                                             Mutual Funds
  40%                                                                                    monetary policy under a flexible inflation targeting framework
                                                             Insurance Companies
  30%
                                                                                         since May 2000, in which the BoT pays attention to inflation
                                                             Financial Corporations      but also to economic growth and stability including financial
  20%
                                                             Commercial Banks            market conditions as well as financial status of households,
  10%
                                                             Central Bank                businesses, and financial institutions By December of each
    0%                                                                                   year, BoT’s MPC proposes its inflation target for the next year
     Dec-07       Dec-08      Dec-09        Dec-10      Dec-11
                                                                                         for Cabinet approval. For 2009 - 2011, BoT’s inflation target
Source: Asian Bonds Online
                                                                                         was set between 0.5% and 3.0%.
Exhibit 108
Historical Foreign Ownership of Government Bonds                                         The Monetary Policy Committee (MPC) signals shifts in
                                                                                         monetary policy stance through announced changes in the
   %
 14                                                                                      key policy rate, set at the 1-day repo rate (Bloomberg Ticker:
                                                                                         BTRR1DAY Index). BoT also offers collateralized deposit and
 12
                                                                                         lending facilities (Standing Facility) to banks at O/N repo rate
 10
                                                                                         ±50bp respectively under the end-of-day Liquidity Adjustment
   8                                                                                     Window.
   6
                                                                                         Open Market Operations (OMOs) are the most actively used
   4
                                                                                         instrument to maintain policy rate. The BoT employs five
   2                                                                                     types of OMOs, namely:
   0
                                                                                             -    Repo and reverse repo transactions,
    2003      2004    2005     2006     2007     2008     2009    2010      2011
Source: Morgan Stanley Research estimates                                                    -    Outright purchases and sales of government bonds
                                                                                             -    Issuance of Bank of Thailand bills/bonds
FX
                                                                                             -    FX swaps
The Thai FX market, though deliverable, is split into
                                                                                             -    Electronic BOT Debt Security (e-PN) Window
onshore/offshore segments that are not fungible. Investors
with supporting securities transactions, direct investments,                             BoT also uses the reserve requirement for commercial banks
trade/commercial needs can use the onshore market (NRBS,                                 as a monetary policy tool.
Non-Resident Bank account for Securities) for both spot and
                                                                                         Derivatives
forward transactions. Other investors are restricted to offshore
general purpose accounts (NRBA, Non-Resident Bank                                        Interest Rate Futures: Currently 3 types of interest rate
Account). Onshore and offshore banks may trade freely with                               futures are traded on the Thai futures exchange: (i) 5y govt
each other in the spot market using the NRBA account,                                    bond futures (Bloomberg: TOBA Comdty), (ii) 3m BIBOR
subject to credit limits.                                                                futures (Bloomberg: TORA Comdty), (iii) 6m THBFIX futures
                                                                                         (Bloomberg: TXBA Comdty)
The THB forward market is almost entirely segregated
between onshore and offshore entities due to BoT restrictions;                           Interest Rate Swaps: The fixing for the interest rate swaps is
at times this has caused notable funding squeezes (this is                               the 6 month FX implied forward-implied interest rate, i.e., the
why foreign financial institutions typically maintain a long THB                         6m THB SOR (Bloomberg Ticker: THFX6M Index). Both
cash position in their NOSTRO accounts) as no bank may                                   onshore and offshore IRS curves are active up to 10 years.

                                                                                                                                                     61
                                                                MORGAN STANLEY RESEARCH

                                                                March, 2012
                                                                EM Local Markets Guide




Cross Currency Swaps: The offshore non- deliverable cross       Relevant Websites
currency swap market is generally illiquid, while foreign
                                                                Bank of Thailand
access to the active onshore CCS market is restricted by BoT.
                                                                http://www.bot.or.th/english
Taxation                                                        Thailand’s Ministry of finance
                                                                http://www2.mof.go.th/index.php
Interest income, in general, is subject to a 15% withholding
                                                                National Statistical Office, Thailand
tax, while income derived from forex transactions is taxed at
                                                                http://web.nso.go.th/
the rate of 3%. Institutional foreign investors operating in
                                                                Thailand bond Market Association
Thailand are liable for corporate income tax but exempt from
                                                                http://www.thaibma.or.th/
withholding tax. Foreign institutional investors which do not
                                                                Morgan Stanley AXJ Rates Strategy Bloomberg Page: IRAX
operate in Thailand are subject to a 15% withholding tax on
government and corporate bonds; though such taxes may be
covered by double taxation treaties between Thailand and the
investor’s country.




                                                                                                                    62
                 MORGAN STANLEY RESEARCH

                 March, 2012
                 EM Local Markets Guide




CEEMEA Markets




                                           63
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Czech Republic
Monetary Policy                                                   FX
The Czech National Bank’s (CNB) primary objective is to           The Czech Koruna (CZK) is a free-floating currency that is
maintain price stability. With effect from January 2010, the      fully deliverable and convertible. The CNB does not intervene
inflation target is annual CPI growth of 2% (with a band of       in the FX market. Spot is mainly quoted against the euro,
±1%). The Bank Board is made up of seven members – the            although other major crosses are common as well. In the spot
CNB governor, 2 vice governors and 4 members, all of whom         market, the average daily market volume is around EUR
are appointed by the President of the Republic for 6 years.       1.5bn, with average transaction size of EUR 5m.

The Board meets 8 times a year to decide on monetary policy,      FX forwards and swaps are deliverable, with the best liquidity
and minutes of each meeting are usually published 8 days          in tenors of one year or less. The average daily market
later. The Bank’s main publication is the quarterly Inflation     volume is around EUR 4bn, with transaction size ranging from
Report, which sets out the forecast for inflation at the          EUR 10m for forwards, to EUR 30m for swaps. Settlement is
monetary policy horizon (about 12-18 months ahead).               T+2.

Exhibit 109                                                       The FX option market is generally liquid, with best liquidity in
CNB forecast for inflation                                        EUR/CZK options of one year or less. Liquidity in USD/CZK
                                                                  options is lower.

                                                                  Fixed Income
                                                                  The Ministry of Finance is responsible for debt management
                                                                  and issuance, with the CNB acting as the Ministry’s agent in
                                                                  the primary market. Around 80% of total debt outstanding is
                                                                  domestic, 14% of which is in T-bills. Average time to maturity
                                                                  of domestic bonds is 6.2 years (at end-2011).

                                                                  Exhibit 110
                                                                  Currency Composition of Government Debt

                                                                                       21%


Source: CNB
The main policy instrument used by the CNB is the 14-day
repo tender; the 2-week repo rate is therefore the key rate                                                                             Domestic

which the Bank targets. Tenders are usually announced three                                                                             External
times a week at around 9.30am. Repos with shorter maturities
are executed from time to time depending on the forecasts of
banking sector liquidity. Due to the systemic liquidity surplus
                                                                                                                     79%
in the Czech banking sector, 2-week tenders are currently
used exclusively for absorbing liquidity.                         Source: Ministry of Finance. Total debt was CZK1,495bn at end-2011.

As an EU member state, the Czech Republic is required to          The most liquid bonds are either the current benchmark 5y
take steps to be prepared (i.e., convergence criteria) to join    and 10y bonds that have been tapped several times, or off-
the euro area at some point in the future, at which time the      the-run 5y and 10y bonds with significant outstanding size.
CNB will cede independent monetary policy to the ECB.             Fixed-rate bonds with annual coupons are typically issued in
However, setting the date for joining the euro area is within     3y, 5y, 10y and 15y maturities. There are four floating-rate
the competence of the member state, and the Czech Republic        notes (FRNs) as well, with maturities ranging from 3y to 10y
has not yet decided to set a target date for euro area entry.     and a reference rate of 6m PRIBOR. There is no inflation-
                                                                  linked market. In 2011 the Ministry started for the first time
                                                                  issuing government savings bonds aimed at retail investors.

                                                                                                                                                   64
                                                                                    MORGAN STANLEY RESEARCH

                                                                                    March, 2012
                                                                                    EM Local Markets Guide




Exhibit 111                                                                         Interest Rate Derivatives
Fixed Income Instruments
                                                                                    The single-currency Interest Rates Swaps (IRS) are the most
                     Bloomberg
 Instrument                             Tenor                  Coupon               liquid interest-rate derivative instrument. The IRS is traded up
                     Ticker
 T-bill              CZTB               3, 6 and 12 months     Zero                 to 20y, although it is generally liquid up to 10y. Tenors up to
 T-bond              CZGB               3-50 years             Fixed, floating
                                                                                    and including 1y fix against 3m PRIBOR (Prague Interbank
Source: Morgan Stanley, Bloomberg.                                                  Offered Rate), while tenors of 2y and longer fix against the 6m
                                                                                    PRIBOR. Average daily market volume is US$50k in DV01,
Primary Auctions                                                                    with average transaction size of US$5-10k in DV01 (bid/offer
                                                                                    around 4bp). Settlement is T+2.
The MinFin publishes a monthly issuance calendar on its
website. The CNB holds primary auctions of government                               The quotations for the calculation of PRIBOR rates are
bonds to direct participants (DPs), of which there are currently                    submitted by reference banks (11 at the time of publication) to
13. Responsibilities of DPs include 1) accounting for at least                      the CNB every business day between 10.30 a.m. and 10.45
1% of the total yearly turnover of all DPs on the secondary                         a.m. local time. The calculation (fixing) of PRIBOR rates takes
market, and 2) subscribing to at least 3% of the total nominal                      place at 11.00 a.m. local time
volume of government bonds offered at auction.                                      For tenors of 2y and less, there is also the Forward Rate
Auctions take place on Wednesdays in multiple price format;                         Agreement (FRA) market, liquid mainly up to 9X12. Liquidity
settlement is T+3. Bids need to be submitted by 12pm local                          and average daily volume are similar to the IRS market.
time, and results are out by 12.15pm local time. At least 85%                       The cross-currency basis is also fairly liquid, and trades out to
of the nominal value of the auction must be sold in the first                       10y. Typical bid/ask is 3-4bp and average transaction size is
round to competitive bids. When the first round results are                         EUR 25-50m.
released, a second (non-competitive) round is offered for the
remainder at the weighted average price of the first round.                         Useful Websites
Investors of Government Debt                                                        Czech National Bank (CNB)
                                                                                    www.cnb.cz
Local banks are the largest holders of domestic Czech debt,
accounting for around 43% of the total. Pension funds and                           Ministry of Finance
insurance companies are also large holders at 27% of total.                         www.mfcr.cz
Non-residents own around 14% of the market. Czech bonds                             Association of Pension Funds
are not included in the main local currency indices.                                www.apfcr.cz
Note though that the financial system in the Czech Republic is                      Czech Insurance Association
largely foreign-owned, and foreign-controlled financial                             www.cap.cz
institutions own about 67% of domestic government debt.
                                                                                    Financial Markets Association
Exhibit 112
                                                                                    http://www.aciforex.cz/
Ownership Distribution of CZK-Denominated. State
Debt                                                                                Morgan Stanley Bloomberg page: MSEX
                                                      CZK bn           % of total
 Total                                                  1182              100%
  Banks                                                  510                43%
  Pension funds and insurance co’s                       318                27%
  Non residents                                          168                14%
  Others                                                 187                16%
Source: Ministry of Finance. As of December 2011




                                                                                                                                                  65
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Hungary
Monetary Policy                                                   FX
The Magyar Nemzeti Bank (MNB) has a primary objective to          The Hungarian forint (HUF) is a fully convertible and
achieve and maintain price stability. The MNB’s medium term       deliverable, free-floating currency. The MNB does not
inflation target since 2007 has been an annual increase in the    intervene in the FX market. In the spot market, the HUF’s
Consumer Price Index of 3% (±1%). The main decision-              most commonly traded pair is the EUR/HUF, quoted to two
making body is the Monetary Council, which has seven              decimal places, but there is also good liquidity in USD/HUF.
members (the Governor, the two Deputy Governors, and five         Daily trade volumes average at EUR 2-2.5bn with transaction
other members elected by Parliament for six years). The           sizes averaging at EUR 5m.
Governor is appointed by the President of Hungary upon a
recommendation by the Prime Minister for a term of six years.     The FX forwards have traded on a deliverable basis since
                                                                  2001. The forward rates are usually quoted against USD and
The Council meets twice a month, with any relevant policy         EUR with tenors reaching 5y. The best liquidity can be found
decisions made at the second scheduled meeting. Minutes           with tenors of 1y or less. The forward market has a typical
are published before the next meeting takes place. MNB staff      daily volume of US$200m, while the swap markets daily
produces quarterly inflation projection covering an 8-quarter     volume reaches US$3bn.
horizon, and is published in the Quarterly Report on Inflation.
                                                                  The FX option market is generally liquid, with best liquidity in
Exhibit 113                                                       EUR/HUF options of one year or less. Liquidity in USD/HUF
MNB Fan Chart of the Inflation Forecast                           options is lower.

                                                                  Fixed Income
                                                                  The Hungarian Debt Management Agency (AKK) is
                                                                  responsible for debt management and issuance. Around 50%
                                                                  of total debt outstanding is domestic, 15% of which is in T-
                                                                  bills. Average maturity of domestic bonds is 4.2 years.

                                                                  Exhibit 114
                                                                  Currency Composition of Government Debt




Source: MNB

The main policy instrument is the two-week MNB-bill, the rate                   49.5%
                                                                                                                                   om
                                                                                                                                  D estic
                                                                                                                          50.5%
                                                                                                                                  External
on which is the key policy rate set by the Council. By
changing this base rate the Council influences money market
rates and indirectly the general economy. The MNB-bill is
available once a week to counterparties through a fixed rate
auction. The MNB also has overnight standing facilities aimed
at preventing extreme fluctuations in interbank rates, fx spot
and swap tenders, and a two-year collateralized lending           Source: AKK. Total debt was HUF 20,500bn at end-2011.
facility aimed at improving banks’ lending capacity.
                                                                  The most liquid bonds are the on-the-run benchmarks,
As an EU member, Hungary has a legal obligation to join the       typically with 3y, 5y and 10y maturities. Fixed-rate bonds with
eurozone and comply with the relevant convergence criteria.       annual coupons are normally issued in 3y, 5y, 10y and 15y
Eurozone membership is a medium-term objective at present.        maturities, while FRNs are issued and tapped much less


                                                                                                                                             66
                                                                               MORGAN STANLEY RESEARCH

                                                                               March, 2012
                                                                               EM Local Markets Guide




frequently. The FRN coupons are typically six-month average                    inflows by foreign investors starting from end-2010, followed
of 3m T-bill yields. There are no inflation-linked bonds.                      by the de facto nationalization of private pension funds
                                                                               resulting in the transfer of their assets to the government in
Exhibit 115
                                                                               Q2 2011, led to a sharp increase in foreigners’ share of the
Fixed Income Instruments
                                                                               HGB market from 26% at end-2010 to 39% at end-2011.
                    Bloomberg
 Instrument                          Tenor                   Coupon
                    Ticker
                                                                               Exhibit 117
 T-bill             HTB              3, 6 and 12 months      Zero
                                                                               Ownership Distribution of HGBs, %
 T-bond             HGB              3, 5, 10 and 15 years   Fixed, floating
Source: Morgan Stanley, Bloomberg.                                               45%

                                                                                 40%

Primary Auctions                                                                 35%

                                                                                 30%
The AKK announces an issuance plan for the forthcoming
                                                                                 25%
three-month period on the 15th of every month. Planned
issuance sizes are announced for each maturity a week                            20%


before the auction.                                                              15%

                                                                                 10%
Issuance is via Primary Dealers (PDs), of which there are 15
                                                                                  5%
at present. PDs are required to buy at least 3% of T-bills and
                                                                                  0%
3% of T-bonds sold in auctions in every calendar half-year.                         2000   2001     2002     2003   2004   2005   2006     2007     2008     2009     2010     2011


Bond auctions are competitive, multiple-price format and take                                     Banks (inc NBH)      PFs and Insurance          Foreign investors          Others

                                                                               Source: AKK, Morgan Stanley
place on Thursday; settlement is T+4. A non-competitive
allocation is also available at the average price of the
                                                                               Interest Rate Derivatives
competitive round. The maximum amount of non competitive
bids is HUF 100m of par if the announced auction amount is                     The single-currency Interest Rates Swaps (IRS) are the most
less than HUF 10bn and HUF 200m if the auction amount is                       liquid derivative instrument, and are generally liquid up to 10y.
more than HUF 10bn.                                                            Tenors up to and including 1y fix against 3m BUBOR
                                                                               (Budapest Interbank Offered Rate), while tenors of 2y and
Exhibit 116
                                                                               longer fix against 6m BUBOR. Average daily market volume
Auction Calendar and Format
                                                                               of the IRS market is US$75k in DV01, with average
                                                                               transaction size of US$5-10k in DV01 (bid/offer around 5bp).
 3-month T-bills                 Tuesdays, weekly                              Settlement is T+2.
 12-month T-bills                Thursdays, biweekly
                                                                               BUBOR fixings are calculated every business day by the NBH
 Interest bearing bills          Tuesdays, Weekly
                                                                               based on quotes submitted by 16 banks. The fix is calculated
                                                                               as the arithmetic mean of contributions, excluding the two
 Bonds                           Thursdays, biweekly
                                                                               lowest and two highest. Contributions are submitted at
 Bids in by                      10.50am local time                            10.30am local time, and the fixings are published at 11am.
 Results out by                  11.30am local time
Source: Morgan Stanley                                                         For tenors of 2y and less, there is also the Forward Rate
                                                                               Agreement (FRA) market, liquid mainly up to 9x12. Liquidity
Investors in Government Debt                                                   and average daily volume are similar to the IRS market.

Foreign investors are the largest owners of local government                   The cross-currency basis is less liquid, and trades out to 10y.
bonds, representing around 40% of the market at the end of                     Typical bid/ask is 10-15bp in 5y, and average transaction size
2011. Banks are also large holders at 34%, followed by                         is EUR 25m.
pension funds and insurance companies at 17%.

Non-resident holdings fell sharply during the 2008 crisis.
However, foreign investors’ holdings increased sharply in
2011, both in absolute terms and as % of total. Increasing


                                                                                                                                                                                      67
                                                     MORGAN STANLEY RESEARCH

                                                     March, 2012
                                                     EM Local Markets Guide




Useful Websites
National Bank of Hungary (NBH)
http://english.mnb.hu/

Hungarian Debt Management Agency (AKK)
www.akk.hu

Hungarian Central Statistical Office
www.ksh.hu

Association of Hungarian Investment Fund and Asset
Management Companies
www.bamosz.hu

Hungarian Foreign Association
http://www.acihungary.hu

Portfolio.hu – Online Financial Journal
www.portfolio.hu

Morgan Stanley Bloomberg page: MSEX




                                                                               68
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Israel
Monetary Policy                                                    towards US$50m. Daily turnover averages at around US$3bn.
                                                                   Note that liquidity is lower every Friday since banks in Israel
The Bank of Israel’s (BoI) main objective is to maintain price     close early for Sabbath.
stability, defined as keeping annual inflation within the 1-3%
range defined by the government. Until October 2011, interest      Forwards are deliverable to 5y. Typically, transaction sizes
rate decisions were made solely by the BoI Governor. Since         are US$20m for forwards and US$50m for swaps, with daily
October 2011, however, a six-member Monetary Committee             volumes averaging at US$1.4bn. The FX option market is
composed of the Governor, the Deputy Governor, an                  generally liquid, with best liquidity found in USD/ILS options.
additional Bank employee appointed by the Governor, and            Options are available out to 5y, and daily trading volumes are
three external members, is in charge of monetary policy            at around US$20m.
decisions. The Governor is appointed by the President of           Exhibit 118
Israel for a term of five years, at the recommendation of the      BoI Intervention
government.
                                                                      4,500                                          Big intervention following            4.4
                                                                                                                     sharp rise in ILS vs USD
The Committee meets on the last Monday of each month to               4,000          ILS Strength
                                                                                                                                                           4.2
determine interest rates, and minutes are published about two         3,500
                                                                                                                                                           4.0
weeks after each meeting. The Bank’s main publication is the          3,000
                                                                      2,500                                                                                3.8
Monetary Policy Report, published biannually about one
                                                                      2,000
month after the end of each half year.                                                                                                                     3.6
                                                                      1,500
                                                                                                                                                           3.4
The effective policy rate is the rate at which the BoI lends to       1,000
                                                                                                                                                           3.2
commercial banks, primarily in collateralized monetary                  500
auctions for deposits/loans of daily, weekly and monthly                -                                                                                  3.0
                                                                            Jan-08           Jan-09         Jan-10              Jan-11            Jan-12
maturity. Other instruments used are overnight loans and
                                                                              BoI monthly net USD purchases, USDmn                       USD/ILS (RHS)
deposits at a spread to the policy rate, makam (BoI bill)
sales/purchases and repo auctions.                                 Source: Bank of Israel, Bloomberg, Morgan Stanley Research



FX                                                                 Fixed Income

The Israeli shekel (ILS) started a de facto move away from a       The Ministry of Finance’s Government Debt Management Unit
crawling peg regime in June 1997 and has followed an               (GDMU) is responsible for issuance and overall debt
official free float regime since June 2005. In practice,           management strategy. As of end-2011, around 82% of
however, the currency was freely floating several years before     government debt is domestic (tradable 60% plus non-tradable
the official regime change in June 2005. The BoI is known to       22%) and 18% external. TBills represent only about 2% of
intervene in the FX market. Since 2009 it has followed a           tradable domestic debt and average time to maturity of
policy of ad hoc discretionary intervention, leaving behind a      tradable bonds is 6.3 years.
reserve accumulation policy. While there are no formal levels,
the BoI typically intervenes during periods of sharp ILS
strength to try to smooth the pace of appreciation and act
against excessive speculative flows. When the BoI comes to
market, usually through locals, it purchases within a range of
US$100m-1bn. During periods of extreme ILS strength, the
BoI has made net purchases of as much as $ 4.1bn in a
month. However, there have not been any notably large
interventions over the past 6 months.

The ILS is both fully deliverable and convertible. The spot
market has a fair level of liquidity where transactions are most
commonly around US$10m, although this figure is rising up


                                                                                                                                                           69
                                                                               MORGAN STANLEY RESEARCH

                                                                               March, 2012
                                                                               EM Local Markets Guide




Exhibit 119                                                                    Exhibit 120
Composition of Government Debt                                                 Fixed Income Instruments
                                                                                Instrument        Bloomberg Ticker          Tenor           Coupon
                  18%
                                                                                Makam             MAKAM                     12 months       Zero
                                                                                T-bills           ILTBIL                    6-12 months     Zero
                                                                Tradable
                                                                domestic        ILGOV             ILGOV                     3-30 years      Fixed
                                                                Non-tradable
                                                                domestic        Shahar*           SHAHAR                    10 years        Fixed
                                                                External
                                                                                ILFRN             ILFRN                     10 years        Floating
         22%
                                                        60%
                                                                                Gilon*            GILON                     10 years        Floating
                                                                                ILCPI             ILCPI                     3-30 years      Inflation-linked
                                                                                Galil*            GALIL                     10-20 years     Inflation-linked
                                                                               Source: Morgan Stanley. *No longer issued at auctions



Source: GDMU. Total amount outstanding as of 31-Dec-2011 is ILS 633bn.
                                                                               Primary Auctions

Fixed rate bonds with annual coupons are the most liquid                       The MinFin announces an annual schedule with auction dates
fixed-income instrument in Israel. They were issued as                         at the end of the year. Issuance sizes are published on a
‘Shahar’ until 2006 and ‘ILGOV’ since 2006. The MinFin                         quarterly basis on the first month of each quarter, and a
typically issues new bonds in 3y, 5y, 10y tenors, and                          detailed monthly issuance plan is published at the last
occasionally 20y and 30y. The most liquid bonds are 3y, 5y,                    Wednesday of every month.
and 10y benchmarks.                                                            There are currently 14 Primary Dealers (PDs) and they are
Israel has the most liquid inflation-linked bond market in                     required to purchase each year the smaller of a) ILS 2bn, or
CEEMEA. Linkers were issued as ‘Galil’ until 2004 and                          b) 5% of fixed coupon bonds issued during a calendar year in
‘ILCPI’ since 2006. Both the principal and interest of linkers                 the competitive auctions only.
are indexed to the latest available Consumer Price Index                       Bond auctions are competitive, multi-price and normally take
(CPI), as published by the Central Bureau of Statistics. ILCPI                 place on Mondays. Settlement is T+1. Non-competitive bids
are typically issued in 5y, 10y, and 30y maturities. The most                  can be placed at the average price of the competitive auction
liquid tenors are 5y and 10y.                                                  up to 24 hours from the auction closing time. The six highest
The floating-rate bonds were called ‘New Gilon’ until 2002,                    ranking PDs are entitled to purchase 20% of the amount
and ‘ILFRN’ since 2007. There are currently only two ILFRNs                    purchased in the competitive auction, all others are entitled to
outstanding, maturing in 2017 and 2020. Interest is paid once                  purchase 10%.
every three months on the last business day of the month.                      Exhibit 121
The annual interest rate is the weighted average yield on                      Auction Calendar and Format
TBills with a 3-12 month maturity in the 10-day period
                                                                                Bills                   Monday
preceding the interest period.
                                                                                Bonds                   Monday
‘ILTBIL’ ‘and Makam’ are short-term Bills issued by the MinFin                                          Competitive multiple price, with a non-competitive
                                                                                Format
and the BoI respectively, with maturity of up to 12 months.                                             component at average price of competitive auction
                                                                               Source: Morgan Stanley
The Makam market is much larger and more liquid than the
TBill market.
                                                                               Investors in Government Debt
                                                                               The domestic savings industry and the Israeli public are the
                                                                               largest investors in domestic bonds, holding 46% and 23% of
                                                                               domestic tradable bonds respectively. Banks (13%) and
                                                                               insurance companies (10%) are also important players. Non-
                                                                               residents own only about 3% of the market, partly because
                                                                               Israeli local currency debt does not feature in most local
                                                                               currency bond indices.


                                                                                                                                                               70
                                                                                                                  MORGAN STANLEY RESEARCH

                                                                                                                  March, 2012
                                                                                                                  EM Local Markets Guide




Exhibit 122                                                                                                       Average daily market volume of the IRS market is US$75-
Ownership Structure of Tradable Bonds (%)                                                                         100k in DV01, with average transaction size of US$5-10k in
  60
                                                                                                                  DV01 (bid/offer around 3-4bp). Settlement is T+2.

                                                                                                                  For tenors of 2y and less, there is also the Forward Rate
  50
                                                                                                                  Agreement (FRA) market, liquid up to18x21. Liquidity and
  40                                                                                                              average daily volume are slightly lower than the IRS market.
  30                                                                                                              The cross-currency basis is not liquid, and trades out to 10y.
                                                                                                                  Typical bid/ask is not wider than 20bp, and average
  20
                                                                                                                  transaction size is US$50 million.
  10



   0
                                                                                                                  Useful Websites
       2000    2001      2002     2003        2004     2005      2006   2007     2008    2009     2010   2011
                                                                                                                  Bank of Israel
       Provident, Mutual, and Pension funds          Insurance     Banks       BoI      Non-residents    Public
                                                                                                                  www.bankisrael.gov.il
Source: Central Bureau of Statistics, Morgan Stanley

                                                                                                                  Israeli Ministry of Finance
Interest Rate Derivatives                                                                                         www.mof.gov.il
The single-currency Interest Rates Swaps (IRS) are the most                                                       The Government Debt Management Unit
liquid instruments. The IRS is traded up to 20y, and is                                                           http://ozar.mof.gov.il/debt/gen/mainpage.asp
generally liquid up to 15y. The floating leg fixes against 3m
                                                                                                                  Central Bureau of Statistics
TELBOR (Tel Aviv Interbank Offered Rate).
                                                                                                                  www.cbs.gov.il
The TELBOR fix is published by Reuters at 12.00am local
                                                                                                                  Tel-Aviv Stock Exchange (TASE)
time and calculated as the average of quotations from 9
                                                                                                                  www.tase.co.il
banks, excluding the highest and lowest rate.
                                                                                                                  Morgan Stanley Bloomberg page: MSEX




                                                                                                                                                                              71
                                                                      MORGAN STANLEY RESEARCH


                                                                      March, 2012
                                                                      EM Local Markets Guide




Poland
Monetary Policy                                                       usually acts to smooth market moves that they deem to
                                                                      excessive. The state-owned BGK bank often carries out FX
National Bank of Poland’s (NBP) primary objective is price            transactions on behalf of the Ministry of Finance.
stability via direct inflation targeting. Since 2004, the inflation
target set by the Monetary Policy Council (MPC) of the NBP            The zloty is now fully convertible and deliverable, and the
has been a 2.5% annual increase in headline CPI with a band           most common traded currency pairs are with the EUR and
of ±1%. The Council consists of the President of the NBP as           USD. There are no restrictions on capital flows between
Chairperson and 9 other members appointed in equal                    Poland and the EU states, and international investors enjoy
numbers by the President of Poland, the Sejm and the Senate           the same treatment as local investors. The spot market has
for a term of 6 years.                                                very good liquidity with EUR 2.5-3bn in daily volumes in
                                                                      normal conditions, and transactions averaging at EUR 5m.
The MPC meets once a month and Minutes of each meeting
are published two weeks later. The NBP also publishes an              Forward market transactions average to EUR 20m, and trade
Inflation Report three times per year with CPI projections for        at 3pips for 3m and 20pips for 12m, showing very good
the next two years.                                                   liquidity. Forwards are traded on both a deliverable and non-
                                                                      deliverable basis with tenors reaching 20y, although the best
Exhibit 123                                                           liquidity is found at tenors of 1y or less. Daily volumes are
NBP yoy CPI Inflation Forecast                                        US$0.2bn for forwards and US$3.5bn for swaps. Options on
                                                                      the PLN are also available in the USD and EUR pairs. Tenors
                                                                      stretch out to 7y and generally trade in sizes of EUR 10-20m.

                                                                      Fixed Income
                                                                      The Polish Ministry of Finance is responsible for debt
                                                                      management and issuance for the country, with the NBP
                                                                      acting as its issuing agent. As of end-2011, around 70% of
                                                                      the debt is denominated in local currency and the rest in
                                                                      foreign currency. TBills account for just 2.4% of tradable
                                                                      domestic debt, and the average maturity of domestic bonds is
                                                                      4.35 years.

                                                                      Exhibit 124

Source: NBP                                                           Currency Composition of State Treasury Debt
The NBP’s main policy tool is the issuance of 7-day NBP
money market bills, the minimum yield on which is the
effective policy rate. Other instruments used are reserve                            31%
requirements and overnight collateralized loans and deposits.

As a new EU member, Poland is also required to at some                                                                                         Domestic
adopt the euro. However, given recent events and the fact
                                                                                                                                               External
that a currency change in Poland requires a constitutional
amendment, Polish officials have indicated that adoption of
the euro is unlikely to happen in the foreseeable future.                                                                         69%


FX
The Polish zloty (PLN) has been a free floating currency since        Source: Ministry of Finance. As of end-2011. Total amount is PLN 767bn

April 2000, before which it followed a crawling peg regime.
The NBP is known to intervene in the market, though there
are no formal levels or published intervention sizes. The NBP

                                                                                                                                                          72
                                                                                       MORGAN STANLEY RESEARCH

                                                                                       March, 2012
                                                                                       EM Local Markets Guide




Poland issues several different types of bonds. The most                               required to purchase a minimum of 0.5% (less than 4 years
frequently issued bonds are fixed-rate, annual coupon, with a                          maturity) and 1.5% (greater than 4 year maturity) of treasury
maturity of 5y and 10y (PS, DS, WS series), but 20y and 30y                            securities sold in auctions during a calendar year.
bonds have also been issued in the past. The second largest
                                                                                       From January 1, 2012, all T-bond and T-bill auctions are
segment of the market are zero coupon bonds (OK series),
                                                                                       carried out in uniform price formula instead of the previously
with a maturity of 2y. The most liquid bonds are the on-the-run
                                                                                       used multiple price format. All successful bidders pay the
2y, 5y and 10y issues, as well as bonds with large amounts
                                                                                       minimum accepted price in the auction. Also from 2012, the
outstanding.
                                                                                       MinFin introduced the option to place non-competitive bids,
Exhibit 125                                                                            the total size of which will not exceed 5% of the total amount
Domestic Marketable Securities by Type                                                 sold in an auction.
                                                  PLN bn                  % of total
                                                                                       Exhibit 127
 Total debt                                          507                      100%     Auction Calendar and Format
    Bonds                                            495                       98%
                                                                                         Bills                                      Mondays
     Coupon                                          303                       60%       Bonds                                      Thursdays
     Zero (2y bonds)                                 109                       21%       Bids in by                                 10.45am local time
     Floating                                         65                       13%       Results out by                             12.30pm local time
     Inflation-linked                                 18                        4%                                                  Competitive uniform price, with a non-
                                                                                         Format
                                                                                                                                    competitive allocation of up to 5%
    Bills                                             12                        2%     Source: Ministry of Finance, Morgan Stanley

Source: Ministry of Finance. As of end-2011
                                                                                       Investors in Government Debt
The MinFin also issues floating-rate notes (WZ series) in 4y-                          The domestic non-bank sector is the largest holder of
11y tenors with coupons linked to 6M WIBOR. Finally, there                             domestic debt with 49% of the market as of end-2011.
are two inflation-linked bonds (IZ series) with a remaining                            Foreign investors own 30% and banks 20%.
maturity of 5y and 11y linked to Polish CPI.
                                                                                       Exhibit 128
Exhibit 126
                                                                                       Ownership Structure of Tradable Domestic Debt (%)
Fixed Income Instruments
                     Bloomberg                                                           70%
 Instrument                              Tenor             Coupon
                     Ticker
                                                                                         60%
 T-bills             PTB                 6m-12m            Zero
                                                                                         50%
 OK series           POLGB               2-year            Zero
 PS series           POLGB               5-year            Fixed                         40%

 DS series           POLGB               10-year           Fixed                         30%

 WS series           POLGB               >10-year          Fixed                         20%

 WZ series           POLGB               4-11year          Floating (6M WIBOR)
                                                                                         10%
 IZ series           POLGB               10- to 15-year    Inflation-linked
                                                                                          0%
                                                                                            2000      2001   2002     2003   2004    2005    2006    2007   2008      2009      2010   2011
Source: Morgan Stanley
                                                                                                                Banks           Non-banking sector          Foreign investors

Primary Auctions                                                                       Source: Ministry of Finance.

The MinFin announces an annual issuance plan with auction                              Assets managed by non-banking financial institutions started
dates and bond types in December. In addition, they                                    to increase noticeably in the early part of the last decade
announce a detailed monthly auction calendar with specific                             following the pension reform of 1999 and the establishment of
bonds and planned sizes at the end of each month.                                      funded open pension funds (OFE). However, the MinFin
Auctions are organised by the NBP as the government’s                                  expects OFE’s purchases of government debt to gradually
issuing agent and usually take place on Thursday. Settlement                           decline in the future due to recent changes in contribution
is T+2. Bids are submitted through a system of Primary                                 rates instituted as part of the government’s deficit reduction
Dealers (PDs), of which there are currently 14. PDs are                                plan. Namely, in May 2011 employee contributions to OFE


                                                                                                                                                                                       73
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




were lowered from 7.3% to 2.3%, and diverted towards newly         POLGBs and inflows accelerated quickly as soon as the
created subaccounts within the pay-as-you-go system,               external risk environment improved in early 2012.
managed by ZUS, Poland’s social insurance institution.
                                                                   Exhibit 131
Contributions to OFE will increase only gradually to 3.5% by
                                                                   Non-Residents’ Holdings of Local Debt, PLN bn
2017. In addition, the proportion of OFE’s assets that can be
invested in equities will be increased from 40% to 62% by            180

2020.                                                                160


The MinFin expects the impact on the debt market to be               140

neutral, as the decline in the demand from OFE will be               120
accompanied by a decline in debt issuance as the Treasury
                                                                     100
reduces the payment it makes to ZUS.
                                                                      80
Exhibit 129
Non-Banking Financial Institutions Assets, PLN bn                     60


                                                                      40
                                                                       Dec-05 Jun-06 Dec-06 Jun-07 Dec-07 Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11

                                                                                                        Non-resident holdings of local debt


                                                                   Source: Ministry of Finance.



                                                                   Interest Rate Derivatives
                                                                   The single-currency Interest Rates Swaps (IRS) are the most
                                                                   liquid instruments. The IRS is traded up to 20y, although it is
                                                                   generally liquid up to 10y. Tenors up to and including 1y fix
                                                                   against 3m WIBOR (Warsaw Interbank Offered Rate), while
                                                                   tenors of 2y and beyond fix against 6m WIBOR.
Source: Ministry of Finance.

                                                                   The WIBOR fix is published by ACI Polska, the Polish
Exhibit 130
                                                                   Financial Markets Association. The fix is calculated as the
Share of Treasury Securities in Non-Banking
                                                                   arithmetic mean of contributions from 14 local banks, trimmed
Financial Institutions Assets (%)
                                                                   of the lowest two and highest two contributions, at 11am local
                                                                   time.

                                                                   Average daily market volume of the IRS market is US$100-
                                                                   120k in DV01, with average transaction size of US$5-15k in
                                                                   DV01 (bid/offer around 3bp). Settlement is T+2.

                                                                   For tenors of 2y and less, there is also the Forward Rate
                                                                   Agreement (FRA) market, liquid up to 21x24. Liquidity and
                                                                   average daily volume are slightly lower than the IRS market.

                                                                   The cross-currency basis is liquid, and trades out to 10y.
                                                                   Typical bid/offer for small sizes is 5bp and average size is
Source: Ministry of Finance.                                       EUR 25-50m.

Non-residents’ share of local market dipped significantly in the
2008 crisis, but recovered quickly from 13% at the end of
2008 to 30% at the end of 2011. Foreign inflows into the
POLGB market are correlated with the wider international risk
environment. Although inflows slowed during the eurozone
crisis at the end of 2011, there were no major outflows from



                                                                                                                                                                    74
                                                    MORGAN STANLEY RESEARCH

                                                    March, 2012
                                                    EM Local Markets Guide




Useful Websites
National Bank of Poland
www.nbp.pl

Polish Ministry of Finance
www.mf.gov.pl

Polish Central Statistics Office
www.stat.gov.pl

Warsaw Stock Exchange (WSE)
www.wse.com.pl

National Depository for Securities
www.kdpw.com.pl

Polish Financial Supervision Authority
www.knf.gov.pl

ACI Polska - Polish Financial Markets Association
www.acipolska.pl



Morgan Stanley Bloomberg page: MSEX




                                                                              75
                                                                            MORGAN STANLEY RESEARCH

                                                                            March, 2012
                                                                            EM Local Markets Guide




Russia
Monetary Policy                                                             FX
The Central Bank of the Russian Federation (CBR) is
                                                                            The Russian rouble (RUB) followed a crawling peg regime
currently in the process of transitioning to an inflation targeting
                                                                            until July 1998, after which the currency devalued and the
regime, with the goal of achieving annual rate of CPI inflation
                                                                            regime shifted to a heavily managed float. Following a second
of 5-6% in 2012, 4.5-5.5% in 2013, and 4-5% in 2014. The
                                                                            devaluation in August 2008 the flexibility around the managed
Bank aims to gradually scale down its direct intervention in
                                                                            float regime increased. The CBR manages the currency
the foreign exchange market and move to flexible exchange
                                                                            against a EUR (45%) and USD (55%) basket.
rate policy.
                                                                            CBR intervention is delivered through MICEX, making it
The CBR’s main decision-making body is the Board of
                                                                            anonymous; however, large-order sizes are often seen as
Directors, composed of the Chairman of the Bank, and ten
                                                                            CBR activity in the market. There is an annual guideline in the
other members (six Deputy Chairmen and four other CBR
                                                                            Main Directions of Monetary Policy to which the CBR
officials). The State Duma appoints the CBR Chairman and
                                                                            adheres. In accordance with the gradual policy shift to an
members of the Board on the proposal of the President of the
                                                                            inflation-targeting regime, the CBR is in the process of
Russian Federation for a term of four years. Board meetings
                                                                            relaxing its control over the exchange rate.
take place once a month but with no exact date pre-
announced. A brief explanatory statement is published after                 There are no restrictions on capital account transactions,
each meeting with the approximate date of the next meeting                  making the rouble both fully convertible and tradable. The
given. The Bank publishes a Quarterly Inflation Review and                  spot market has very good liquidity with daily volumes of
once a year, Guidelines for the Single State Monetary Policy                US$15 billion and transaction sizes averaging at US$25
covering the following three years.                                         million. The forward and swap markets have average
                                                                            transaction sizes at US$25 million. The option market has
The operational target of the CBR’s interest rate policy is the
                                                                            maturities available to 5y and average daily volumes and
short-term interbank money market rate, which it influences
                                                                            transaction sizes of US$300 million and US$30 million,
via a variety of instruments. The main liquidity absorption tool
                                                                            respectively.
which sets the lower bound for market rates is the standing
overnight deposit facility. The main liquidity provision                    Exhibit 133
instruments are the standing overnight loan facility and short-             CBR Intervention
term open market repo operations. Other tools used by the                    40,000
                                                                                                                                       RUB Strength
                                                                                                                                                                41

Bank are reserve requirement ratios, issuance of Bank of
                                                                             20,000                                                                             39
Russia bonds (OBRs) and outright sales and purchases of
government bonds (OFZs).                                                          0                                                                             37

Exhibit 132
                                                                             -20,000                                                                            35
CBR Key Policy Rates and Interbank Rate
                                                                                                                          .
                                                                             -40,000                                                                            33
   7.00

   6.00                                                                      -60,000                                                                            31


   5.00
                                                                             -80,000                                                                           29
                                                                                   Aug-08   Feb-09     Aug-09    Feb-10       Aug-10   Feb-11    Aug-11   Feb-12
   4.00
                                                                                  Net USD Purchases, USDmn       Net EUR Purchases, USDmn         RUB Basket (RHS)
   3.00
                                                                            Source: Central Bank of Russia, Bloomberg, Morgan Stanley Research
   2.00
    19-Jan-11 22-Mar-11 23-May-11 22-Jul-11 22-Sep-11 23-Nov-11 02-Feb-12

          Interbank credit rate, 1 day, %    CBR deposit rate, %
          Repo auction rate, %               Fixed Repo, %
Source: CBR, Morgan Stanley Research




                                                                                                                                                              76
                                                                                  MORGAN STANLEY RESEARCH

                                                                                  March, 2012
                                                                                  EM Local Markets Guide




Fixed Income                                                                      Primary Auctions
The Russian Ministry of Finance (MinFin) is responsible for                       The MinFin announces quarterly issuance calendars on its
debt management and issuance for the country, with the CBR                        website. Auctions are held weekly on Wednesdays by the
acting as the MinFin agent. Around 80% of the government                          CBR in a multiple price format. Dealers can place 20% of all
debt is domestic and 20% external. Russia has no TBills                           their bids as non-competitive which are filled at the average
outstanding, while the average maturity of domestic bonds is                      price of competitive fills. The MinFin issues size and yield
6.4 years (3.1 years for OFZ-PD, see below).                                      guidance ahead of the auction, but dealers can place bids
                                                                                  outside of that guidance. The CBR can also sell bonds on the
Exhibit 134
                                                                                  secondary market. Normal settlement convention is same day
Currency Composition of Government Debt
                                                                                  (T+0), or next day (T+1) after trading hours.

           21%                                                                    Exhibit 136
                                                                                  Auction Calendar and Format
                                                                                   Bonds                    Wednesdays
                                                                                   Bids in by               11-11:30am local time
                                                                                   Results out by           12.30pm local time
                                                              Domestic                                      Competitive multiple-price; 25% of total bids can be non-
                                                                                   Format
                                                                                                            competitive at average of price of competitive fills
                                                              External            Source: Morgan Stanley

                                                                                  Investors in Government Debt
                                                                                  The main holders of OFZs are domestic Russian banks
                                           79%                                    (including CBR), which own 70% of the market. The state
                                                                                  development bank, Vnesheconombank (VEB), owns another
Source: Ministry of Finance. Total amount is RUB 5,975bn                          16% and is the only domestic real-money-type investor in
                                                                                  OFZs through its pension fund mandate. According to MinFin
There are two main types of domestic government bonds:                            published information, foreign investors’ share of the OFZ
Federal Loan Obligations (OFZs) and Federal Savings Bonds                         market is less than 5%, one of the lowest levels in EM.
(GSOs). Issuance and trading is concentrated in OFZ-PD                            However, we think that the actual figure is higher and
(fixed coupon bonds with bullet repayment), and liquidity is                      probably closer to 10%, given difficulties in collecting
best in 3-5y tenors. A significant amount of OFZ-AD (step-up                      ownership information. Also, recent reforms are expected to
coupon with a sinking fund) are outstanding as well. As of                        lead to a substantial increase in foreign investors’ involvement
March 2012, the outstanding amount of OFZ-PD was RUB                              in the OFZ market.
1,900bn, and RUB 1,060bn for OFZ-AD. The outstanding
amount of GSOs is RUB 550bn. OFZ-AD and GSO are mainly                            Exhibit 137
held by domestic investors.                                                       OFZ Holdings by Type of Investor, % of Total
                                                                                                                       Non-residents, 3%
Exhibit 135                                                                                         Other domestic*,
                                                                                                          12%
Composition of Domestic Bonds                                                                                                                    Russian banks,
                                                                                                                                                      30%
                                                                                          Bank of Russia,
                      16%                                                                      10%




                                                                                                VEB, 16%
                                                                                                                                           Sberbank, 29%

                                                                         OFZ-PD
                                                                                  Source: Morgan Stanley Research, MinFin; As of mid-2011; *Incl. private asset managers.
                                                                         OFZ-AD
                                                           54%
                                                                         GSO      Following reforms in 2002, Russia adopted a multi-pillar
           30%                                                                    pension system under the umbrella of the state Pension Fund
                                                                                  of the Russian Federation. All citizens participate in the basic
                                                                                  first pillar, but individuals born after 1967 also accumulate part
                                                                                  of their wage fund into a second-pillar personal account and
Source: Ministry of Finance. As of March 2012. Total amount is RUB 3,510bn
                                                                                  can choose either VEB or a private management company to


                                                                                                                                                                       77
                                                                                   MORGAN STANLEY RESEARCH

                                                                                   March, 2012
                                                                                   EM Local Markets Guide




manage the funds. VEB’s mandate currently allows it to invest                      institutions such as Euroclear, thus opening trading in OFZ to
in bonds issued by the Russian government and its agencies,                        any foreign investor with a Euroclear account. The recent
domestic corporates (backed by deposits) and international                         passage of the Federal Law “On the Central Depository”
financial organizations, as well as mortgage securities. Private                   paves the way for the start of the process of selection of a
pension funds also have significant funds under management                         CD. The Russian regulators have set a deadline of July 1,
(over RUB 1 trillion) but have a very small proportion of their                    2012, but there is a risk this might be delayed into the
funds invested in government bonds.                                                autumn.

                                                                                   Interest Rate Derivatives
Recent Reforms of OFZ Market
                                                                                   The cross-currency swaps (ccy) are the most liquid interest
The Russian bond market will likely be the fastest-growing
                                                                                   rate derivatives in Russia, with the floating leg fixed against
local EM debt market in 2012, rising by 35%. The total
                                                                                   3m USD Libor. Liquidity is the greatest in the 1-3y sector, but
outstanding size of the domestic bond market is expected to
                                                                                   trading exists up to 10y tenors.
reach almost RUB 6 trillion by 2013, a more than four-fold
increase since end-2008. To facilitate this growth and heavy                       The interest rate swaps (IRS) are less liquid than the ccy, and
issuance that it entails, the government has recently taken                        fix against 3m MosPrime for all trades. MosPrime is the
steps to modernize the OFZ market and ease access for                              National Foreign Exchange Association (NFEA) fixing of
foreign investors.                                                                 reference rate based on the offered rates of Russian rouble
                                                                                   deposits as quoted by 13 Contributor Banks. It is calculated
Exhibit 138
                                                                                   as an arithmetic average of the rates provided by banks after
Growth in EM Local Bond Markets in 2012
                                                                                   disregarding the highest and the lowest. It is published every
 40
                                                                                   business day at 12.30pm Moscow time.
 35
 30                                                                                For shorter-dated maturities (<1y), FX forwards are traded
 25                                                                                more commonly. Average daily volume is around US$1bn,
 20                                                                                with average trade sizes of US$10-20m.
 15

 10                                                                                Useful Websites
  5
                                                                                   Central Bank of Russia
  0
      RU   ID    TH   IN   SA SG MY MX PH HU KO TW TU PD CZ HK                     www.cbr.ru
                2012 Net domestic bond issuance / Total domestic bond market       Ministry of Finance
Source: Morgan Stanley Research; Figures for Russia include OFZ (AD+PD) plus GSO   www.minfin.ru

Starting from February 2012, Russian government bonds                              Moscow Interbank Currency Exchange
trade alongside corporates on the Main Market section of                           www.micex.com
MICEX, which has a much greater number of participants and                         National Foreign Exchange Association
simpler trading procedures. The same trading rules that apply                      www.nva.ru
to corporate bonds now apply to OFZs, including over-the-
counter trading. Initial placement of OFZs will continue to take                   Morgan Stanley Bloomberg page: MSEX
place on the FGS (Federal Government Securities) section of
MICEX.

The next important step in the liberalisation of the OFZ market
should take place in the second half of 2012, when domestic
regulators are expected to select a Central Depository (CD),
which will have the exclusive right to certify security
ownership in Russia. The new CD will have the right to open
nominee accounts for foreign depository and settlement




                                                                                                                                                78
                                                                                          MORGAN STANLEY RESEARCH

                                                                                          March, 2012
                                                                                          EM Local Markets Guide




South Africa
Monetary Policy                                                                           extreme appreciation in the ZAR, and to accumulate FX
                                                                                          reserves. The SARB does not adhere to any formally known
The South African Reserve Bank’s (SARB) primary objective                                 levels or intervention sizes.
is price stability, via an inflation-targeting regime. Since, Feb
2009 the inflation target has been 3 to 6% of headline CPI (for                           In special circumstances, usually related to large foreign
all urban areas). The SARB’s main decision body is the MPC,                               direct investment, the SARB engages in off-market
consisting of the Governor, three deputy governors, and four                              transactions to purchase the related FX inflows. Local
senior officials of the bank. Each member holds a single vote.                            corporates need SARB approval to externalise their assets.
                                                                                          Domestic unit trusts (mutual funds) are allowed to invest up to
The timetable for MPC meetings is available before the start                              30% of their assets offshore, while insurance companies and
of the year. The MPC meets every two months, although the                                 pension funds have a 20% cap on offshore allocations.
SARB may conduct unscheduled meetings should they deem
it necessary. The meeting process includes three days of                                  In the spot market, USD/ZAR is the most commonly traded
formal deliberations, followed by a consensus vote. The                                   pair. With a bid/offer spread at 50pips, daily volume at US$3
decision is announced at a media conference (usually early                                billion and transaction sizes averaging at US$5 million, the
afternoon, GMT), webcast live on the SARB website, and the                                currency has very good liquidity.
press release is also published. The SARB also publishes a                                FX forwards on the ZAR are offered with tenors reaching 10y
Monetary Policy Review (MPR) twice a year, with CPI                                       and are generally most liquid up to 1y. Forwards and swaps
projections for the next 2 years.                                                         are deliverable and have daily volumes at around US$1 billion
Exhibit 139                                                                               and US$9 billion, respectively. The ZAR also has a liquid and
SARB yoy CPI inflation* forecast                                                          developing options market. Note that options have transaction
                                                                                          sizes at around US$20 million and tenors similar to that of the
                                                                                          forward market.

                                                                                          Fixed Income
                                                                                          The Treasury is responsible for debt management and
                                                                                          issuance for the country, with the SARB acting as its issuing
                                                                                          agent. As of 3Q11, around 90% of total debt outstanding was
                                                                                          denominated in local currency, 14% of which are in CPI
                                                                                          linkers and 16% in Bills. The average maturity of tradable
                                                                                          fixed bonds is 9.8 years, and for linkers is 12.1 years.

                                                                                          Exhibit 140
Source: SARB; MPR Nov’11. *CPIX for metropolitan and other urban areas until the end of   Total Debt Composition
2008; CPI for all urban areas thereafter.
                                                                                                        10%
The SARB’s key policy rate is the repo rate, implemented via
weekly repurchase auctions with the commercial banks, with
a goal to keep short-term rates in line with the policy rate. In
addition, the SARB offers end-of-day facilities to commercial                                                                                     Domestic

banks and uses a range of open market operations, to
manage liquidity and implement their monetary policy stance.                                                                                      External



FX
The South African rand (ZAR) is a freely floating currency,
with no exchange controls to non-residents in trading. The
                                                                                                                           90%
SARB is known to intervene in the FX market, though it                                    Source: Morgan Stanley, Haver. As of 3Q11. Total amount is R966 billion (US$128 billion)
typically acts through locals. It usually intervenes to mitigate


                                                                                                                                                                                79
                                                                           MORGAN STANLEY RESEARCH

                                                                           March, 2012
                                                                           EM Local Markets Guide




Unlike other CEEMEA markets, new issues are infrequent in                  Investors of Government Debt
South Africa, as the Treasury usually taps existing bonds.
                                                                           The Public Investment Corporation (PIC), which manages the
Therefore, issuance is focused on specific bonds, rather than
                                                                           Government Employees Pension Fund, is the single-biggest
benchmark tenors (e.g., 2y, 5y and 10y). The longest-maturity
                                                                           investor of domestic debt with 27% of the market. Other
fixed-rate bond issued in South Africa is 30 years.
                                                                           private, domestic institutional investors are also major holders
Another unusual feature of the bond market is that some                    of local debt with around 30% share. Local banks have
bonds (the R157 and the R186) have a three-way principal                   gradually increased their holdings of local currency debt since
redemption. For example, on September 16, 2012, the R157                   2006 and now own 13% of the market.
(with maturity on September 16, 2015) will split into three
                                                                           Exhibit 143
issues, with equal amounts maturing on September 15, 2014,
                                                                           Ownership Structure of Tradable Domestic Debt (%)
2015 and 2016. The R186 (maturing in December 2026) will
                                                                            70%
split in a similar manner in December 2024. In recent years,
the Treasury has introduced bonds with bullet redemptions                   60%

(e.g., R201, R206, R207, etc.). While the R157 is still the
                                                                            50%
most liquid bond, these new bonds are becoming more and
more liquid as their outstanding size has increased.                        40%


The Treasury also issues inflation linked bonds, with the                   30%

redemption and coupon linked to CPI with a 3-month lag. For
                                                                            20%
example, a trade with value date mid-October 2011 will
reference to CPI interpolated from the June and July CPI                    10%

levels; published in July and August, respectively. The
longest-maturity linker issued is 30 years.                                  0%
                                                                              Jan-00          Jan-02         Jan-04         Jan-06     Jan-08        Jan-10        Jan-12
                                                                                  Public Investment Corporation       Reserve Bank   Banks      Non-monetary Private Sector
Exhibit 141
                                                                           Source: SARB, Morgan Stanley. *Non-monetary private sector includes non-residents
Fixed Income Instruments                                                   Foreign ownership has increased substantially over the last
 Instrument               Bloomberg Ticker     Tenor          Coupon       two years, from 14% of the market in 2009 to 22% in 2010,
 T-bill                   SATB
                                               6 and 12
                                                              Zero         and 29% as of end-2011. Daily data of foreign bond flows is
                                               months
                                               Up to 30       Fixed, CPI   also available (SABO Index), and suggests net inflows
 T-bond                   SAGB, SACPI
                                               years          linked       continue to trend higher, though vulnerable to outflows during
Source: Morgan Stanley
                                                                           periods of wider market weakness.
Primary Auctions
                                                                           Exhibit 144
The SARB conducts auctions on behalf of the Treasury. The                  Cumulative Non-Resident Bond Inflows, ZARbn
auctions for fixed-rate bonds are every Tuesday, and the                    200

bonds to be auctioned are announced on the Wednesday                        180

before. Bills and CPI-linked bonds are auctioned every Friday,              160

and the bonds/bills to be auctioned are announced on the                    140

Monday preceding. Auctions are conducted on a uniform-yield                 120

basis and open only to Primary Dealers, of which there are                  100

currently eight, and results are out at 11.30am local time.                  80

                                                                             60
Exhibit 142
                                                                             40
Auction Calendar and Format                                                  20
 Bills                                       Fridays
                                                                              0
 Bonds                                       Tuesdays
 Inflation-linked bonds                      Fridays                        -20
 Bids in by                                  11am local time                  Jan-06         Jan-07         Jan-08         Jan-09    Jan-10        Jan-11       Jan-12
 Results out by                              11.30am local time                                  Cumalative non-resident bond flows (since Jan'06) ZARbn
 Format                                      Uniform yield                 Source: SARB, Morgan Stanley, Bloomberg
Source: Morgan Stanley




                                                                                                                                                                              80
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Interest Rate Derivatives
The single-currency Interest Rates Swaps (IRS) are the most
liquid instruments. The IRS is traded up to 30y, although it is
generally liquid up to 10y. The floating legs fix against the 3m
Johannesburg Interbank Agreed Rate (JIBAR), and the
convention is Act/365.

The JIBAR fix is calculated daily by SAFEX as the arithmetic
mean of quotes received from nine contributing banks,
excluding the top and bottom two.

Average daily market volume of the IRS market is US$100k in
DV01, with average transaction size of US$5-15k in DV01
(bid/offer around 3-4bp). Settlement is T+0.

For tenors of 2y and less, there is also the Forward Rate
Agreement (FRA) market, liquid up to 9x12. Liquidity (around
5bp on US$5-10k) and average daily volume (US$ 80k) are
slightly lower than the IRS market.

The cross-currency basis trades out to 20y, and is liquid up to
10y. Typical bid/ask is 6bp, and average transaction size is
US$25 million.

Useful Websites
South African Reserve Bank (SARB)
www.reservebank.co.za
National Treasury
www.treasury.gov.za
Bond Exchange of South Africa (BESA)
www.bondexchange.co.za
Johannesburg Stock Exchange (JSE)
www.jse.co.za
South African Futures Exchange (SAFEX)
www.safex.co.za
Morgan Stanley Bloomberg page: MSEX




                                                                                             81
                                                                     MORGAN STANLEY RESEARCH

                                                                     March, 2012
                                                                     EM Local Markets Guide




Turkey
Monetary Policy                                                       FX
The Central Bank of the Turkey’s (CBT) primary objective is           The Turkish lira (TRY) is a free floating currency having left
to achieve and maintain price stability. The bank follows an          behind a crawling peg system in 2001. The CBT conducts
inflation-targeting regime with a three-year horizon, and for         daily USD purchases or sales with both fixed and optional
the period 2012-14 the target has been set at 5% year-end             quantities, when they deem it necessary. They do so to
yoy CPI. The target has an uncertainty band of ±2%.                   mitigate extreme moves in the currency and offset the
The main decision-making body is the Monetary Policy                  exchange rate impact of speculative flows. They also
Committee (MPC), currently consisting of six members: the             conduct USD purchases in order to accumulate FX reserves.
Governor, three Deputy Governors, one member from the                 In addition, on rare occasions, the CBT directly intervenes in
CBT’s Board and one member appointed by the Governor. The             the FX market, and publishes the interventions.
Governor is appointed for a term of five years by the Council of      The lira is fully convertible, and the spot market is one of the
Ministers. The MPC meets once a month and Minutes of each             most liquid in EM, enjoying average daily volumes at around
meeting are released a week later. The CBT’s main publication         US$6bn. Spot is generally quoted against the USD, although
is the quarterly Inflation Report, which sets out the Bank’s          EUR crosses are regularly traded, and other crosses such as
inflation forecast over the next two years.                           with the ZAR or ILS are also traded. Typical transaction sizes
Exhibit 145                                                           are at around US$5m.
CBT Inflation and Output Gap Forecasts                                The TRY forward market offers tenors to 10y and is most liquid
                                                                      up to 2y. The average daily volumes reach US$2-2.5bn with
                                                                      trade sizes for forwards and swaps averaging at US$10m and
                                                                      US$200m, respectively. The TRY option market is also diverse
                                                                      and offers tenors that range to 10y.

                                                                      Exhibit 146
                                                                      USD Buying and Selling Auctions and USD/TRY
                                                                       2.0


                                                                       1.8


                                                                       1.6



Source: CBT                                                            1.4


The CBT’s main policy instrument is the one-week repo, which           1.2
is the official policy rate. The Bank also has an overnight
borrowing and lending facilities at rates below and above the          1.0
policy rate used to temporarily adjust monetary policy. For             Jan-06 Oct-06 Jul-07 Apr-08 Jan-09 Oct-09 Jul-10 Apr-11 Jan-12

example, in order to stem the depreciation of the Turkish lira, at            USD Buying Auction             USD Selling Auction            USD/TRY
the end of 2011 the CBT sharply raised the rate on the                Source: Undersecretariat of Treasury. As of end-2011. Total amount is TRY 518bn.

overnight lending facility and reduced the proportion of funding
provided at the one-week repo. The average funding rate went          Fixed Income
up even though the main policy rate was unchanged. The CBT            The Undersecretariat of Treasury is responsible for debt
also introduced a 1-month repo. Other tools actively used by          management and issuance for the country, with the CBT
the Bank to affect liquidity conditions include reserve               acting as its issuing agent. Lira-denominated debt accounts
requirement ratios on banks’ liabilities.                             for 70% of central government debt, and foreign currency


                                                                                                                                                         82
                                                                                                MORGAN STANLEY RESEARCH

                                                                                                March, 2012
                                                                                                EM Local Markets Guide




another 30%. Turkey doesn’t have any TBills outstanding,                                         Inflation-linked bonds were introduced in February 2007 and
while the average maturity of domestic debt is 2.6 years.                                        are typically issued in 5y and 10y tenors. Redemption and
                                                                                                 (semi-annual) coupons are indexed to the headline
Exhibit 147
                                                                                                 Consumer Price Index (CPI) with a three-month lag, as
Currency Composition of Central Government Debt
                                                                                                 published by the Turkish Statistical Institute (TurkStat).

                                                                                                 Floating rate notes (FRN) account for a large share of
                                                                                                 outstanding debt but are mainly old issues with 5y and 7y
              30%
                                                                                                 maturity that are now rarely tapped by the Treasury.
                                                                                                 Coupons are quarterly or semi-annual and linked to the
                                                                                                 weighted average yield of zero coupon auctions during either
                                                                        TRY                      the previous 91 or 182 days (depending on the coupon
                                                                        Foreign currency         frequency). Finally, the Treasury has four revenue-indexed
                                                                                                 bonds outstanding, with semi-annual coupons linked to
                                                                                                 revenues of four state-owned companies. Revenue bonds
                                                                                                 represent a very small fraction (<1%) of domestic debt.
                                                           70%


                                                                                                 Primary Auctions
                                                                                                 The Treasury announces a three-month ahead borrowing
Source: Undersecretariat of Treasury. As of end-2011. Total amount is TRY 518bn.                 strategy on the last business day of each month. The
There are four main types of bonds outstanding: fixed-rate,                                      document includes monthly targets for redemptions and
zero-coupon, CPI-linked and floating rate. Fixed-rate and                                        borrowing, and an issuance calendar with dates and specific
zero bonds are the most frequently issued and most liquid                                        bonds to be issued. Target sizes of individual auctions are
government securities. Zeros are typically short-dated, sub-                                     not announced prior to the auction date.
2y issues. The most actively traded zero is called the                                           The CBT is the Treasury’s fiscal agent and conducts
“Benchmark”. Fixed-rate bonds with semi-annual coupons                                           domestic borrowing auctions. Auctions are multiple-price
are issued with tenors from 2y to 10y. Their share of                                            format and open to all investors (either institutional or
outstanding debt continues to grow given the Treasury’s                                          individual), but a system of primary dealers (PDs, of which
strategy of increasing the average maturity of government                                        there are 12 as of the time of writing) also exists. Bidders are
debt.                                                                                            required to deposit collateral in an amount of 1% of their bids
Exhibit 148                                                                                      on the auction day, but PDs are exempt from this
Total Debt Composition                                                                           requirement.

                                                                                                 PDs and public agencies have the right to submit non-
                    17%                                                                          competitive bids prior to the announcement of auction
                                                                                                 results. PDs also have the right to purchase additional
                                                     30%
                                                                                                 amounts through option bids at the average price of the
                                                                                                 auction until 14:00pm on the issue date.
                                                                 Fixed-rate

                                                                 Floating rate (incl revenue)    Exhibit 149
                                                                 Zero coupon                     Auction Calendar and Format
      24%                                                        CPI
                                                                                                  Bills and Bonds           Dependent on monthly plan
                                                                                                  Non-competitive bids by   10:30am local time
                                                                                                  Competitive bids by       11:45am local time
                                                                                                  Results out by            1-2pm local time
                                          29%
                                                                                                  Format                    Multiple price
Source: Morgan Stanley, Bloomberg. As of March 2012. Total amount is TRY 383bn.
                                                                                                                            At average price of competitive auction, with a
                                                                                                  Non-competitive
                                                                                                                            further option to purchase more at average price
                                                                                                 Source: Morgan Stanley




                                                                                                                                                                       83
                                                                                                               MORGAN STANLEY RESEARCH

                                                                                                               March, 2012
                                                                                                               EM Local Markets Guide




PDs have an obligation to purchase at least 3% in each                                                          Interest Rate Derivatives
month and at least 5% in each three month period, on a net
                                                                                                                Unlike most markets in CEEMEA, the cross-currency swaps
basis, of either the securities issued, or programmed to be
                                                                                                                (ccy) are the most liquid instruments in Turkey. Ccy is traded
issued by the Treasury, depending on which is lower. Non-
                                                                                                                out to 20y, although it is generally liquid up to 10y. The
comp purchases are taken into account in the calculation of
                                                                                                                floating leg fixes against 3m USD Libor, and the convention
the obligation, while purchases through option bids are not.
                                                                                                                is Act/360.
In addition, purchased amounts are weighed differently
depending on the maturity of the bond.                                                                          Average daily market volume of the ccy market is US$100-
                                                                                                                150k in DV01, with average transaction size of US$5-10k in
Investors in Government Debt
                                                                                                                DV01 (bid/offer around 4-8bp). Settlement is T+2.
The Turkish domestic government debt market is dominated
                                                                                                                The single-currency interest rate swap (IRS) is far less liquid,
by local banks, which own 59% of the market, followed by
                                                                                                                although is improving. The floating leg is fixed against 3m
domestic corporates at 19%. Foreign investors’ share of the
                                                                                                                TRLIBOR, which is managed by the Banks Association of
market has also grown in recent years to 17%, while
                                                                                                                Turkey. There are currently 13 participating banks. The fixing
domestic institutional investors such as mutual funds are
                                                                                                                is published at 11:15am local time for tenors from overnight
only a small component of the market with a less than 4%
                                                                                                                to 1y, and calculated by taking the arithmetic average of all
share. Social security in Turkey is still largely dominated by
                                                                                                                contributions, after removing the highest three and the lowest
the state pay-as-you-go system. Private pension funds exist
                                                                                                                three quotations.
but participation is optional. As of end-2010, mutual funds
and pension funds had asset under management of TRY
                                                                                                                Useful Websites
30bn and TRY 12bn, respectively, about a third of which was
invested in government debt.                                                                                    Central Bank of Turkey
                                                                                                                www.tcmb.gov.tr
Exhibit 150
Ownership Structure of Domestic Debt (%)                                                                        Turkish Treasury
                                                                                                                www.treasury.gov.tr
  70

                                                                                                                Turkish Statistical Institute (TurkStat)
  60
                                                                                                                www.turkstat.gov.tr
  50
                                                                                                                Istanbul Stock Exchange (ISE)
  40                                                                                                            www.ise.org
  30                                                                                                            Capital Markets Board of Turkey
  20
                                                                                                                www.cmb.gov.tr

  10
                                                                                                                Banks Association of Turkey
                                                                                                                www.tbb.org.tr
   0
        2004         2005      2006            2007           2008          2009        2010        2011
                                                                                                                http://www.trlibor.org
          Banks (inc CBT)   Retail investors          Corporate investors     Mutual funds     Non-residents    Morgan Stanley Bloomberg page: MSEX
Source: Undersecretariat of Treasury, Morgan Stanley.




                                                                                                                                                                            84
                MORGAN STANLEY RESEARCH

                March, 2012
                EM Local Markets Guide




LatAm Markets




                                          85
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Argentina
Monetary Policy                                                     recently focused on fighting currency depreciation, as it is a
                                                                    net buyer of foreign exchange. Its daily purchases and policy
Unlike other central banks in the region, which target inflation,   are readily disclosed.
Argentina has a unique monetary policy, targeting M2 growth.
At the beginning of the year, the central bank of Argentina         Foreign investors gain exposure to ARS through the offshore
discloses its inflation target and the projected total change in    market in NDFs. There is a liquid curve for NDFs extending
the money stock; this is disclosed in the Monetary Program.         out to seven years. The fixings for NDFs are determined by
There is a monthly monetary report which looks at financial         the Trade Association for the Emerging Markets, EMTA, and
and monetary issues, called the Monetary Report. As of the          the central bank also releases a fixing for ARS.
January 2012 Monetary Report, the BCRA has a full year              Onshore forwards are also available; however, the curve is an
target M2 at 26.4%Y (+/- 4%) This is the target M2 range that       abbreviated version of the NDF market as it extends only to a
the central bank uses to conduct its monetary policy, which is      one year maturity.
significantly different than most other central banks globally.
                                                                    Market making in foreign exchange options is available both
The central bank carefully monitors the interbank interest rate     onshore and offshore.
market, and sets the reverse repo interest rate. The call loan
rate, an annualized rate of interest in the interbank market,       Fixed Income
generally tracks the rate set by the BCRA for the reverse repo
interest rate. There are call loan rates extending from             The fixed income market for Argentina domestic and external
overnight through 30 days out. The market for repo securities       bonds is arguably the most complex among its Latin America
is based on sterilization paper issued by the BCRA and              peers. There is an active market for both domestic and
performing (non-defaulted) Treasury Bonds, PGs and Bogars.          external paper; however, due to the 2001 debt default and the
While the bank targets its repo rate, the market rate for           2005 debt exchange, the bond market is crowded with various
borrowing and lending between private entities is BADLAR,           types of issues and structures. Active external bonds exist for
and this rate generally tracks the target repo rate.                not only USD but also for JPY and EUR.

                                                                    Exhibit 151
FX                                                                  Total Debt Stock Composition
The Argentine peso (ARS) is an actively managed floating
currency, where the currency is not convertible. Onshore spot
trading takes place in two markets: the MAE (Mercado Abierto
Electrónico) and the MEC (Mercado Electrónico de Cambios).                                                               38.7%
Companies in Argentina are required to convert foreign
exchange receipts from exports into local currency during a
specified timeframe. The currency has had episodes of rapid
sell-offs along with devaluations such as in the 2002 crisis,
                                                                     61.3%
and the BCRA is known to sell USD positions to support its
currency.                                                                                                                       Domestic
The peak hours of liquidity for the local currency are from                                                                     External
8:30am to 12:30pm EST, with daily average volume of
US$400-600 million. The central bank publishes a Weekly             Source: Morgan Stanley Research, Bloomberg, Ministerio de Economia y Produccion
Foreign Exchange Report, which details the BCRA’s net
purchases of foreign exchange, the benchmark exchange               The majority of domestic securities are Treasury bonds, which
rate, foreign exchange transactions excluding BCRA and              were issued ex post debt default in 2001. There is also
cereal & oil grain exporters’ foreign exchange transactions. In     Treasury debt linked to the 2005 debt exchange which
2010, the BCRA was a net daily purchaser of foreign                 includes Discount, Par and Quasi-Par series; alongside these
exchange; meanwhile, during the financial crisis in 2008, it        issues are the Bonar and Boden Series of bonds. Sovereign
was a net seller of foreign exchange. The central bank has          bonds are usually Euroclearable, whereas central bank debt

                                                                                                                                                      86
                                                                        MORGAN STANLEY RESEARCH

                                                                        March, 2012
                                                                        EM Local Markets Guide




can only be settled locally. Within the sovereign bond space,           Exhibit 153
inflation-linked bonds make up a significant portion of the debt        Sovereign Debt Issuance 2010 (in US$ thousand)
outstanding.
                                                                         16000
Both the BCRA and the Treasury issue CPI-linked bonds,                   14000
where the principal is protected from inflation. Argentine CPI
                                                                         12000
is subject to a one-month lag plus ten days. The INDEC                                     Domestic
(Instituto Nacional de Estadistica) releases the CPI index,              10000             External
where the last reading for December 2011 was 9.5%Y                         8000
inflation. Many of these inflation-linked bonds have GDP
                                                                           6000
warrants attached to them, which pay a coupon linked to the
performance of Argentina GDP, and are legacy assets of the                 4000
2005 debt restructuring.                                                   2000
Exhibit 152                                                                    0
Domestic Debt Instruments                                                       Jan-11 Mar-11 Mar-11 Jul-11 Aug-11 Aug-11 Oct-11 Dec-11
                                                                        Source: Morgan Stanley Research, Bloomberg
 Domestic Debt Instruments
 Name                     Type                Coupon       Maturities
                                                                        While Exhibit 153 shows a healthy issuance of securities in
 Boden                    Floating            Semiannual   2012-15
                                                                        foreign currency, high spreads on its debt, reflected by the
 Bogar                    Infflation-linked   Monthly      2018, 2020
                                                                        elevated CDS spreads, have deterred the government from
 ARS Discount             Infflation-linked   Semiannual   2033
                                                                        accessing external funding markets. Therefore, the
 Bonar                    Floating            Quarterly    2012-17
                                                                        government has relied on intra-public finance, which makes
 Nobac                    Floating/Fixed      Semiannual   2012-14
                                                                        use of direct loans with neighboring countries (e.g., Argentina
 ARS Quasi Par            Fixed               Semiannual   2045
                                                                        has direct issuance lines with Venezuela in place).
 ARS Par                  Step Coupon         Semiannual   2038
 Bocon                    Infflation-linked   Monthly      2012-18      Interest Rate Derivatives
Source: Morgan Stanley Research, Bloomberg
As Argentina has emerged from its series of debt crises, the            There are BADLAR interest rate futures where the underlying
Treasury has issued bonds in debt swaps with GDP warrants               asset is a fixed-term deposit contract, for ARS, where the
attached, and in 2005 these have been detached and now                  contract size is generally quoted as 100,000. These futures
trade separately in the market. There is an active market for           are quoted 18 months out regularly.
GDP warrants with fair liquidity.
                                                                        While the BCRA has its own reference repo rate, a market for
Primary Auctions                                                        Badlar has developed which is an interbank market and is
                                                                        quoted as the average rate between banks on wholesale
The BCRA auctions securities weekly on Tuesdays and its                 CDs. The country has regulations which would foster the
securities settle t+1. There are two types of securities, which         development of an interest rate swap market; however,
the central bank auctions: Lebacs (Letras del Banco Central)            liquidity remains low.
and Nobacs (Notas dl Banco Central). In 2011, Argentina has
issued paper in local currency, ARS (domestic) and in foreign           There are inflation-linked swaps based on the Coeficiente de
currencies: USD, EUR and JPY. Issuance has been fairly                  Estabilizacion de Referencia, CER, and this is calculated by
consistent throughout the year, with more frequent issues               the central bank according to the consumer price index. The
coming from the domestic sector in ARS, as there are                    index is intended to adjust deposits and loans in ARS. While
consistent auctions of these types of securities.                       the index is available as a reference to swap inflation between
                                                                        counterparties, liquidity is low for this product for now.




                                                                                                                                    87
                                                             MORGAN STANLEY RESEARCH

                                                             March, 2012
                                                             EM Local Markets Guide




The Local Investor Base
Unlike Peru or Colombia, where private pension funds are a
dominant local investor, in Argentina, pension funds are
nationalised and have been since 2008. Pension funds
(AFJPs) now complement the outstanding holdings by the
BCRA and other government agencies which hold Treasury
bonds. In place of pension fund demand, insurance
companies are now the primary investor in local markets,
where over 50% of insurance AUM is held in government
bonds.

Useful Websites
Banco Central de la Republica Argentina
www.bcra.gov.ar

Ministerio de Economia
www.mecon.gov.ar

EMTA web site
www.emta.org




                                                                                       88
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Brazil
Monetary Policy                                                     USD. The BCB releases a daily fixing rate for USD-linked
                                                                    instruments called the PTAX800, used for fixing of all non-
Brazil moved to an inflation-targeting regime in 1999, after
                                                                    deliverable instruments. The PTAX is a weighted average per
abandoning a crawling-peg system aimed at keeping the FX
                                                                    volume calculated by the BCB and is published daily.
rate within a predetermined band. The inflation target is set by
the so-called National Monetary Council (CMN), comprising           Exhibit 154
the central bank governor, the finance minister and the             BM&F Open Short USD (Long BRL) Positions
planning minister. Currently, the target is 4.50% +/- 2%.
                                                                      290,000

It is worth noting that although the central bank (Banco
                                                                      240,000
Central do Brasil – BCB) is not independent, as set forth by
the constitution, recent administrations have allowed the bank        190,000

to act with de facto independence, allowing the BCB to                140,000

become more effective in delivering monetary policy due to
                                                                       90,000
greater credibility. The central bank president was elevated to
ministerial status several years ago.                                  40,000


The Selic rate, set by the bank’s monetary policy committee           -10,000

(Copom), is the main policy tool. The effective Selic rate is set     -60,000

in the collateralized O/N market.
                                                                     -110,000

The relevant CPI headline index is the IPCA, which is
                                                                     -160,000
published monthly (in addition to bi-weekly IPCA-15 pre-                    Jan-08   Apr-08   Jul-08   Oct-08   Jan-09   Apr-09   Jul-09   Oct-09   Jan-10   Apr-10   Jul-10

releases).                                                          Source: BM&F
                                                                    Offshore BRL NDFs are fixed two days prior to settlement
Monetary policy meetings take place eight times per year on
                                                                    through the PTAX rate. Onshore forwards and futures are
predetermined dates.
                                                                    fixed one day before settlement.
FX
                                                                    The onshore market is quite liquid, with a daily turnover of
The Brazilian real (BRL) has operated within a dirty or             around US$16 billion. The offshore NDF market trades with
managed-float regime since 1999. Though BRL is not                  daily volumes of around US$3-4 billion, with the best liquidity
convertible, there have been some initial steps taken in the        in tenors of one year or less. The total volume adding non-
last few years to move towards full convertibility, including       deliverables and spot is approximately US$20 billion.
allowing exporters to have full discretion over export revenues
                                                                    The existence of a very strong and deep onshore market is
and allowing BRL to be used to settle transactions between
                                                                    fairly unique to Brazil. The local futures exchange provides
Brazilian and international financial institutions in specific
                                                                    transparency, tighter bid-asks and lower haircuts through its
cases. Nevertheless, the central bank still requires the
                                                                    daily-margin mechanism – a system deemed potentially crisis-
registration of all trades on its system (Sisbacen).
                                                                    proof, as several external and internal financial distress
Brazil has the largest international reserves in Latin America,     situations did not disrupt the functioning of the FX market,
with over US$350 billion as of January 2012.                        even during the most turbulent times.

The spot market trades OTC and is registered in Sisbacen            The onshore futures market opens at 9am and closes at 6pm
(Central Bank system). Spot settlement is usually T+2 (though       Sao Paulo time. The spot market opens at 9 am and closes at
in some instances T+1 / T+0 is available). Although the             5pm Sao Paulo time, with most liquidity concentrated around
volume in the spot market is increasing as the BCB pushed           10am-1:30pm/ 02:30-4:30pm.
for more transparency and convertibility, still the bulk of
                                                                    The FX option markets is quite liquid as well, with both
volumes are in non-deliverable format, in the form of local
                                                                    onshore listed options and offshore OTC options in the form
futures cash settled in BRL (listed at the BM&F futures
                                                                    of NDOs (non-deliverable options) The liquidity in options is
exchange) and offshore in the form of NDFs cash settled in
                                                                    more concentrated offshore though, given the greater


                                                                                                                                                                               89
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




flexibility with strikes and option types, while the onshore      Exhibit 156 highlights the different debt instruments currently
market is more limited to pre-defined vanilla strikes listed at   issued by the NT. The most widely traded instruments by
BM&F and mostly short-dated. The options market has deep          foreign participants are the fixed-rate NTN-F bonds and the
pricing from one month to one year, but also two years            inflation-linked NTN-B instruments. Government bonds have a
(which, although less liquid, also has liquidity).                daily liquidity of about US$6-7 billion. Though government
                                                                  securities are exempt from income and capital gains tax,
The actual liquidity in BRL resides on the floor of the BM&F
                                                                  foreigners need to pay the IOF tax on investments unwound
exchange, so all pricing given by dealers in spot or NDFs is
                                                                  within 30 days of inception. There is also a 6% upfront IOF tax
hedged via futures; therefore dealers carry a natural large
                                                                  for all fixed income investments by foreigners. To trade in the
basis risk position either between spot and futures or between
                                                                  local market, foreign investors need to open a special
NDFs and futures.
                                                                  investment account, commonly known as the 2689 account.
The BCB releases a daily fixing rate for USD-linked
                                                                  Exhibit 156
instruments called the PTAX. This fixing applies to bonds and
derivatives settled onshore and offshore. The PTAX is the         Domestic Debt Instruments
average of effective rates of transactions in the interbank FX
market. This average is then weighted by the volume of the        Name       Type                Coupon      Index          Maturities   Out.    % of total
corresponding transactions.                                       LFT        Floater             -           Selic(O/N)     2m-6yrs      516.8   28.3%
Offshore BRL NDFs are fixed two days prior to settlement
                                                                  LTN        Fixed Rate          Zero        -              3m-6yrs      288.6   26.8%
through the PTAX rate. Onshore forwards are fixed one day
before settlement.                                                NTN-F      Fixed Rate          Semi        -              3-10yrs      236.9   11.0%

The Brazilian FX market is officially open from 9:00am – 6:00     NTN-B      Inflation-linked    Semi        IPCA           3-40yrs      354.1   28.6%
pm in BMF and the spot market trades from 9:00 am to 5:00
                                                                  NTN-C      Inflation-linked    Semi        IGP-M          3- 20yrs     26.9    3.5%
pm local time.                                                    Source: Banco Central do Brasil, Tesouro Nacional as of Dec 2011

Fixed Income
The National Treasury (NT) issues all debt in the country. Its    LTN and NTN-F: LTN and NTN-F bonds are becoming the
strategy for several years has been to extend the duration of     most actively traded instruments from an investment point of
domestic debt while increasing the share of fixed-rate            view, even though they (individually) make up a relatively
instruments (away from floating-rate debt).                       small share of outstanding government securities. The
                                                                  government’s focus on issuing an increasing number of fixed-
Brazil uses a business/252 day-count convention for domestic      rate bonds is likely to make these instruments the most liquid
BRL-denominated assets, calculated as follows:                    going forward. Offshore investors tend to be most involved in
da = (1+r)1/252-1                                                 the back end of the curve, while locals tend to be most active
                                                                  in the shorter end
da = daily accrual, r = rate in %
                                                                  NTN-B: These bonds are linked to the IPCA inflation index.
Exhibit 155
                                                                  Though they are significantly less liquid than fixed-rate bonds,
Total Debt Stock Composition                                      NTN-Bs are often used by foreigners wishing to make longer-
                                                                  term bets on Brazilian inflation and rates. Liquidity in the NTN-
                                                                  Bs is concentrated mostly in the front end.




                                          External
                                          Domestic


Source: Banco Central do Brasil


                                                                                                                                                   90
                                                                                                 MORGAN STANLEY RESEARCH

                                                                                                 March, 2012
                                                                                                 EM Local Markets Guide




Exhibit 157                                                                                      The government continues to focus on issuing more fixed-rate
Foreign Holdings of Local Fixed Income as                                                        debt and is trying to push out duration as much as demand
Percentage of Total                                                                              will allow it. Thus far this year, the National Treasury has
 14%                                                                                             issued most heavily in the 2014 and 2021 NTN-F maturities.
 12%
                                                                                                 Interest Rate Derivatives
 10%
                                                                                                 The CDI rate is set in the uncollateralised interbank O/N
  8%
                                                                                                 interest rate used as the funding rate for pricing derivatives
  6%                                                                                             (similar to Libor in other markets). The CDI is closely linked to
  4%                                                                                             the Selic rate.

  2%                                                                                             The DI futures are traded in the BM&F exchange and are
                                                                                                 effectively fixed-maturity date OIS contracts with the
  0%
    Jan-        Aug-     Mar-      Oct-     May-       Dec-       Jul-     Feb-      Sep-
                                                                                                 underlying instrument being the CDI rate. The contracts trade
     07          07       08        08       09         09         10       11        11         on a yield basis and the cash flow of the contract resembles
Source: Tesouro Nacional                                                                         that of a zero-coupon bond.
                                                                                                 Liquidity is irregular across the curve, with the Jan ’13 and
Primary Auctions
                                                                                                 Jan ’14 being the most liquid contracts at present. Daily
The National Treasury establishes weekly auctions of its                                         turnover in the DI market is around US$20-25 billion.
securities. It releases a monthly issuance calendar with sizes                                   In addition to DI futures, there is a growing DI swaption
and maturities to be auctioned on a rolling basis. Authorised                                    market that is traded at the BM&F onshore (and OTC
bidders include brokers, pension funds, mutual funds and                                         offshore).
banks.
                                                                                                 The Local Investor Base
Exhibit 158
                                                                                                 Most Brazilian pensioners are covered by a mandatory pay-
NTN-F Issuance in 2011 (BRL)                                                                     as-you-go pension system that is managed publicly. The
 3,500                                                                                           public pension system generates hefty deficits (currently
 3,000                                                                                           around 13% of GDP) due to the low retirement age and past
 2,500                                                                                           concessions. Unlike other Latin American countries, the
 2,000                                                                                           private (optional) pension system is small and currently only
 1,500                                                                                           covers around 6 million pensioners. In total, the pension
 1,000                                                                                           system has a 15% position in government debt (as of Dec
  500                                                                                            2011). But the pension funds also invest in mutual funds that
   -                                                                                             have positions in government debt – thus bringing the total
         Jan-11 Feb-11 Mar-11 Apr-11 May-11 Jun-11   Jul-11 Aug-11 Sep-11 Oct-11 Nov-11 Dec-11

Source: Tesouro Nacional
                                                                                                 exposure to around 50%. Local hedge funds also play an
                                                                                                 important role in the derivatives market.
                                                                                                 Useful Websites
                                                                                                 Banco Central do Brasil
                                                                                                 www.bcb.gov.br
                                                                                                 Brazilian Ministry of Finance
                                                                                                 www.fazenda.gov.br
                                                                                                 Brazilian Future and Mercantile Exchange (BM&F)
                                                                                                 www.bmf.com.br
                                                                                                 Brazilian Association of Security Dealers (ANDIMA)
                                                                                                 www.andima.com.br
                                                                                                 Brazilian Security and Exchange Commission (CVM)
                                                                                                 www.cvm.gov.br


                                                                                                                                                                 91
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Chile
Monetary Policy                                                    circumstances, but it must publicly inform and justify its
                                                                   interventions.
As an autonomous institution, the primary objective of the
central bank is to maintain currency stability, and it generally   In November 2010, the central bank changed limits on foreign
pursues this policy through price stability. Second, the central   investment, raising the limits for pension funds to 80% from
bank ensures the normal functioning of domestic payments,          60%, which began in December 2010. This move started with
which has the added benefit of maintaining low and stable          three successive increases every three months beginning in
inflation. Third, it ensures the normal functioning of external    December. The increase in limits encourages local investors
payments and has the power to determine foreign exchange           to invest capital abroad and thereby weaken the Chilean
policy. The policy of achieving low and stable inflation is part   peso. To further weaken the peso, the Chilean central bank
of a broader framework of achieving a sustained path of            announced an intervention at the start of 2011, driving a rapid
growth and full employment.                                        8% depreciation in the currency.

The central bank targets inflation as its monetary approach        The central bank manages foreign exchange reserves in two
and, since 2007, the objective of the bank is to target annual     portfolios – the investment and liquidity portfolios; the former
inflation of the consumer price index (CPI) at 3%, with a range    holds foreign currency assets with an average duration of 13
of plus or minus 1%. Policy decisions are conducted at             months and the latter is used for shorter-term needs. Foreign
monthly meetings and these dates are announced six months          currency composition is generally in USD and EUR in a ratio
in advance. Statements of monetary policy are released in the      of 60% to 40%. The central bank currently holds US$39,210
official gazette (Diario Oficial). Minutes are released on the     million of foreign currency reserves.
fifth business day prior to the next Monetary Policy Meeting.      The FX spot market is liquid and settles T+1 with daily
The bank implements its monetary policy through a target           average turnover of approximately US$1.5-2 billion (plus NDF
nominal interbank interest rate, known as the monetary policy      volume of around $600m), traded on the screens and OTC.
rate (tasa de politica monetaria, TPM) and conducts open           Investors should also be mindful of the holiday schedule, as
market operations generally through purchasing and selling         there is credit risk, if onshore participants trade pesos for
promissory notes using a repurchase agreement.                     USD. If there is a holiday in the United States, and a
                                                                   participant delivers pesos, USD cannot be delivered from the
While the central bank target rate is the TPM, CLP TASA            United States, posing a 24-hour credit risk. The peak hours of
CAMARA is the rate that trades on the floating leg of swaps.       liquidity for the currency are from 7:00am-10:00am EST.
CAMARA is an interbank rate. Since 2001, the mean spread           Spreads in the spot market between bid and offer are usually
between the overnight CAMARA and the TPM rate is .01,              at .50 pips.
suggesting that CAMARA follows TPM very closely; however,
the two rates should not be thought of as interchangeable.         The forward market is non-deliverable for foreign investors;
CAMARA is akin to a LIBOR rate, but this rate closely tracks       thus, there is an active NDF market for currency forwards
the TPM.                                                           both onshore and offshore. The central bank releases the
                                                                   fixing for CLP, which is used for NDFs daily. Local investors
IPC (Inflación del IPC e IPCZ) is currently at 4.40%YoY.           could have forwards deliverable in local currency, CLP.
The current governor of the central bank is Rodrigo Vergara,       An FX options market is also available (although it is fairly
who has served since 2011.                                         illiquid), where trades are cash settled in USD; the most
                                                                   frequently traded types of options are vanilla calls and puts,
FX                                                                 but more exotic structures are available.
The Chilean peso (CLP) is a free-floating exchange rate            Cross-currency swaps are also available and are similar to
though it is not fully convertible. The exchange rate has been     interest rate swaps, but instead are funded at Libor instead of
freely floating since September 1999, when the last                the CAMARA rate.
restrictions on foreign exchange were removed; however,
these restrictions can be reinitiated. The central bank is
empowered to intervene in the market under exceptional


                                                                                                                                92
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




Fixed Income                                                      Exhibit 160
                                                                  Domestic Debt Securities (in CLP million)
The majority of Chile’s debt outstanding is issued in Chilean
                                                                                                                                                 % of
pesos, where currently domestic bonds represent over 90% of        Name         Type           Coupon             Maturities   Outstanding       total
total debt outstanding.
                                                                   PDBC         Fixed          Zero               2012-15      2,386,250         17.5%
The average maturity of Chile’s dollar debt is 4.7 years,
                                                                   BCP          Fixed          Semiannual         2012-13      1,622,650         12.8%
whereas the domestic debt has an average maturity of 7.4
years.                                                             BTP          Fixed          Semiannual         2014-20      524,975           7.3%
                                                                                Inflation-
Exhibit 159                                                        BCU          linked         Semiannual         2012-24      4,342,975         32.4%
                                                                                Inflation-
Total Debt Stock Composition                                       PRC          linked         Semiannual         2014-30      715,875           1.6%
                                                                                Inflation-
                                                                   BTU          linked         Semiannual         2015-40      1,718,100         28.6%
                                                                  Source. Morgan Stanley Research Bloomberg


                                                                  Primary Auctions
                                                                  Both the central bank and the Ministry of Finance can issue
                                                                  debt, and there is regular issuance in both nominal and
                                                                  inflation-linked securities. The Ministry of Finance makes its
                                                                  auction dates available on its website. The Ministry of
                                                                  Finance’s auction tables are quoted in UFs; for example, the
                                            External
                                                                  Ministry of Finance plans to issue 2,400 million of 10y BTU
                                            Domestic
                                                                  bonds on 11 April 2012. This equates to approximately CLP
                                                                  54,000 million in bonds and US$110 million. This type of
Source: Morgan Staley Research, Bloomberg
                                                                  issuance is replicated generally on a monthly basis and is
Only Chile Government International Bonds are issued in
                                                                  spread in equal issuance size, across the curve in 7y, 10y,
dollars. The rest of the bonds are issued in CLP, local
                                                                  20y and 30y space. The central bank also has a monthly
currency and represent various series respectively, viz.,
                                                                  liquidity schedule for its pace of issuance as well, and the
PDBC, BCP, BTP, BCU, PRC and BTUs, where each of these
                                                                  auction calendar from the central bank is generally released
bonds are issued by the central bank. BTUs are also issued
                                                                  at the beginning of the year in January, in close proximity to
by the Ministry of Finance.
                                                                  the previous year’s Annual Report.
Chile uses a unit of account called the Chilean Unidad de
                                                                  Exhibit 161
Fomento (UF), which is a real monetary unit based on the
                                                                  Sovereign Debt Issuance 2010 (in USD thousand)
inflation rate. Currently the exchange rate of UFCLP is 22468.
Sizes of bond issuance for example are listed in UFs. UF           16,000
bonds, issued by the Ministry of Finance, represent the bulk of    14,000                    Domestic     External
new issuance and daily trading in these issues amounts to          12,000
approximately US$250 million. There is also the Bolsa de           10,000
Comercio, the local exchange, which accounts for about a
                                                                    8,000
quarter of daily trading volumes.
                                                                    6,000

Despite developed financial and capital markets, Chile has          4,000
issued much less dollar debt than its peers, due to its strong      2,000
financial and capital markets.                                          0
                                                                         Jan-    Feb- Mar-   Apr- May-   Jun-   Jul-   Aug- Sep- Oct-    Nov- Dec-
                                                                          11      11   11     11   11     11     11     11   11   11      11   11
                                                                  Source: Morgan Staley Research, Bloomberg




                                                                                                                                                 93
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Chile continues to issue debt in accordance with its debt
distribution – favouring domestic issues. In summer 2010,
Chile issued a US$1 billion bond, Chile Pagare Reajustable
Cupon, but there have been no other dollar issuances since
then.

Interest Rate Derivatives
There are vanilla interest rate swaps, which swap fixed versus
floating cash flows in the same currency, CLP. The floating
leg is based on the CAMARA fixing. The convention is to
quote swaps on an ACT/360 basis, and swaps over one year
are generally paid semi-annually. There is an active and liquid
curve from 3m to 10y out.

Additionally, there is a UF versus CAMARA swap quoted from
the short end through the 20y sector. Up to the 18-month
tenor, the convention is as a zero-coupon swap and, from
2-year onwards, swaps are quoted on a semi-annual basis,
with both on an actual/360 basis.

Local Investor Base
Institutional investors, namely pension funds, hold a sizeable
amount of investable funds. Pension funds alone have over
US$100 billion in AUM. Pension funds are categorised based
on their limit to invest in equity securities as a percentage of
total assets under management. There are five different fund
categories (A, B, C, D and E), and each category assigns the
fund with a certain amount of foreign securities it can hold
along with other types of investments.

Useful Websites
Banco Central De Chile (Central Bank)
www.bcentral.cl/eng/index.htm

Gobierno de Chile (Ministry of Finance)
www.minhda.cl/english

Superintendencia de AFPs (pension fund regulator)
www.safp.cl




                                                                                             94
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Colombia
Monetary Policy                                                     central bank plans to continue to intervene for 3 months to
                                                                    fight currency appreciation.
According to the central bank, Banco de la República, the
primary objective of monetary policy is to maintain a low and       The Bolsa de Valores oversees daily trading activities in the
stable inflation rate, and to achieve a long-term GDP growth        foreign exchange markets. The peak hours of liquidity for the
trend. The bank’s Board of Directors sets quantitative inflation    currency are from 9:00am to 11:00 pm EST, with daily
targets, which are defined as the yearly variation of the           average volume ranging from US$1.5-2 billion.
consumer price index (CPI), released by DANE                        The forward market is non-deliverable; thus, an active NDF
(Departamento Administrativo Nacional de Estadística). The          market has developed and there are active quotes from the
bank currently has an inflation target range between 2% and         short-dated contracts through five years out. NDFs follow an
4%, with 3% as the specific target. Since 2007, the inflation       ACT/360-day count convention and generally settle T+2. The
target range has generally been 100bp wide.                         fixing for COP in the NDF market is released by the
There are seven voting members on the Board of Directors of         Superintendencia Bancaria de Colombia. This rate fixing is
the central bank; the current Governor is José Darió Uribe.         scheduled for release daily from 5:00pm to 6:00pm EST, and
The central bank releases an Inflation Report monthly, which        the fixing can be released at anytime during that hour.
describes its decisions on setting the base rate in Colombia.       Liquidity in FX options is best in vanilla structures and options
The base rate in Colombia is called the repo auction base rate      would be on a non-deliverable basis, cash settled at
or sometimes the base rate for liquidity-expansion auctions,        expiration.
where today the central bank holds this rate at 5% with
inflation at 3.90%. The central bank adjusts this repo rate in      Exhibit 162 shows positioning in the NDF market since 1997
order to conduct its monetary policy objectives.                    with the total outstanding contracts in millions for short
                                                                    USD/COP (long COP) positions. From the trend, it can
FX                                                                  generally be inferred that an active NDF market has
                                                                    developed along with an increased bias to be long the peso.
The Colombian peso (COP) is a managed floating currency,
and the currency is not convertible. The central bank               Exhibit 162

maintains a foreign exchange policy, which should be                Short USD/COP NDF Contracts (in millions)
consistent with its monetary policy, (i.e. if the central bank is    30000
raising its base rate it would not be simultaneously purchasing
                                                                     25000
foreign exchange reserves). Intervention is an acceptable and
possible method for the bank to pursue its inflation target, and     20000
if foreign exchange policy and monetary policy do not
coincide, the bank will usually carry out sterilisation. Since       15000
October 14, 2010, the central bank has linked itself to the
                                                                     10000
Colombian Foreign Currency Clearing House, which broadens
its methods available for purchasing dollars on the market.           5000

Colombia has maintained a floating exchange rate since                    0
September 1999; however, scheduled auctions and direct                    Jan-97 Jan-99      Jan-01 Jan-03 Jan-05   Jan-07 Jan-09   Jan-11
spot intervention do occur on a consistent basis in the FX spot     Source: Banco de la República
markets. The central bank used to maintain an intervention
rule that if COP moves more than 2% from its previous 20-day        Fixed Income
average rate, the central bank intervenes in the spot markets
to dampen volatility, but has since exited this program.            The majority of Colombia’s debt is in local currency, and while
                                                                    the country does have some international bonds on the
Since Feb. 3, 2012, the central bank has been back                  market, domestic issues seem to be most favourable for the
intervening in markets, purchasing US$20 million daily. The         government.



                                                                                                                                             95
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Exhibit 163                                                        Where UVR(t) = days after 15th day of the month (m), having
Total Debt Stock Composition                                       (t) between 1 and 31. (m-1) is the monthly inflation percent
                                                                   variation for month (m-1). The latest read on this index was
                                                                   202.096 as of Mar 26, 2010.

                                                                   While there is a repo market for domestic TES bonds, liquidity
                                                                   remains fairly low for such activity.

                                                                   Exhibit 164
                                                                   Domestic Debt Instruments (in COP million)
                                                                                                                                           % of
                                                                    Name             Type              Coupon   Maturities   Outstanding   total
                                                                    Local
                                                                    TES              Fixed             Annual   2012-26      94,029,867    87%
                                                                                     Inflation-
                                                Domestic            TES UVR          linked            Annual   2012-13      5,025,962     5%
                                                                                     Inflation-
                                                External            TES IPC          linked            Annual   2012-14      1,851,637     2%
                                                                    Global
Source: Morgan Staley Research, Bloomberg                           TES              Fixed             Annual   2013-41      6,972,163     6%
                                                                   Source. Morgan Staley Research, Bloomberg
The most prevalent type of bond is the domestic government
bond, TES B, which stands for Tesoreria B, and these bonds
                                                                   Primary Auctions
extend from maturities of one month to nearly 15 years out.
There are two types of Colombian TES bonds, nominal and            The central bank announces the scheduled auction sizes in
inflation-linked, with the former known as TES Tasa Fija and       coordination with the Ministry of Finance and generally
the latter as TES UVR. The government also has international       auctions are announced up to one day prior to it taking place.
bonds outstanding issued in COP and USD.                           TES Tasa Fija bonds, fixed-coupon bonds, generally come to
                                                                   market on Wednesdays and all auctions are based on a
Of particular note, the local TES market is virtually closed for
                                                                   Dutch system.
foreigners, as withholding taxes are equivalent to income
taxes and therefore investing and trading these securities, as     Exhibit 165
a foreigner, is economically unfeasible. The Colombia TES          Sovereign Debt Issuance 2011 (in US$ million)
bonds are similar in nature to the Peru Soberanos.                  4,500,000

Domestic fixed-coupon bonds settle T+3, and these bonds             4,000,000
pay coupons annually, while international bonds generally           3,500,000
                                                                                            External
have semi-annual coupon frequencies. Colombia also has an           3,000,000
active inflation-linked bond market.                                                        Domestic
                                                                    2,500,000

In linkers, Colombia issues two types of bonds: TES IPC and         2,000,000
UVR. TES IPC bonds pay a fixed real spread over the annual          1,500,000
CPI inflation. Coupons are paid annually and the inflation          1,000,000
index is adjusted monthly. While many of these bonds are              500,000
outstanding, this security has been phased out by the                            0
government.                                                                  03/03/11 04/28/11 06/23/11 08/18/11 09/20/11 11/10/11

TES UVR bonds are issued in real value units (UVR), where          Source. Morgan Staley Research, Bloomberg

the coupon is fixed and pays on an adjusting principal, which      Issuance has historically been dominated by domestic paper;
adjusts according to the UVR inflation-adjusted index. New         however, Colombia issues 2bn in USD denominated bonds in
bonds issued by the government follow this procedure and the       2011.The issuance of domestic paper is generally on a
calculation for the Units of Real Value Inflation-adjusted index   gradual basis and appetite remains strong for these types of
follows this formula.                                              issues.
UVR(t )  UVR(15, m) * (1  l (m  1))^ (t / Dm)


                                                                                                                                           96
                                                                 MORGAN STANLEY RESEARCH

                                                                 March, 2012
                                                                 EM Local Markets Guide




Interest Rate Derivatives                                        Useful Websites
The interest rate swap market is moderately developed (more      Ministry of Finance
than Peru, less than Brazil and Mexico). The central bank has    www.minhacienda.gov.co
a target rate, Indicador Bancario de Referencia, (IBR), and
                                                                 Banco de la Republica (Central Bank)
this is developing into a tradable curve. The regulator, which
                                                                 www.banrep.gov.co
oversees foreign exchange trading, Bolsa de Valores de
Colombia, has initiated a futures market for TES B domestic      National Statistics Department
bonds, where a synthetic 5-year bond anchors the futures         www.dane.gov.co
price.
                                                                 Bolsa de Valores de Colombia
                                                                 www.bvc.com.co
The Local Investor Base
                                                                 Superintendencia Financiera (Regulator)
Local pension funds and companies represent the largest
                                                                 www.superfinanciera.gov.co
investor class for domestic TES bonds. Local governments
and quasi-sovereigns such as Ecopetrol hold another large
share of the domestic debt. These entities alone hold
approximately 75% of TES bonds. Banks are another large
holder of fixed income securities and local hedge funds seem
to be a non-factor with respect to the local investor base.
Since domestic TES bonds are virtually unavailable to foreign
investors, most of the purchasers and holders of these
securities are from local investors, as they avoid the steep
withholding taxes imposed on foreigners.




                                                                                                           97
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Mexico
Monetary Policy                                                     forwards are fully-deliverable and trade as swaps. Beyond
                                                                    that, basis swaps are available with maturities up to 20 years.
Perhaps one of the most critical changes to take place in the
last few years helping to develop Mexico’s local markets was        Mexico is currently in process of building up reserves.
the adoption of a formal inflation-targeting mechanism in early     International reserves stood at US$148 billion as of March
2005. Banco de Mexico’s (Banxico) shift towards a more              2012.
transparent mechanism using the so-called tasa de fondeo or         The government announced an options-based mechanism
O/N rate as its principal monetary tool provided much-needed        designed to allow it to accumulate international reserves
visibility to a market that had outgrown the previous system        without systematically distorting the functioning of the
based on a less precise corto (effectively a signal-based           currency market. The mechanism was successfully used from
mechanism that aimed to absorb liquidity in the market).            1996-2001 and has been in use since March 2010 without
The O/N rate is published on a daily basis and is effectively a     impeding normal FX market operations. At the end of 2011,
weighted average of one-day repo transactions between               Banxico introduced a weak side intervention mechanism,
brokers and banks cleared through Indeval. In addition, there       where they will intervene if the currency weakens by more
is a 28-day TIIE benchmark rate that is actively used by            than 2% on any given day. This intervention is via three
commercial banks and is determined on a daily basis by              auctions during the trading day, and Banxico will sell up to
Banxico through a series of surveys among banks. This rate          USD400m dollars per day.
is comparable to Libor rates in other countries. The relevant
CPI headline index is the IPC, which is published monthly (in       Fixed Income
addition to bi-weekly pre-releases).                                Exhibit 166

Open market operations are carried out on a daily basis by          Total Debt Stock Composition
Banxico, namely via BondesD instruments. Though Banxico
officially targets a 3% (+/- 1%) inflation level at present, the
fact that it has failed to meet the mid-point of the target since
its inception could potentially result in an upward revision of
the target.

Monetary policy meetings take place on a monthly basis on
predetermined dates.                                                                                              Domestic
                                                                                                                  External
FX
The Mexican peso (MXN) has been a free-floating currency
without any restrictions on purchases or sales since 1994.
                                                                    Source: Banco de Mexico
MXN trades on a fully deliverable basis.                            The importance of local debt continues to increase in Mexico,
The highly liquid FX market in Mexico is one of the most            now making up more than 80% of the total debt stock in the
widely traded in the world (and the most liquid in Latin            country. Of that, federal government securities comprise
America), with daily turnover of MXN of close to US$10 billion.     about 63% of Mexico’s fixed income market, with the
                                                                    remainder made up by IPAB bonds, corporates and other
The fixing is set by Banxico on a daily basis via a series of       smaller issues.
surveys averaging at least four quotes from local banks.

Mexico is the only Latin American country operating a
deliverable forward market open to non-residents.

In addition to spot, FX forwards are quite liquid up to 2 years
(with the most liquid points between 3 and 12 months). FX



                                                                                                                                 98
                                                                                       MORGAN STANLEY RESEARCH

                                                                                       March, 2012
                                                                                       EM Local Markets Guide




 Exhibit 167                                                                           Primary Auctions
 Domestic Debt Instruments (in MXNbn)
                                                                                       The finance ministry establishes weekly auctions of its
                                                                                       securities. It releases quarterly outlooks and issuance
Name        Type      Coupon          Maturities              Outstanding    % total
                                                                                       calendars with maturities and sizes. Authorised bidders
Cetes       Fixed     Zero-coupon     28, 91, 182, 364d       705,609.12     22.16%    include brokers, pension funds, mutual funds and banks.

Bonos       Fixed     Semiannual      3, 5, 10, 20, 30yr      1,638,841.94 51.47%      As the local market stabilized post financial crisis, the
                                                                                       government has returned to issuance patterns seen pre
BondesD     Float     28-day          3, 5yr                  701,716.91     22.04%    financial crisis. The government has issued greater duration,
                                                                                       bringing the average duration of Mexican debt to 6.7 years.
Udibonos Linker       Semiannual      3, 10, 20, 30yr         138,160.94     4.34%
 Source: Secretaria de Haciend EC389489 Corpa y Credito Publico
                                                                                       For the first time in 2010, in addition to the predetermined
 Bonos: These are the most actively traded instruments from                            auctions, the government decided to focus on issuing specific
 an investment point of view, as they make up the largest                              maturities via a syndicated placement system. This is designed
 share of federal government securities. Further, Bonos are                            to allow the government to create much more liquid points on
 Euroclearable and are largely tax-empted for foreigners by the                        the curve by issuing significantly larger amounts in a one-off
 finance ministry. Average daily turnover tends to be around                           fashion (instead of having to wait for an entire quarter’s-worth of
 US$1.5 billion.                                                                       auctions to reach the same amount of outstanding issue (see
                                                                                       Exhibit 169).
 Udibonos: These bonds are linked to inflation and are
 denominated in inflation-indexed units called unidad de                               Exhibit 169
 inversion (UDI). The UDI is a non-monetary unit used to                               Government Bono Issuance by Quarter (MXNbn)
 translate all aspects of Udibonos into MXN for payment. The                            30000
 UDI is calculated using the biweekly CPI releases by Banxico.
                                                                                                                                                                Q1 2011
 Though they are less liquid than Bonos, Udibonos are                                   25000
                                                                                                                                                                Q2 2011
 important as they serve as real-rate benchmarks for corporate
                                                                                        20000                                                                   Q3 2011
 issuers. These bonds tend to be buy-and-hold instruments
                                                                                                                                                                Q4 2011
 mostly owned by local pension funds in Mexico.                                         15000
                                                                                                                                                                Q1 2012
 The development of the local market has continued to attract
                                                                                        10000
 interest from foreign investors. We expect foreign holdings of
 Bonos to increase in 2012 from last year’s levels. Though                               5000
 there are no entry/exit taxes/restrictions for foreigners, a local
                                                                                             0
 custodian is required for government securities.
                                                                                                     3 year          5 year         10 year        20 year      30 year
 Exhibit 168                                                                           Note: Spikes in 5Y and 10Y issuance due to use of syndicated placement
 Foreign Holdings of Local Fixed Income (% of Total)                                   Source: Banco de Mexico


  50%
                                                                                       Interest Rate Derivatives
  40%                                                                                  TIIE: The interest rate swap market is one of the deepest and
                                                                                       most extended in Latin America. TIIE swaps are now widely
  30%
                                                                                       traded in Mexico and offshore. The 28-day TIIE rate (similar to
                                                                                       Libor in other markets) functions as the floating benchmark for
  20%
                                                                                       TIIE swaps.
  10%
                                                                                       The curve is well developed and ranges from 3 months (3x1
                                                                                       or three periods of 28 days) to 30 years (390x1). However,
   0%
    Jan-05     Jan-06      Jan-07   Jan-08     Jan-09      Jan-10   Jan-11   Jan-12    the curve is most liquid in 2s, 5s and 10y. The TIIE market
 Source: Banco de Mexico
                                                                                       allows for investors to sell rates, a prospect that is more
                                                                                       difficult through Bonos, given the underdevelopment of the
                                                                                       local repo market.


                                                                                                                                                                          99
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




In addition to the outright TIIE market, there is a growing TIIE   Useful Websites
swaption market, with expirations ranging from 1-5 years.
                                                                   Banco de Mexico
                                                                   www.banxico.gob.mx
The Local Investor Base
                                                                   Mexican Ministry of Finance
Local pension funds and mutual funds make up the lion’s
                                                                   www.shcp.gob.mx
share of the local investor base. Pension funds known as
Afores currently have over US$120 billion in assets under          Pension Funds Regulatory Commission (CONSAR)
management, while mutual funds manage around US$78                 www.consar.gob.mx
billion. The growth of this industry over the last 10 years is
                                                                   National Commission for Securities and Banking (CNBV)
due primarily to the creation of the private pension fund
                                                                   www.cnbv.gob.mx
system in Mexico since the late 1990s. Growth continues to
take place at a very rapid pace, with Afores exhibiting            Mexican Derivatives Exchange Market
annualised rates of growth of assets under management of           www.mexder.com
over 20% for the last few years. We expect this rate of growth
to continue.

In March 2008, the pension fund regulator, Consar, created a
new breakdown allowing for five different types of funds with
different risk profiles.

Inclusion into Global Indexes
Mexico’s inclusion into Citigroup’s World Government Bond
Index (WGBI) in October 2010 represented an important
structural improvement that is likely to deepen the local
market going forward.




                                                                                                                           100
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Peru
Monetary Policy                                                     Offshore investors gain access to PEN forwards through
                                                                    NDFs as the currency is not deliverable in forwards space.
The objective of the BCRP is to preserve monetary stability,        Liquidity is highest for those maturities in the front end of the
and it does this through inflation-targeting. Currently, the        curve, and NDFs are quoted through five years out. For
inflation target is set at 2% with a tolerance of plus or minus     NDFs, the Superintendencia de Banca y Seguros releases a
1%. This inflation target has been in place since 2007, and         daily fixing at 1:35pm EST.
the inflation rate is calculated by the National Statistics and
Information Institute, where the inflation rate is the observed     In options space, as with most of its peers, vanilla structures
rate in the Consumer Price Index for Metropolitan Lima. Until       are available on a non-deliverable basis, but liquidity is fairly
2006, this inflation target had been set at 2.5%. The last          low.
reading of the CPI index in December 2011 was at 4.23%Y.            Fixed Income
The central bank conducts its monetary policy, i.e., targeting      Reflecting the high degree of dollarisation in its economy,
the inflation rate, through adjusting its reference interest rate   Peru has substantially more dollar debt than its peers, other
or interbank rate. The BCRP publishes a schedule at the             than Argentina.
beginning of the year, which delineates when central bank
meetings will occur, and Inflation Reports are released             Exhibit 170
quarterly. In these meetings, the BCRP gives the reasoning          Total Debt Stock Composition
for its monetary policy and its outlook going forward.
The current Governor of the Board of Directors of the Central
Reserve Bank of Peru is Julio Velarde. In order to meet its
monetary policy objectives, the central bank has a target
reference rate for the interbank lending market. The BCRP
uses open market operations to influence the interbank
lending rate. To lower the interbank rate, the BCRP
purchases BCRP Certificates of Deposits or Treasury Bonds
under repurchase agreements, and to elevate the interbank                                                          External
rate, the BCRP generally issues central bank securities to                                                         Domestic
withdraw liquid funds from the interbank market.
                                                                    Source: Morgan Stanley Research, Bloomberg
Monetary policy meetings take place on a monthly basis on
                                                                    Domestic bonds in local currency are known as Soberanos,
predetermined dates. The current overnight interbank rate in
                                                                    and these have maturities extending from the front end
Peruvian new sols is 4.25%.
                                                                    through to the 35-year sector, where Peru currently has bonds
FX                                                                  maturing as late as 08/12/2046. Soberanos can be
                                                                    characterised as either Tasa Fija or VAC bonds, or nominal
The Peruvian new sol (PEN) is an actively managed free-             and inflation-linked, respectively.
floating currency and is not convertible, while the BCRP does
intervene in the spot market to manage exchange rate                Soberanos throughout the curve can be converted into GDNs,
volatility. While Peru has its own independent exchange rate,       can be settled in USD and are Euroclearable. This structure in
the US dollar is used frequently in transactions in Peru and        the fixed income market allows foreign investors to buy and
the two currencies coexist in the country. Dollarisation has        hold local bonds and keep their holdings in Euroclearable
decreased steadily over the last five years due to stable           accounts as opposed to local custody accounts, as would be
inflation dynamics and lower volatility in the exchange rate.       the case in purchasing bonds in other Latin American
The BCRP can intervene directly in the spot market but also         countries.
issues dollar-denominated CDs at times.
                                                                    VAC bonds have also been phased out in Peru and are no
The peak hours of liquidity for the local currency are from         longer issued. Though these bonds could technically be
9:00am to 1:00 pm EST, with daily average volume ranging            traded, there are very few sellers in the market, and this
from US$800 million to US$1 billion (400-600mn in the spot          would be very difficult.
market, 200-300 in NDF).


                                                                                                                                  101
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




Nominal domestic bonds generally pay on a semi-annual                                 Government bonds are issued using Dutch auctions, and
basis, and represent the bulk of debt outstanding. Peru                               bonds are issued monthly. Two days prior to the auction, the
government bonds have fixed coupons ranging from 3.83-                                Ministry of Finance announces the size of the issuance, and
12.25%.                                                                               bonds settle on a T+1 basis.
Compared to other countries in Latin America such as Chile                            Interest Rate Derivatives
and Colombia, Peru has a good amount of international bonds
both issued in USD and PEN, but with the majority of these                            There are interbank market rates extending from overnight to
issues in USD. Coupons on external dollar paper range from                            the 1-year sector for both USD and PEN, and there are also
6.55-9.87%.                                                                           LIMABOR rates for the same maturity types. Interbank rates
                                                                                      are the rates at which borrowers and lenders agree to borrow
Exhibit 171
                                                                                      and lend money and LIMABOR interest rates are based on a
Domestic Debt Instruments (in PEN million)                                            survey of banks; however, there is no true interest rate
                                                                              % of    derivatives market for Peru local rates.
Name                   Type          Coupon           Maturities     Out.     total
Soberanos- tasa
fija                   Fixed         Semiannual       2012-46        21,761   86%
                                                                                      The Local Investor Base
                       Inflation-
Soberanos- VAC         linked        Semiannual       2012-50        1,256    8%      Private pension funds and foreigners are the dominant
                       Zero                                                           holders of domestic government bonds, and they hold a
BCRP CDs               Coupon        Semiannual       2012-13        11,990   6%
Source: Morgan Staley Research, Bloomberg                                             combined amount of over 80% of domestic bonds
Inflation in Peru is kept in check both through inflation-                            outstanding. Pension funds concentrate their holdings in
targeting by the central bank and through linker issuance.                            longer-duration paper and inflation-linked bonds. Five pension
There are currently five outstanding inflation-linked bonds                           funds alone in Peru hold over 50% of Soberanos outstanding,
issued by the government, all denominated in PEN, with                                and they are called AFPs. These AFPs have adopted an
maturities extending from 2013 through 2050 and real                                  investment management approach described as a Multifund
coupons of 3.83-7.4%.                                                                 Management System, which is an investment approach that
These bonds are called VAC bonds, pay semi-annual                                     allocates capital according to specified risk-tolerance
coupons, and are linked to the monthly CPI inflation rate. The                        measures. There are three levels of risk, with level three
VAC, Valor Adquisitivo Constante, is subject to a one-month                           being the highest-risk investment fund. In addition to
lag.                                                                                  sophisticated pension investors, insurance companies also
                                                                                      participate in the market, but these institutions hold less than
Primary Auctions                                                                      5% of bonds outstanding.
Issuance in 2011 has been almost entirely in Peruvian new
sols, and issuance has been fairly light with slightly over                           Useful Websites
US$800 million in debt issued in 2011, relative to other LatAm
                                                                                      Ministry of Finance
countries.
                                                                                      wwww.mef.gob.pe
Exhibit 172
                                                                                      Banco Central de la Republica de Peru (Central Bank)
Sovereign Debt Issuance 2011 (PENmn)
                                                                                      www.bcrp.gob.pe
 14,000,000

 12,000,000                                                                           National Statistics Department
                                                                                      www.inei.gob.pe
 10,000,000       DOMESTIC

  8,000,000
                  EXTERNAL
                                                                                      Bolsa de Valores de Lima
                                                                                      www.bvl.com.pe
  6,000,000

  4,000,000                                                                           Superintendencia de Banca, Seguros y AFP (Banking,
                                                                                      insurance, and pension system regulator)
  2,000,000
                                                                                      www.sbs.gob.pe
          0
        1/24/2011 05/12/11 6/30/2011 8/24/2011 9/21/2011 11/14/11 12/07/11
Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                 102
                      MORGAN STANLEY RESEARCH

                      March, 2012
                      EM Local Markets Guide




Product and Market Focus




                                                103
                                                                                                MORGAN STANLEY RESEARCH

                                                                                                March, 2012
                                                                                                EM Local Markets Guide




EM as an Asset Class since 1994
Juha Seppala
                                rd
First Published Jun 23 , 2011: EM as an Asset Class since 1994

In a previous report, see EM Profile: The Case for EM Local                                      We also display two possible candidates for the optimal
Markets (April 4, 2011), we analyzed the risk-return                                             portfolio. The reason for this is that we didn’t find a typically
characteristics of EM local bonds compared to several                                            used criterion, but chose the portfolio with the highest
different asset classes for dollar-based investors in the                                        Sharpe ratio as a compelling method in our case. As it turns
Mean-Variance setting using weekly 2002-10 data. We found                                        out, the maximum Sharpe ratio criterion would have been
the optimal allocation to EM, in particular to EM local bonds,                                   even less risk-averse than portfolio A. As a matter of fact, it
FX-hedged, to be 75%. We chose to concentrate on 2002-10                                         would have been very close to the minimum variance
in our study as those are the years for which we have                                            portfolio. Since we believe our investors are more risk-
available data for EM local bond returns. Those are also the                                     tolerant than that, we chose to use as well an alternative
                                                                                                                                                                 18
years when EM stabilized as an asset class.                                                      portfolio selection criterion, Telser’s Safety First criterion.
One can argue that the returns during the last 10 years have                                     The optimal portfolios have 24% and 42% allocated to EM
been ‘too good to be true’. That is, the stabilization referred                                  according to the Sharpe and Telser criteria, respectively, the
to in the previous paragraph was a once-in-a-lifetime                                            rest being allocated to US bonds, and dollar index for low-
opportunity which has already been priced in. For this                                           risk portfolios and gold for high-risk portfolios.
reason, we have now extended the dataset to include years
                                                                                                 Exhibit 173
1994-2001. The 1990s were not good for EM. For example,
                                                                                                 The Efficient Portfolio Frontier, 1994-2010
the MSCI EM equity index was 541 at the beginning of 1994
and 489 at the end of 1999.                                                                       Expected Return
                                                                                                  12%
                                                                                                  11%                                                                   Efficient Set
The problem with the new dataset is that we do not have                                                                                            B
                                                                                                  10%
                                                                                                                                                                        Preferred Portfolios
data on EM local bonds for those years. However, from                                                 9%
previous studies – see EM FX in 2011 – A                                                              8%                                                                Equally Weighted
                                                                                                                    A                                                   Portfolio
Contemporaneous Approach (January 13, 2011) and EM                                                    7%
                                                                                                                                                                        Telser Portfolio
                                                                                                      6%
Quantitative Strategy Update: What if Treasuries Sell Off?                                                                                                C
                                                                                                      5%                                                                Sharpe Portfolio
Part Two (May 12, 2011) – we know that, for the most part,                                            4%
EM assets are highly correlated and one can approximate                                               3%
the returns of different assets by looking at the MSCI EM                                                  1%        3%          5%          7%
                                                                                                                                                   Risk
                                                                                                                                                          9%        11%          13%           15%

equity index and EMBI bond index, in particular, and we can
                                                                                                 Portfolio                                          A          Sharpe      Telser            B
construct a proxy for the purpose of this analysis.                                              Return                                           6.76%        4.77%       7.31%           9.44%
                                                                                                 Risk                                             3.39%        2.05%       3.96%           8.42%
EM in Global Portfolio 1994-2010                                                                 Sharpe Ratio                                      2.00         2.33        1.85            1.12

We study the risk-reward relationship using the Mean-                                            Asset class weights
Variance Optimisation framework and compare the results to                                       EM Equity                                          -        -                -              -
                                                                                                 EM Sovereign Credit                               1.00% -                   6.49%         47.29%
a ‘benchmark portfolio’, labeled C in Exhibit 173, where each                                    EM Money Market                                  34.64%   24.40%           35.91%         40.51%
                                              17
asset class gets an equal 9.09% allocation. Exhibit 173                                          S&P 500                                           0.85%    0.28%            0.96%          0.13%
                                                                                                 MSCI World Equities                                -        -                -              -
shows how, by moving from an equally weighted portfolio C                                        Commodities                                        -       0.18%             -              -
to portfolio A, one can reduce the portfolio risk from 8.4% to                                   Real Estate                                        -        -                -              -
                                                                                                 Gold                                              2.20%    2.44%            3.06%         12.07%
3.4% without giving up any expected return. Similarly, by                                        JP Morgan GBI                                      -      27.29%             -              -
moving from an equally weighted portfolio C to portfolio B,                                      JP Morgan US Agg Bond                            57.77%   16.52%           53.58%           -
one can increase the expected return from 6.8% to 9.4%                                           Dollar Index                                      3.54%   28.88%             -              -
                                                                                                 Source: Morgan Stanley Research, Bloomberg.
without increasing the portfolio risk.


17                                                                                               18
     Given the history of weekly returns since January 1994, we estimate the variance-                Telser’s Safety First Criterion maximizes the expected return subject to the constraint that
     covariance matrix for the 11 different asset classes in our sample. We then compute the          probability of return being less than a predetermined level of disaster return -7.5% is less
     set of efficient portfolios – that is, the set of portfolios which cannot be improved on         than 99.99%. For more discussion, please see J. Seppala, “The Diversification of
     without either decreasing the expected return (given by average historical return) or            Currency Loans: A Comparison between Mean-Variance and Safety First Criterions”,
     increasing the risk (given by the standard deviation of the associated portfolio).               European Journal of Operational Research, 1994.


                                                                                                                                                                                               104
                                                                         MORGAN STANLEY RESEARCH

                                                                         March, 2012
                                                                         EM Local Markets Guide




Exhibit 174                                                               gold, which has almost twice as high expected return than
Risk and Return by Asset Class, 1994-2010                                 the second performing asset class, EM equity. The other
                                              Expected Return     Risk    assets used in the optimal allocation are US bonds and
EM Equity                                          6.94%        22.95%    dollar index.
EM Sovereign Credit                               10.75%        13.04%
EM Money Market                                    8.15%         6.72%    We also wanted to see what would happen if the ‘the storm
S&P 500                                            9.11%        18.03%    of the century’ 2008-09 crisis was removed from the data.
MSCI World Equities                                5.73%        16.90%    For that purpose, in Exhibit 181 and Exhibit 182 we present
Commodities                                        7.00%        22.55%
Real Estate                                        5.96%        23.13%
                                                                          the results for the 1994-2007 subsample. The optimal EM
Gold                                               8.67%        16.36%    allocation is now much larger than in the previous cases,
JP Morgan GBI                                      6.47%         5.93%    28% using the Sharpe criterion and 55% using the Telser
JP Morgan US Agg Bond                              6.22%         4.15%    criterion. The assets in the optimal portfolios are S&P 500,
Dollar Index                                      -0.91%         8.13%
Source: Morgan Stanley Research, Bloomberg.
                                                                          commodities, bonds and dollar index.

Exhibit 174 displays the average returns and their volatilities           Finally, we wanted to assess how stable have the EM
over our sample period. EM sovereign credit has the highest               correlations been in our sample. In Exhibit 183, Exhibit 184
expected return while US bond index has the lowest risk.                  and Exhibit 185 we show the correlations between different
While gold has lower expected return than S&P 500, its risk               asset classes and EM equity, sovereign credit, and money
is also lower and it is much less correlated with other assets.           market, respectively. The correlations are remarkably stable.
Similarly, dollar allocation despite its negative expected                In particular, EM equity is highly correlated with other EM
return is due to the fact that it is the only asset class which is        assets, S&P 500 and world equities. It is positively correlated
significantly negatively correlated with other assets.                    but less so with commodities, real estate and gold, and it is
                                                                          basically not correlated with bonds and dollar index.
Sensitivity Analysis                                                      The correlations for EM sovereign credit can be divided into
                                                                          two groups: higher correlation group consisting of other EM
We also performed sensitivity analysis to these results using             assets, equities and real estate, and lower correlation group
subsamples of our dataset. The results are presented in                   made up of commodities, gold, bonds and dollar index. EM
Appendix 1 below. We first move to the 1990s. Exhibit 175                 money market, on the other hand, has three different
and Exhibit 176 present the efficient portfolio frontier,                 categories: high correlation (EM and equities), low
associated asset class weights, and risk-return                           correlation (commodities, real estate, gold and world bonds),
characteristics using 1994-1999 data. The optimal portfolio               and negative correlation (US bonds and dollar index).
has 18% or 28% allocated to EM using Sharpe and Telser
criteria, respectively. The best performing asset during this             Conclusion
period was S&P 500. Hence, not surprisingly, the allocation               Using weekly data from January 1994 to December 2010 on
varies between bonds and S&P 500 together with EM and                     11 different asset classes, we find that the optimal allocation
some dollar exposure for diversification purposes.                        on EM assets should be between 20% and 40%, depending
Results for 2000-05 sub-period are presented in Exhibit 177               on the risk tolerance level. This conclusion holds even when
and Exhibit 178. EM allocation in the optimal portfolio is                we analyze the data using different subsamples. However, in
either 33% or 34%, depending on portfolio selection criterion             each sub-period we used, the best performing asset class
used. The other instruments in the optimal portfolios are the             changed. In the full sample, the best performer was EM
best performing asset class, commodities, real estate, US                 sovereign credit, in 1994-99 it was S&P 500, in 2000-05
bonds and dollar index. In the 2006-10 sub-period,                        commodities, and in 2006-10 gold. The estimated EM
summarized in Exhibit 179 and Exhibit 180, the optimal EM                 correlations turn out to be remarkably stable.
allocation is either 25% or 17%. The best performing asset is




                                                                                                                                     105
                                                                                        MORGAN STANLEY RESEARCH

                                                                                        March, 2012
                                                                                        EM Local Markets Guide




Appendix 1: Subsample Results
I. 1994-99                                                                               II. 2000-05
Exhibit 175                                                                              Exhibit 177
The Efficient Portfolio Frontier, 1994-99                                                The Efficient Portfolio Frontier, 2000-05
 Expected Return                                                                          Expected Return
 28%                                                           Efficient Set              19%                                                      Efficient Set
                                                                                          17%
 23%                                                           Preferred Portfolios                                                                Preferred Portfolios
                                                                                          15%
                                                                                                            B
                                                               Equally Weighted           13%                                                      Equally Weighted
 18%
                                                               Portfolio                                                                           Portfolio
                               B                                                          11%
                                                               Telser Portfolio                                                                    Telser Portfolio
 13%                                                                                       9%
                                                               Sharpe Portfolio
                                                                                                     A                                             Sharpe Portfolio
            A                                                                              7%
  8%
                                                                                           5%

  3%                                                                                       3%
       1%          3%    5%         7%          9%       11%       13%            15%           1%          6%            11%            16%           21%                26%
                                         Risk                                                                                     Risk

Portfolio                            A          Sharpe     Telser           B            Portfolio                            A          Sharpe    Telser             B
Return                             6.43%        5.69%      10.13%        12.53%          Return                             7.55%        6.29%     9.93%           11.74%
Risk                               2.63%        2.30%      4.72%         6.30%           Risk                               2.97%        2.31%     4.64%           6.31%
Sharpe Ratio                        2.45         2.47       2.15          1.99           Sharpe Ratio                        2.54         2.72      2.14            1.86

Asset class weights                                                                      Asset class weights
EM Equity                             -            -          -              -           EM Equity                             -            -         -                 -
EM Sovereign Credit                   -            -          -              -           EM Sovereign Credit                  4.05%         -       19.14%            37.02%
EM Money Market                     19.20%       18.11%     28.01%         27.05%        EM Money Market                     28.45%       32.76%    14.99%              -
S&P 500                              9.43%        6.07%     24.06%         38.02%        S&P 500                               -            -         -                 -
MSCI World Equities                   -            -          -              -           MSCI World Equities                   -            -         -                 -
Commodities                          1.46%        1.35%      1.88%          1.62%        Commodities                          4.65%        2.54%     8.90%            13.58%
Real Estate                           -            -          -              -           Real Estate                          8.17%        4.47%    14.93%            21.20%
Gold                                  -            -          -              -           Gold                                 1.59%        2.31%     2.01%             5.56%
JP Morgan GBI                       43.85%       44.92%     24.04%         29.36%        JP Morgan GBI                         -            -         -                 -
JP Morgan US Agg Bond                 -            -        19.66%          3.95%        JP Morgan US Agg Bond               42.69%       39.10%    40.03%            22.64%
Dollar Index                        26.07%       29.55%    2.35%             -           Dollar Index                        10.41%       18.81%      -                 -
Source: Morgan Stanley Research, Bloomberg                                               Source: Morgan Stanley Research, Bloomberg


Exhibit 176                                                                              Exhibit 178
Risk and Return by Asset Class, 1994-99                                                  Risk and Return by Asset Class, 2000-05
                                      Expected Return                Risk                                                       Expected Return                 Risk
EM Equity                                  0.43%                   20.09%                EM Equity                                  7.90%                     19.30%
EM Sovereign Credit                       10.43%                   16.89%                EM Sovereign Credit                        12.36%                     9.27%
EM Money Market                            8.05%                    6.50%                EM Money Market                            7.93%                      5.12%
S&P 500                                   22.14%                   14.31%                S&P 500                                    0.73%                     17.62%
MSCI World Equities                       15.12%                   12.20%                MSCI World Equities                        -0.31%                    15.37%
Commodities                                5.81%                   15.44%                Commodities                                17.76%                    23.21%
Real Estate                                0.95%                   11.40%                Real Estate                                12.16%                    13.89%
Gold                                      -4.36%                   12.02%                Gold                                       11.20%                    14.68%
JP Morgan GBI                              5.53%                    5.38%                JP Morgan GBI                              6.76%                      6.17%
JP Morgan US Agg Bond                      5.50%                    4.45%                JP Morgan US Agg Bond                      6.87%                      4.03%
Dollar Index                               1.10%                    7.22%                Dollar Index                               -1.32%                     8.28%
Source: Morgan Stanley Research, Bloomberg                                               Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                          106
                                                                                    MORGAN STANLEY RESEARCH

                                                                                    March, 2012
                                                                                    EM Local Markets Guide




III. 2006-10                                                                         IV. 1994-2007
Exhibit 179                                                                          Exhibit 181
The Efficient Portfolio Frontier, 2006-10                                            The Efficient Portfolio Frontier, 1994-2007
 Expected Return                                                                      Expected Return
 23%                                                                                  12%                                                       Efficient Set
                                                           Efficient Set
 21%                                                                                  11%                         B
                                                                                                                                                Preferred Portfolios
 19%                                                       Preferred Portfolios       10%
 17%                                B
                                                                                       9%                                                       Equally Weighted
 15%                                                       Equally Weighted
                                                                                                                                                Portfolio
                                                           Portfolio                   8%
 13%                                                                                               A                                            Telser Portfolio
                                                           Telser Portfolio            7%
 11%
                                                                                       6%                                                       Sharpe Portfolio
  9%                                                       Sharpe Portfolio
            A
  7%                                                                                   5%
  5%                                                                                   4%
  3%                                                                                   3%
       1%          6%            11%            16%         21%               26%           1%      3%      5%        7%       9%       11%     13%        15%         17%
                                         Risk                                                                                  Risk

Portfolio                            A          Sharpe    Telser          B          Portfolio                               A        Sharpe    Telser               B
Return                             6.01%        5.12%     8.16%        14.95%        Return                                7.40%      5.21%     8.38%              9.74%
Risk                               2.48%        2.02%     4.19%        12.25%        Risk                                  3.31%      2.15%     4.26%              6.34%
Sharpe Ratio                        2.43         2.53      1.95         1.22         Sharpe Ratio                           2.24       2.42      1.97               1.54

Asset class weights                                                                  Asset class weights
EM Equity                             -            -        -                -       EM Equity                               -           -         -                 -
EM Sovereign Credit                   -            -       5.83%           31.72%    EM Sovereign Credit                     -           -        4.32%            15.64%
EM Money Market                     23.25%       24.45%   11.39%             -       EM Money Market                       40.83%      27.59%    50.32%            58.06%
S&P 500                               -            -        -                -       S&P 500                                4.37%       1.27%     8.57%            15.43%
MSCI World Equities                   -            -        -                -       MSCI World Equities                     -           -         -                 -
Commodities                           -            -        -                -       Commodities                            2.94%       1.44%     5.75%            10.86%
Real Estate                           -            -        -                -       Real Estate                             -          0.20%      -                 -
Gold                                 6.95%        4.88%   12.91%           52.02%    Gold                                    -          0.54%      -                 -
JP Morgan GBI                         -           9.23%     -                -       JP Morgan GBI                           -         25.63%      -                 -
JP Morgan US Agg Bond               49.03%       33.00%   64.48%           16.26%    JP Morgan US Agg Bond                 49.72%      18.62%    31.04%                  -
Dollar Index                        20.77%       28.43%    5.39%             -       Dollar Index                           2.14%      24.70%      -                 -
Source: Morgan Stanley Research, Bloomberg                                           Source: Morgan Stanley Research, Bloomberg
Exhibit 180                                                                          Exhibit 182
Risk and Return by Asset Class, 2006-10                                              Risk and Return by Asset Class, 1994-2007
                                        Expected Return              Risk                                                    Expected Return                Risk
EM Equity                                   12.61%                 29.47%            EM Equity                                   7.97%                    19.92%
EM Sovereign Credit                         9.11%                  11.56%            EM Sovereign Credit                         10.85%                   12.70%
EM Money Market                             8.08%                   8.50%            EM Money Market                             8.81%                     5.78%
S&P 500                                     3.17%                  22.09%            S&P 500                                     11.25%                   15.56%
MSCI World Equities                         1.44%                  22.59%            MSCI World Equities                         7.94%                    13.78%
Commodities                                 -4.39%                 28.37%            Commodities                                 10.99%                   19.67%
Real Estate                                 3.17%                  38.17%            Real Estate                                 6.02%                    14.06%
Gold                                        21.21%                 21.83%            Gold                                        6.53%                    14.66%
JP Morgan GBI                               6.80%                   6.25%            JP Morgan GBI                               6.44%                     5.71%
JP Morgan US Agg Bond                       6.33%                   3.95%            JP Morgan US Agg Bond                       6.07%                     4.09%
Dollar Index                                -2.15%                  8.94%            Dollar Index                                -1.42%                   7.62%
Source: Morgan Stanley Research, Bloomberg                                           Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                       107
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




V. Correlation Results
Exhibit 183
EM Equity Correlations
                        Full Sample   1994-1999   2000-2005   2006-2010   1994-2007
EM Sovereign Credit           0.59        0.65        0.43        0.74        0.53
EM Money Market               0.68        0.58        0.50        0.83        0.58
S&P 500                       0.65        0.53        0.57        0.78        0.55
MSCI World Equities           0.77        0.66        0.69        0.88        0.69
Commodities                   0.32        0.13        0.10        0.55        0.13
Real Estate                   0.48        0.37        0.27        0.61        0.34
Gold                          0.19        0.09        0.16        0.24        0.20
JP Morgan GBI                 0.05       -0.07       -0.02        0.19       -0.01
JP Morgan US Agg Bond        -0.10       -0.10       -0.18       -0.05       -0.13
Dollar Index                 -0.25        0.04       -0.12       -0.54       -0.10
Source: Morgan Stanley Research, Bloomberg
Exhibit 184
EM Sovereign Credit Correlations
                        Full Sample   1994-1999   2000-2005   2006-2010   1994-2007
EM Equity                     0.59        0.65        0.43        0.74        0.53
EM Money Market               0.48        0.36        0.51        0.70        0.40
S&P 500                       0.39        0.40        0.21        0.63        0.30
MSCI World Equities           0.44        0.44        0.26        0.69        0.34
Commodities                   0.18        0.10        0.11        0.37        0.09
Real Estate                   0.33        0.34        0.15        0.54        0.23
Gold                          0.08        0.03        0.22        0.07        0.09
JP Morgan GBI                 0.13       -0.02        0.31        0.22        0.11
JP Morgan US Agg Bond         0.22        0.17        0.35        0.19        0.23
Dollar Index                 -0.11        0.17       -0.20       -0.44        0.01
Source: Morgan Stanley Research, Bloomberg
Exhibit 185
EM Money Market Correlations
                        Full Sample   1994-1999   2000-2005   2006-2010   1994-2007
EM Equity                     0.68        0.58        0.50        0.83        0.58
EM Sovereign Credit           0.48        0.36        0.51        0.70        0.40
S&P 500                       0.40        0.19        0.19        0.67        0.23
MSCI World Equities           0.54        0.32        0.32        0.79        0.36
Commodities                   0.32        0.12        0.14        0.56        0.15
Real Estate                   0.38        0.05        0.19        0.56        0.17
Gold                          0.29        0.23        0.37        0.30        0.31
JP Morgan GBI                 0.37        0.16        0.53        0.44        0.35
JP Morgan US Agg Bond        -0.02       -0.21        0.15        0.05       -0.04
Dollar Index                 -0.56       -0.29       -0.59       -0.76       -0.45
Source: Morgan Stanley Research, Bloomberg




                                                                                                                108
                                                                           MORGAN STANLEY RESEARCH

                                                                           March, 2012
                                                                           EM Local Markets Guide




EM Flows – Should I Stay or Should I Go?
Juha Seppala
                      th
First Published Sept 30 , 2011: EM Flows - Should I stay or Should I Go?

In EM Quantitative Strategy Update: Go with the (EM) Flow                   Exhibit 186
(July 29, 2011), we analyzed what determines and, more                      TED Spread vs. Euribor-OIS Spread
importantly, predicts the demand for EM assets as captured                   500                                                                          2.5
by total private flows to EM countries. We found that the total              450
flow to the EM in 2011 should be in the same ballpark it was                 400                                                                          2
in 2009 and 2010. That is, the demand for EM assets is likely                350
to remain strong. The biggest risk to this call is a sudden                  300                                                                          1.5
reduction in DM liquidity as measured by reduced M2 or an                    250
increased global risk aversion as measured by an upward                      200                                                                          1
spike in TED spread.                                                         150
However, in retrospect, we see two potential problems in                     100                                                                          0.5
using TED spread as the measure of global risk aversion in                    50
the current situation. First, when the short-term US Treasury                  0                                                                          0
                                                                               Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11
rates are close to the zero bound, TED spread is probably                                   Ted Spread                      Euribor-OIS Spread (RHS)
not as sensitive as it would otherwise be. Second, the
                                                                            Source: Morgan Stanley Research, Bloomberg.
current confidence crisis is very Europe-centered, so
different measures might be more appropriate. This is
                                                                            Exhibit 187
demonstrated in Exhibit 186, which displays TED spread and
                                                                            Regressing Euribor-OIS Spread in 2002-2008 Data
Euribor-OIS spread since the beginning of 2007. While these
                                                                            (R2 = 0.89)
spreads did co-move until the summer 2010, the recent
                                                                                               Variable       Coefficient           Std Err        t-Statistic
increase in TED spread is nowhere near current elevated                     Constant                                 0.02              0.01              1.82
levels in Euribor-OIS spread                                                TED Spread                               0.00              0.00             26.16
                                                                            ECRI                                    -0.02              0.00            -13.65
The reason we chose TED spread in the first place was that
                                                                            Source: Morgan Stanley Research, Bloomberg.
it worked much better than VIX in our sample and different
Libor-OIS measures were not available before late 2001.
                                                                            Predicting EM Flows with Euribor-OIS Spread
How to simultaneously measure the current spike in global
risk aversion and have a long series of this variable? The                  As in the July EMQSU, we used the following variables in our
answer is suggested by Exhibit 186. If TED spread and                       investigation – in addition to Euribor-OIS spread: 1) GDP-
Euribor-OIS spread did co-move before 2010, why not                         weighted policy rate differential between DM (represented by
regress the latter on the former, possibly with other variables,            G10 countries) and EM countries; 2) GDP-weighted inflation
using 2002-2008 and then use fitted values for 2000-2001 to                 differential; 3) real GDP growth differential; and 4) the total
generate the missing observations?                                          G10 M2. Again as we pointed out in the July EMQSU, the
                                                                            factors that determine EM flows can be divided into two
That is exactly what we did. The best other variable,
                                                                            camps: “push factors” – which direct flows out of developed
somewhat surprisingly, turned out to be ECRI weekly leading
                                                                            markets, and “pull factors” – which attract more flows to
index growth rate we discussed in the above mentioned July
                                                                            emerging market economies
29 EMQSU. The results are presented in Exhibit 187.
                                                                            The first three (EM vs DM differentials) represent both push
                                                                            and pull factors; M2 as a proxy for global liquidity is a push
                                                                            factor; and Euribor-OIS spread as our preferred measure of
                                                                            global risk appetite is a pull factor. Exhibit 188 through
                                                                            Exhibit 192 summarize our models for EM private flows, 0 to
                                                                            4 quarters ahead.



                                                                                                                                                         109
                                                                                 MORGAN STANLEY RESEARCH

                                                                                 March, 2012
                                                                                 EM Local Markets Guide




Exhibit 188                                                                       Exhibit 193
Explaining EM Total Private Flows (R2 = 0.74)                                     Actual and (after 4Q2010) Predicted EM Flows
        Variable             Coefficient             Std Err      t-Statistic       350,000
Constant                     -148,770.00           38,266.02           -3.89        300,000
M2                                  0.02                0.00           10.54        250,000
Euribo-OIS                   -212,035.27           32,489.09           -6.53        200,000
Real Rate Diff                -24,234.37            8,424.87           -2.88        150,000
Source: Morgan Stanley Research, Bloomberg, Haver, IMF.                             100,000
                                                                                     50,000
Exhibit 189
                                                                                          0
Predicting EM Flows 1Q Ahead (R2 = 0.64)
                                                                                    -50,000
Variable                     Coefficient             Std Err      t-Statistic      -100,000
Constant                     -175,411.15           48,147.49           -3.64       -150,000
M2(t-1)                             0.04                0.01            5.88                Mar-00     Mar-02     Mar-04      Mar-06     Mar-08   Mar-10   Mar-12
Euribor-OIS(t-1)             -193,199.15           39,417.71           -4.90
                                                                                  Source: Morgan Stanley Research (forecasts), IMF (actual).
Groth Diff(t-1)               -34,314.84           14,501.26           -2.37
Real Rate Diff(t-1)           -18,880.32            9,973.59           -1.89
Source: Morgan Stanley Research, Bloomberg, Haver, IMF.
                                                                                  In
Exhibit 190
                                                                                  Exhibit 193, we display the realized EM total private flows
Predicting EM Flows 2Q Ahead (R2 = 0.50)                                          during 1Q2000-4Q2011 and our model generated predictions
                      Variable    Coefficient           Std Err    t-Statistic
                                                                                  for 1Q2011-3Q2012 using the realizations of explanatory
Constant                          -333,950.89         78,601.59         -4.25
                                                                                  variables as they could be observed at the end of 3Q2011
M2*(t-2)                                 0.03              0.00          6.24
Euribor-OIS(t-2)                  -141,618.86         49,086.40         -2.89
                                                                                  (EM flows in 2011 are still unavailable). The models predict
Inflation Diff(t-4)                 32,559.71         11,186.82          2.91     that total flow to the EM in 2011 will be USD0.97 trillion
Source: Morgan Stanley Research, Bloomberg, Haver, IMF.                           compared to USD1.20 trillion in 2010 and USD0.95 trillion in
                                                                                  2009. Moreover, they predict that the total flows in 4Q2011-
Exhibit 191
                                                                                  3Q2012 would be only USD0.89 trillion or 25% reduction
Predicting EM Flows 3Q Ahead (R2 = 0.51)
                                                                                  from the last published (1Q2010-4Q2010) numbers.
Variable                     Coefficient            Std Err       t-Statistic
Constant                     -486,195.26         112,435.98            -4.32      What would be the effect on EM assets? It is hard to move
Flow(t-3)                          -0.54               0.23            -2.34      from quarterly data to higher frequency assets, but we did
M2*(-3)                             0.04               0.01             5.05      find that changes in EM flows predict future changes in
Euribor-OIS(t-3)             -232,920.81          75,536.67            -3.08      EMFX levels. According to our econometric work, a 25%
Inflation Diff(-4)             45,772.18          13,623.81             3.36      reduction in EM flows in one quarter corresponds to 5%
Source: Morgan Stanley Research, Bloomberg, Haver, IMF.                           depreciation in the GDP-weighted EMFX index in the next
Exhibit 192
                                                                                  quarter.
Predicting EM Flows 4Q Ahead (R2 = 0.47)                                          Conclusion
                      Variable    Coefficient           Std Err    t-Statistic
Constant                          -326,798.50         93,890.37         -3.48     Our revised model predicts that total flow to the EM in 2011
M2(t-4)                                  0.02              0.00          6.04     will be USD0.97 trillion compared to USD1.20 trillion in 2010
Policy Rate Diff(t-4)               27,581.24         11,194.12          2.46     and USD0.95 trillion in 2009. This downward revision in our
Source: Morgan Stanley Research, Bloomberg, Haver, IMF.                           estimate is caused by the upward spike in our new preferred
The overall trend in flows to EM is driven by DM liquidity as                     global risk aversion measure, Euribor-OIS spread. We
captured by G10 M2, but occasional drawdowns are best                             predict that the total flows in 4Q2011-3Q2012 will be only
explained and predicted by our preferred risk aversion                            USD0.89 trillion or a 25% reduction from the last published
measure, Euribor-OIS spread. It remains statistically                             (2010) numbers. This predicts a 5% depreciation in the GDP-
significant until three quarters ahead.                                           weighted EMFX index. Flows should bounce back in
                                                                                  3Q2012.




                                                                                                                                                              110
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Valuing Latin America Rates
Juha Seppala
First Published December 2nd, 2010: Valuing Latin American Rates

We present a macro-financial model for fair risk premia             Exhibit 194
and apply it to the Brazil DI, Mexico TIIE, and Chile               Back-Testing for DI Jan’11 Contract – Number of
CLPxCamara curves. Back-testing of our models                       Correct Predictions = 10, Number of Incorrect
indicates a success rate in the range of 70% in predicting          Predictions = 4
interest rate movements.                                                                                            Predicted     Actual
                                                                         Date         Market           Fair          Change       Change
Nominal interest rates can be decomposed into a                         29-May-09
                                                                        26-Jun-09
                                                                                       9.74
                                                                                       9.92
                                                                                                       9.75
                                                                                                      9.84                    1      18        Correct
compensation for postponing the use of funds (real interest               31-Jul-09    9.84            9.85                  -8       -8       Correct
                                                                        28-Aug-09      9.76            9.76                   1       -8       Incorrect
rate), a compensation for the variation in the value of nominal         25-Sep-09     10.23           10.21                   0       47       Correct
                                                                         30-Oct-09    10.34           10.25                  -2      11        Incorrect
balances (expected inflation), a compensation due to                    27-Nov-09     10.29           10.35                  -9       -5       Correct
uncertainty about variations in real interest rates and inflation       31-Dec-09     10.50           10.38                   6       21       Correct
                                                                        31-Jan-10     10.32           10.29                 -12      -18       Correct
during the holding period (risk premia) 19.                             27-Feb-10     10.48           10.38                  -3      16        Incorrect
                                                                        31-Mar-10     10.40           10.36                 -10       -8       Correct
                                                                         30-Apr-10    11.12           11.01                  -4      72        Incorrect
Alternatively, local nominal curves can be expressed as a               31-May-10     10.96           11.03                 -11      -16       Correct
                                                                        30-Jun-10     11.36           11.25                   7       40       Correct
sum of the expected path of the nominal short-term interest               28-Jul-10   10.83           10.39                 -11      -53       Correct
rates, which are commonly forecast by economists, plus fair         Source: Bloomberg, Morgan Stanley

risk premia, whose computations are far from being a trivial
exercise, and have increasing importance for longer interest        Exhibit 195
rate tenors. Hence, interest rates strategy depends crucially       Current and Fair Prices of Brazil DI Contracts
on good assessment of how much investors are willing to             According to Our Model
receive, on average, as compensation for risk.                        13.0
                                                                                               Market Price    Fair Price

To give an idea of the performance of our model, Exhibit 194
                                                                      12.5
displays the ex-ante predictions and ex-post realized rates for
the DI Jan’11 contract. The model predicted correctly the             12.0
changes in the DI rate in 10 out of 14 months. We view this
result quite respectable.                                             11.5


Exhibit 195 summarizes our model’s current fair value
                                                                      11.0
estimates. The short tenors, up to Jan’12 contract, are cheap
while the longer tenors, up to Jan’22 contract, are rich.             10.5


                                                                      10.0
                                                                     Aug-10      Aug-12    Aug-14     Aug-16     Aug-18     Aug-20    Aug-22       Aug-24
                                                                    Source: Morgan Stanley, Bloomberg, central bank of Brazil


                                                                    The Objective
                                                                    The objective of our econometric macro-finance models is to
                                                                    carve out a measure of fair risk premia, i.e., a measure of the
                                                                    compensation for risk, given some observed (and forecast)
                                                                    values of a set of macro-finance fundamentals.

                                                                    Suppose that the long-term rates go up. This could be due to
                                                                    the fact that the market is expecting more hikes in the future
                                                                    or because the risk premia have gone up or some
                                                                    combination of both. In order to gauge whether this increase
                                                                    in rates is justifiable by the fundamentals, one needs to have
19
     There can also be a compensation due to liquidity problems.


                                                                                                                                                           111
                                                                                            MORGAN STANLEY RESEARCH

                                                                                            March, 2012
                                                                                            EM Local Markets Guide




a disciplined view of what is driving the move: premia or                                   fundamentals without too much rigidity typical of theoretical
expectations. In Latin American countries, it is possible to                                models.
distinguish the two by looking at survey data published by
                                                                                            We begin by obtaining a crude measure of empirical
central banks in the region. This allows us both to develop a
                                                                                            (observable) ‘forward’ risk premia as the difference between
fair value model for risk premia and compare our economists’
                                                                                            overnight forward rates and expected overnight rates. The
expectation with consensus expectations. If either component
                                                                                            latter are extracted from the Brazilian central bank’s surveys
has a significant divergence then that may flag a trade
                                                                                            of economic forecasts, available on its web page. By
opportunity.
                                                                                            integrating these forward risk premia, we can obtain ‘spot’ risk
We will start publishing the results of our models under                                    premia, the respective measure for spot prices.
different scenarios in the, EM Quantitative Strategy Update.
                                                                                            The next step consists of estimating the statistical relationship
We welcome feedback from our readers on what kinds of
                                                                                            between the observed risk premium and a set of financial and
scenarios interest them most.
                                                                                            macroeconomic fundamentals. These explanatory variables
Our objective is cross-sectional – that is, we want to have a                               span the set of factors that we have found to be relevant to
current snapshot of fair values across the curve – and not                                  explain the behavior of the risk premia. Some variables reflect
time-series path forecast for any single instrument as was                                  global financial conditions (among them VIX and EM CDX),
done in a recent publication, EM Profile: The Near-Term Path                                others capture idiosyncratic aspects of the country of interest
for EM Rates — A Scenario Approach (November 5, 2010).                                      (e.g., FX volatility and sovereign CDS spreads), while others
However, one could also use the present model as a                                          capture macroeconomic conditions (among them activity or
complementary approach for obtaining path forecasts for                                     production indexes, and different price indices).
different curve instruments, assuming one is willing to make
                                                                                            After this, we use the predictions of the model to compute the
assumptions about future risk premia. Alternatively, one could
                                                                                            fair risk premium based on observed and forecast values of
use our fair risk premium to forecast future risk premia given
                                                                                            the fundamentals. As such, our model can be used not only to
macro-finance data observable today.
                                                                                            determine the current measure of fair risk premium, but also
The Model                                                                                   to forecast the dynamic evolution of this premium in the short
                                                                                            term.
The objective of our econometric macro-finance model is to
carve out a measure of fair risk premia, i.e., a measure of the                             Finally, we combine our fair risk premium estimates with
compensation for risk, given some observed (and forecast)                                   Brazil, Chile and Mexico monetary policy target rate
values of a set of macro-finance fundamentals.                                              expectations to price all available DI contracts in Brazil,
                                                                                            CLPxCamara in Chile, and TIIE contracts in Mexico.
Extensive literature deals with the term structure of interest                              However, consensus expectations are only used for 2012 and
rates and calculations of the risk premia for developed                                     beyond. For 2010 and 2011 we use the monetary policy path
economies, principally for the US. The risk premia models                                   forecast by Morgan Stanley’s economists.
range from purely econometric affine models, which use
unobservable latent factors, to fully fledged general                                       Without entering into a discussion of technical details, one
equilibrium macroeconomic models, which explain the yield                                   final point concerning the treatment of the curves is worth
curve with specific relationships between observable                                        mentioning. To obtain well behaved spot and forward curves,
macroeconomic fundamentals. 20                                                              monetary policy expectations, and risk premia, we apply a
                                                                                            smoothing procedure. This guarantees that the obtained risk
Our model follows a middle way approach by establishing an                                  premium does not adopt counterintuitive shapes and reduces
econometric relationship between the risk premia and a set of                               possible measure errors in market prices or estimates of
relevant financial and macroeconomic variables. The                                         market expectations. 21 Exhibit 196 illustrates how we find
approach is flexible enough to capture idiosyncratic                                        empirical forward risk premia. Notice how the risk premia are
characteristics of Brazil, Chile and Mexico and macro-finance                               negative when the expected Selic is above the O/N forward
                                                                                            rate.

20
  For a review of the former, see Monika Piazzesi “Affine Term Structure Models,” in the
Handbook of Financial Econometrics, volume 1, 2009. For examples of the latter applied to
nominal term structure, see Federico Ravenna and Juha Seppala Monetary Policy and
                                                                                            21
Rejections of Expectations Hypothesis, 2007, and Monetary Policy, Expected Inflation, and     We follow Lars Svensson “Estimating Forward Interest Rates with the Extended Nelson &
Inflation Risk Premium, 2007.                                                               Siegel Method”, Quarterly Review, Sveriges Riksbank, 1995.


                                                                                                                                                                              112
                                                                                                                       MORGAN STANLEY RESEARCH

                                                                                                                       March, 2012
                                                                                                                       EM Local Markets Guide




Chile and Mexico have one additional complication compared                                                             Exhibit 198
to Brazil. Brazil DI contracts are essentially spot (zero-                                                             Risk Premia and Monetary Policy in Chile
coupon) rates, so that repricing the curve using monetary
policy expectations and risk premia estimates is straight-                                                               400
                                                                                                                                                             5Y Risk Premium             O/N Exp (RHS)
                                                                                                                                                                                                                   9

forward. In Chile and Mexico, CLPxCamara and TIIE                                                                                                                                                                  8
contracts are swap contracts and quoted yields are par yields.                                                           300

Hence, we first convert par rates to spot rates, and then                                                                                                                                                          7

calculate the fair spot curve, and finally convert the spot curve                                                        200                                                                                       6
to par curve. The attached Technical Appendix explains the
details of these conversions.                                                                                                                                                                                      5
                                                                                                                         100
                                                                                                                                                                                                                   4
Exhibit 196
How to Obtain Empirical Risk Premia?                                                                                      0                                                                                        3
13                                                                                                               320   September-04   October-05    November-06 December-07        January-09      February-10
              (Forward) Risk Premium (RHS)    DI Curve         O/N Forward         Expected Selic
                                                                                                                                                                                                                   2
                                                                                                                 260    -100
12                                                                                                                                                                                                                 1

                                                                                                                 200
                                                                                                                        -200                                                                                       0

11                                                                                                               140   Source: Bloomberg, Morgan Stanley, central bank of Chile
                                                                                                                       Exhibit 199
                                                                                                                 80
                                                                                                                       Risk Premia and Monetary Policy in Mexico
10

                                                                                                                 20
                                                                                                                       500                                                                                             10
                                                                                                                                                                       5Y Risk Premium             O/N Exp (RHS)
 9                                                                                                               -40   450
     0             1                2          3                4              5                    6        7
                                                                                                                                                                                                                       9
Source: Bloomberg, Morgan Stanley, central bank of Brazil                                                              400

                                                                                                                       350
Exhibit 197 displays the dynamics of risk premia and policy                                                                                                                                                            8

rate in Brazil. The risk premium tends to increase when the                                                            300

target rate is on hold between the end of the cutting cycle and                                                        250                                                                                             7

the start of the hiking cycle. When hikes begin (or, more                                                              200
precisely, when uncertainty about when they will start                                                                                                                                                                 6
                                                                                                                       150
declines) risk premia collapse very fast. Moreover,
disregarding the spike of October 2008 in Brazil, they tend to                                                         100
                                                                                                                                                                                                                       5
be the highest/lowest when the target rate is at the                                                                    50
lowest/highest point within the easing/hiking cycle.                                                                      0                                                                                            4
                                                                                                                          Jan-04      Jan-05       Jan-06    Jan-07       Jan-08          Jan-09         Jan-10
Exhibit 197
                                                                                                                       Source: Bloomberg, Morgan Stanley, central bank of Mexico
Risk Premia and Monetary Policy in Brazil
                                                                                                                       Our EM Profile: Hiking Cycles and Term Spreads in Latin
 1400                                                                                                   21
                                                    3yr Risk Premium         O/N Selic (RHS)                           America (May 26, 2010) addressed the behavior of the term
 1200                                                                                                   19             spreads in the major Latin American countries around the
                                                                                                                       beginning of hiking cycles. Our result was that the curve
 1000                                                                                                   17
                                                                                                                       flatteners were attractive risk-reward trades in the months
  800                                                                                                   15             immediately before and after the month when central banks
                                                                                                                       began their hiking cycles. The collapse in risk premium
  600                                                                                                   13
                                                                                                                       identified in the previous paragraph is one major factor behind
  400                                                                                                   11
                                                                                                                       this pattern.

  200                                                                                                   9



   0                                                                                                    7
 January-04     February-05        March-06    April-07         May-08        June-09

Source: Bloomberg, Morgan Stanley, central bank of Brazil



                                                                                                                                                                                                                   113
                                                                                                              MORGAN STANLEY RESEARCH

                                                                                                              March, 2012
                                                                                                              EM Local Markets Guide




Exhibit 202 show the average differences for each month                                                       Exhibit 202
before and after the beginning of the hiking cycle relative to                                                Cumulative Change in 5Y Risk Premia Relative to
the level at the beginning of the cycle. 22 The differences                                                   the Start of the Hiking Cycle in Mexico
clearly document how the risk premia begin to tighten before                                                   300
the hikes begin and continue to do so for several months
afterwards. Typically, longer tenors tighten more.                                                             250


Exhibit 200                                                                                                    200
Cumulative Change in Risk Premia Relative to the
                                                                                                               150
Start of the Hiking Cycle in Brazil
                                                                                                               100
     4
                                                                      1yr               3yr
     3                                                                                                             50
                                                                      5yr               7yr

     2                                                                                                              0
                                                                                                                         -11 -10   -9    -8   -7   -6   -5    -4   -3   -2   -1   0   1   2   3   4
     1                                                                                                                                         Months from the beg. of hiking cycle
                                                                                                                   -50
                                                                                                              Source: Bloomberg, Morgan Stanley, central bank of Mexico
     0
              -5    -4        -3       -2        -1     0        1      2       3             4       5
 -1                Months from the beg. of hiking cycle                                                       Historical Back-Test
 -2                                                                                                           Exhibit 203 displays the ex-ante predictions and ex-post
                                                                                                              realized rates for the DI Jan’11 and Jan’17 contracts. Exhibit
 -3
                                                                                                              204 and Exhibit 205 display the ex-ante predictions and ex-
 -4                                                                                                           post realized rates for the 5y CLPxCamara and 10y TIIE
Source: Bloomberg, Morgan Stanley, central bank of Brazil                                                     contracts. Shorter tenors depend crucially on overnight
                                                                                                              interest rate predictions whereas longer tenors are more
Exhibit 201                                                                                                   sensitive to risk premia in addition to long-term expectations.
Cumulative Change in 5Y Risk Premia Relative to                                                               Given the difficulty of forming monetary policy expectations
the Start of the Hiking Cycle in Chile                                                                        seven years ahead, it should not be a surprise that the
                                                                                                              success ratio of our back-testing is better for Jan’11 tenor,
     200
                                                                                                              where the model predicted correctly the changes in the DI
                                                                                                              rate in 10 out of 14 months. Nonetheless, the success ratio of
     150
                                                                                                              8, 9 and 8 out of 12 months 23 for Jan’17, 5y CLPxCamara,
                                                                                                              and 10y TIIE, respectively, is still quite respectable, in our
     100
                                                                                                              view.

         50                                                                                                   In particular, the model has been able to predict quite well the
                                                                                                              tightening of the Jan’17 rate that has taken place during the
          0                                                                                                   first half of 2010. As we noted above and in our EM Profile:
              -11 -10    -9    -8     -7    -6    -5   -4   -3   -2    -1   0       1         2   3       4   Hiking Cycles and Term Spreads in Latin America (May 26,
     -50                            Months from the beg. of hiking cycle                                      2010), the curve flatteners have been attractive risk-reward
                                                                                                              trades in the months immediately before and after the months
 -100                                                                                                         when central banks begin their hiking cycles.
Source: Bloomberg, Morgan Stanley, central bank of Chile




22
     The hiking cycles begin in November 2004, April 2008 and April 2010 in Brazil, in July
     2007 and July 2010 in Chile, and in April 2007 in Mexico. The contribution of the last hike              23
                                                                                                                   In addition to two months which were inconclusive.
     for Brazil is used only for the months before the beginning of the cycle.


                                                                                                                                                                                                  114
                                                                                  MORGAN STANLEY RESEARCH

                                                                                  March, 2012
                                                                                  EM Local Markets Guide




Exhibit 203                                                                       Technical Appendix
Back-Testing for DI Jan’17 Contract – Number of
                                                                                  The par yield for a certain bond maturity is the coupon rate
Correct Predictions = 8, Number of Incorrect
                                                                                  that causes the bond price to equal its par value.
Predictions = 4
                                           Predicted         Actual               In order to compute rates for different maturities, we
     Date          Market        Fair       Change           Change
    29-May-09      11.75        11.87                                             interpolate the observable yields based on the Svensson
    26-Jun-09      12.25        12.26                   12     50     Correct
      31-Jul-09    12.54        12.85                    1     29     Correct     model. This interpolation method is broadly used in our
    28-Aug-09
    25-Sep-09
                   12.64
                   12.72
                                12.91
                                12.76
                                                        31
                                                        27
                                                               10
                                                               8
                                                                      Correct
                                                                      Correct
                                                                                  methodology to transform par yields to spot rates and vice
     30-Oct-09     13.10        13.26                    4     38     Correct     versa 24.
    27-Nov-09      13.10        13.46                   16     0      N/A
    31-Dec-09      13.10        13.12                   36     0      N/A
    31-Jan-10      13.02        13.18                    2     -8     Incorrect
    27-Feb-10      12.72        12.61                   16    -30     Incorrect   Compute the Spot Rates from Par Yields
    31-Mar-10      12.26        12.19                  -11    -46     Correct
     30-Apr-10     12.41        12.16                   -7     15     Incorrect   The price for an n-period bond is:
    31-May-10      12.30        12.20                  -25    -11     Correct
    30-Jun-10      12.18        12.35                  -10    -12     Correct
      28-Jul-10    11.93        12.39                   17    -25     Incorrect   P  d1 * C  d 2 * C    d n * C  d n * F
Source: Bloomberg, Morgan Stanley
                                                                                  Where:
Exhibit 204
Back-Testing for 5Y CLPxCamara Contract –                                         P  price of a bond
Number of Correct Predictions = 9, Number of                                      C  coupon
Incorrect Predictions = 3
                                                                                  F  face value
                                           Predicted         Actual
     Date          Market           Fair    Change           Change               d i  discount factor of the ith coupon date
     10-Dec-09      5.27            4.97
       8-Jan-10     5.21            4.85      -30              -6     Correct
     10-Feb-10      4.99            4.66      -36             -22     Correct     Considering the par value of the bond and rearranging the
     10-Mar-10      4.81            4.55      -33             -18     Correct
       9-Apr-10     4.93            5.17      -26              12     Incorrect   equation:
     11-May-10      5.08            4.82       24              15     Correct
     10-Jun-10      5.05            4.78      -26              -3     Correct
       10-Jul-10    4.96            4.63      -27              -9     Correct      P                   C
       7-Aug-10     5.06            4.79      -33              10     Incorrect       1 and              yn
       9-Sep-10     4.94            4.77      -27             -12     Correct      F                   F
      12-Oct-10     4.90            4.99      -17              -4     Correct
     12-Nov-10      5.14            4.91        9              24     Correct
                                                                                             Annualized Par Yield
       1-Dec-10     5.45
Source: Bloomberg, Morgan Stanley
                                    4.98      -23              31     Incorrect
                                                                                   yn 
                                                                                          Coupons Payments per year
Exhibit 205
Back-Testing for 10Y TIIE Contract – Number of                                    P  d1 * C  d 2 * C    d n * C  d n * F
Correct Predictions = 8, Number of Incorrect
Predictions = 4

     Date          Market           Fair
                                           Predicted
                                            Change
                                                             Actual
                                                             Change                1  d1 * y n  d 2 * y n    d n * y n  d n
     30-Nov-09      8.17            8.18
     30-Dec-09      8.32            8.27        1              15     Correct
     30-Jan-10      8.20            8.16       -5             -12     Correct

                                                                                   1  d1  d 2    d n 1 * y n  d n *  y n  1
     28-Feb-10      7.96            7.98       -4             -24     Correct
     30-Mar-10      7.88            7.90        2              -8     Incorrect
      30-Apr-10     7.75            8.07        2             -13     Incorrect
     30-May-10      7.64            7.19       32             -11     Incorrect

                                                                                                 1  d1  d 2    d n 1 * yn
     30-Jun-10      7.30            6.62      -45             -34     Correct
       30-Jul-10    6.92            6.26      -68             -38     Correct
     31-Aug-10      6.61            5.90      -66             -31     Correct      dn 
     30-Sep-10
      29-Oct-10
                    6.46
                    6.42
                                    5.41
                                    5.36
                                              -71
                                             -105
                                                              -15
                                                               -4
                                                                      Correct
                                                                      Correct                                 1  yn
       1-Dec-10     7.20            7.23     -106              78     Incorrect
Source: Bloomberg, Morgan Stanley
                                                                                  Using the equation above it is possible to find each discount
Future Work                                                                       factor and consequently the respective spot rate.

Even though consensus expectations are not readily available
in CEEMEA, one may be able to approximate these missing
variables and replicate the process there as well.
                                                                                  24
                                                                                    For more information see Lars Svensson “Estimating Forward Interest Rates with the
                                                                                  Extended Nelson & Siegel Method”, Quarterly Review, Sveriges Riksbank, 1995.


                                                                                                                                                                     115
                                                                           MORGAN STANLEY RESEARCH

                                                                           March, 2012
                                                                           EM Local Markets Guide




In order to come up with the d value, we compute iteratively               Compute the Par Yields from Spot Rates
                                       r  y1 .
                                       n
  d n1   , using as initial condition: 1                                  The price for an n- period bond is:

The procedure is done as follows:                                          P  d1 * C  d 2 * C    d n * C  d n * F

          1                 1  d1 * y2        1  d1  d 2 * y3        1  d1  d 2    d n1  d n  * y n  d n
d1              ,   d2                  , d3                        ,
       1  y1                  1  y2                  1  y3
                                                                                                 1  dn
                                                                            yn 
          1  d1  d 2    d n 1 * yn                                             d1  d 2    d n1  d n
…,   dn 
                       1  yn
                                                                                         Spot Ratei
Translating to spot rates:
                                                                           ri 
                                                                                  Coupon Payments per year
 d i  1  ri   ri  d i i  1
                                   1
                i             


                                                                           d i  1  ri 
                                                                                         i

Spot Ratei  Coupon Payments per year * ri
                                                                                                    1
This way we get the spot rates for each coupon date.                       Starting with d1            and so on, we get the entire par
                                                                           yield curve.       1  r1



                                                                                  1  d1        1 d2                         1 dn
                                                                           y1           , y2           , …, y n 
                                                                                    d1          d1  d 2            d1  d 2    d n1  d n

                                                                           Annualized Par Yield  y n * Coupon Payments per year




                                                                                                                                       116
                                                                  MORGAN STANLEY RESEARCH

                                                                  March, 2012
                                                                  EM Local Markets Guide




The Risk Premia in CEEMEA
Juha Seppala, Robert Habib
                          th
First Published December 15 , 2011: The Risk Premia in CEEMEA

In December 2010, we launched a macro-financial model for          interest them most. Our objective is cross-sectional – that is,
fair LatAm rates curves (see EM Profile: How to Value Brazil       we want to have a current snapshot of fair values across the
Rates? July 29, 2010, and EM Profile: How to Value Latin           curve – and not time-series path forecast for any single
America Rates? December 2, 2010). We have now extended             instrument as was done in a recent publication, EM
the model to five CEEMEA countries: Czech Republic,                Quantitative Strategy Update: EMTrend – The Near-Term
Hungary, Poland, Israel and South Africa. As a matter of fact,     Path for EM Rates (October 13, 2010). However, one could
the procedure could be and has been implemented for                also use the present model as a complementary approach
Turkey as well. The only problem is that recently (see Turkey      for obtaining path forecasts for different curve instruments,
Economics: Decisive Action from CBT: Banks Enter a Dark            assuming one is willing to make assumptions about future
Corridor, October 26, 2011) the Central Bank of Turkey             risk premia. Alternatively, one could use our fair risk premium
introduced an interest rate corridor of 5.75-12.5% for the         to forecast future risk premia given macro-finance data
effective policy rate. Given that, at any given day, the policy    observable today.
rate can be anything within this band, our method for
                                                                   The Model
evaluating the fair market rates based on expected policy
rates unfortunately cannot work. Once CBT returns to more          The core idea is to closely analyze consensus data of Brazil,
orthodox monetary policy, we will add Turkey back to the           Chile and Mexico, look for patterns, and use them to
model.                                                             calibrate our model, which may be viewed as an improved
                                                                   version of the first-order difference of a Taylor rule. (We
The objective of our econometric macro-finance models is to
                                                                   explain the mathematical details of the model in the
carve out a measure of fair risk premia, i.e., a measure of the
                                                                   appendix.) To maintain its simplicity, only three explanatory
compensation for risk, given some observed (and
                                                                   variables are used: real rate, inflation rate and short-term
forecasted) values of a set of macro-finance fundamentals.                                                                2
                                                                   forecasts. Nevertheless, the explanatory power ( R ) is
Suppose that the long-term rates go up. This could be due to
                                                                   strong enough to explain more than 90% of total variations in
the fact that market is expecting more hikes in the future, or
                                                                   market consensus. The residual and unexplained portion of
because the risk premia have gone up or some combination
                                                                   the expectations is relatively small for these three countries
of both. In order to gauge whether this increase in rates is
                                                                   and may be attributed to country-specific idiosyncratic factors
justifiable by the fundamentals, one needs to have a
                                                                   such as FX regimes, size and openness of the markets, EU
disciplined view of what is driving the move: premia or
                                                                   membership and convergence, and relevance of the local
expectations. In Latin American countries, it is possible to
                                                                   pension fund community.
distinguish the two by looking at survey data published by
central banks in the region. This allows us both to develop a      We then make the key assumption that such idiosyncratic
fair value model for risk premia and compare our economists’       factors also play a relatively small role for several other EM
expectation with consensus expectations. If either                 countries and apply the calibrated model using the explanatory
component has a significant divergence, then that may flag a       variables of Czech Republic, Hungary, Israel, Poland and
trade opportunity.                                                 South Africa. Hence, this sub-model builds a time series of
                                                                   monetary policy expectations for these countries.
In CEEMEA countries, the situation is more complicated as
the high-quality surveys with sufficient history and forecast      As explained in our previous publications (see EM Profile:
horizon are not available. However, we believe that we have        How to Value Brazil Rates? July 29, 2010, and EM Profile:
succeeded in circumventing this problem with the aid of            How to Value Latin America Rates? December 2, 2010,
LatAm surveys. Below we describe how our method works.             where our risk premia model is elaborated, the historical risk
                                                                   premia can then be computed by simply subtracting policy
We will start publishing the results of our models under
                                                                   expectations for the market curve. Thereafter, these can be
different scenarios in future issues of EMQSU. We welcome
                                                                   regressed over CDS and other explanatory variables to
feedback from our readers on what kinds of scenarios
                                                                   obtain a measure of fair risk premia.

                                                                                                                             117
                                                                                                  MORGAN STANLEY RESEARCH

                                                                                                  March, 2012
                                                                                                  EM Local Markets Guide




Exhibit 206                                                                                        Exhibit 208
Policy Rate and Market Consensus in Brazil                                                         2y Expectation and Model Prediction in Brazil
21                                                                                                 20


19
                                                                                                   18

17

                                                                                                   16
15

                                                                                                   14
13


11                                                                                                 12


 9
                                                                                                   10

 7
 Jul-04    Dec-05      May-07       Oct-08       Mar-10    Aug-11   Jan-13   Jun-14      Nov-15
                                                                                                    8
Source: Banco Central do Brasil, Morgan Stanley Research.                                           Aug-04       Aug-05           Aug-06             Aug-07              Aug-08   Aug-09   Aug-10   Aug-11

                                                                                                                    Policy Rate      2y Expectation          Predicted

Exhibit 207                                                                                        Source: Banco Central do Brasil, Morgan Stanley Research.
Policy Rate and Market Consensus in Chile
                                                                                                   Exhibit 209
10
                                                                                                   2y Expectation and Model Prediction in Chile
 8
                                                                                                   8



 6

                                                                                                   6

 4



                                                                                                   4
 2




 0
 Jul-04       Dec-05       May-07            Oct-08       Mar-10    Aug-11      Jan-13             2



Source: Central Bank of Chile, Morgan Stanley Research.
                                2                                                                  0
Besides the high R , the model also explains well some                                             Aug-04     Aug-05              Aug-06            Aug-07               Aug-08   Aug-09   Aug-10   Aug-11

systematic bias we (surprisingly) observed in the data.                                                                Policy Rate         2y Expectation      Predicted


Focusing on Brazil and Chile for a moment, consensus                                               Source: Central Bank of Chile, Morgan Stanley Research.

numbers expect the Central Bank of Chile to raise policy                                           Applying the Model to CEEMEA
rates and Brazil to do the opposite most of time.                                                  We use real rates, inflation rates and (Morgan Stanley) short-
Exhibit 206 and Exhibit 207 illustrate this systematic bias; the                                   term interest rate forecasts for Czech Republic, Hungary,
thick line represents the evolution of the target policy rate                                      Israel, Poland and South Africa, and insert them into the
and each thin ‘hairline’ describes market consensus for the                                        model (that was calibrated using data from Brazil, Chile and
following years.                                                                                   Mexico). We summarize the results and report rate
                                                                                                   expectations and risk premia in Exhibit 210 and Exhibit 211.
Exhibit 208 and Exhibit 209 compare model prediction with                                          Comparing 5y spot risk premia across countries, it is
2y market consensus data. Despite its simplicity, our model                                        probably not surprising that they are widest in Hungary,
matches market expectation extraordinarily well. It is also                                        Poland and South Africa and tightest in Czech Republic and
encouraging that one model can manage two different                                                Israel. The general shape of risk premia is U-shaped,
countries whose monetary policies expectations are at                                              reflecting increased uncertainty about near-term economic
opposite ends of the spectrum.                                                                     conditions and policy response.




                                                                                                                                                                                                       118
                                                                                                MORGAN STANLEY RESEARCH

                                                                                                March, 2012
                                                                                                EM Local Markets Guide




Exhibit 210                                                                                      Exhibit 213
Model-Generated Policy Rate Expectations                                                         5y Risk Premia in MEA
                      9m           1y             1.5y       2y          3y               5y      2.5
Czech Republic        1.15        1.62            2.41      2.83        2.72             1.47
Hungary               6.70        6.74            6.77      6.71        6.37             5.58     2.0
Israel                2.76        3.11            3.71      4.01        3.86             2.75
Poland                4.33        4.57            4.96      5.05        4.43             2.31     1.5

South Africa          5.85        6.21            6.78      7.04        6.93             6.20
                                                                                                  1.0
Source: Morgan Stanley Research, Bloomberg.
                                                                                                  0.5
Exhibit 211
Model-Generated Spot Risk Premia                                                                  0.0

                       9m           1y             1.5y       2y          3y           5y        -0.5
Czech Republic        70bp        51bp             7bp      -28bp       -56bp        -24bp
Hungary               74bp        74bp             66bp      59bp        59bp         91bp       -1.0
Israel                4bp         -8bp            -33bp     -52bp       -57bp         7bp
Poland                66bp        59bp             39bp      23bp        17bp         69bp       -1.5
South Africa          17bp         9bp             -7bp     -17bp       -10bp         56bp
Source: Morgan Stanley Research, Bloomberg.                                                      -2.0
                                                                                                    Jul-05       Jul-06               Jul-07       Jul-08            Jul-09            Jul-10       Jul-11

Exhibit 212 and Exhibit 213 display time series of 5y risk                                                     Israel     South Africa

                                                                                                 Source: Morgan Stanley Research, Bloomberg.
premia in all CEEMEA countries we are studying. It is worth
noting that in all cases we also observe that risk premia have                                   Exhibit 214
generally been positive and fairly mean reverting. Also, both                                    Chile vs. Czech and Israel 5y Spot Risk Premium
Poland and Hungary display big run-ups in risk premia                                             2.5

recently as the economic uncertainty in Europe has spread
                                                                                                  2.0
to CEEMEA.

On the other hand, it is interesting to note that this increased                                  1.5


uncertainty has pushed down risk premia in three countries
                                                                                                  1.0
considered to be relatively risk-free: Chile, Czech Republic,
and Israel. Hence, their risk premia have recently showed                                         0.5

similar trends as shown in Exhibit 214.
                                                                                                  0.0
Exhibit 212
5y Risk Premia in CEE                                                                            -0.5

 2.5
                                                                                                 -1.0
                                                                                                   May-06           May-07                May-08            May-09            May-10            May-11
 2.0                                                                                                                          Chile        Czech   Israel

                                                                                                 Source: Morgan Stanley Research, Bloomberg.
 1.5


                                                                                                 However, it is somewhat peculiar that the 5y risk premium in
 1.0
                                                                                                 Czech Republic is below both Chile and Israel. As we
 0.5                                                                                             emphasized in GEM Credit Trade Idea: Buy Czech 5Y CDS
                                                                                                 (November 9, 2011), markets regard Czech Republic to be
 0.0
                                                                                                 relatively safer than its peers, despite its deep economic and
                                                                                                 financial linkage with Europe at the moment. As a matter of
-0.5
                                                                                                 fact (see page 122), we think this less than justified risk
-1.0                                                                                             premium warrants a Czech 2s10s steepener.
  May-06         May-07         May-08             May-09      May-10           May-11
                  Czech Rep   Hungary    Poland

Source: Morgan Stanley Research, Bloomberg.




                                                                                                                                                                                                         119
                                                                          MORGAN STANLEY RESEARCH

                                                                          March, 2012
                                                                          EM Local Markets Guide




Historical Back-Test                                                       Exhibit 216
In Exhibit 215-Exhibit 219, we display the back-testing                    Back-Testing for 10y Hungary Rate – Number of
results for Czech Republic, Hungary, Israel, Poland, and                   Correct Predictions = 9, Number of Incorrect
South Africa. Since typically risk premium has the largest                 Number = 4
effect on the 10y sector valuations (short rates are driven by                                                      Predicted   Actual
monetary policy expectations and the longer term rates are                       Date        Market          Fair    Change     Change
heavily influenced by convexity bias), we are displaying the                   31-Oct-10      6.42           6.74       39
                                                                              30-Nov-10       7.17           7.48      32          75    Correct
results for 10y swap rates in each country. The future                        31-Dec-10       7.18           7.48      31          1          -
monetary policy path is taken from Morgan Stanley                             31-Jan-11       6.99           7.37      30         -19    Incorrect
economists’ forecasts. The testing period varies somewhat                     28-Feb-11       6.82           7.18       37        -17    Incorrect
depending on the availability of data needed for the model.                   31-Mar-11       6.84           7.14       36          2    Correct
                                                                               30-Apr-11      6.64           6.97      30         -21    Incorrect
The model has the following hit ratios for different countries:               31-May-11       6.68           7.01      33          4     Correct
57% for Czech Republic, 69% for Hungary, 50% for Israel,                      30-Jun-11       6.75           6.88      33          7     Correct
                                                                                31-Jul-11     6.92           6.55      12         17     Correct
60% for Poland, and 69% for South Africa. With the
                                                                              31-Aug-11       6.48           6.20      -36        -44    Correct
exception of Israel where the model has underperformed                        30-Sep-11       7.41           7.22      -27        93     Incorrect
since September 2010, we view these results as quite                           31-Oct-11      7.25           7.31      -18        -15    Correct
respectable.                                                                  30-Nov-11       7.37           7.64       6          12    Correct
                                                                              12-Dec-11       7.43           7.68       26          5    Correct
Finally, in Exhibit 220-Exhibit 224 we show the market and                 Source: Morgan Stanley Research

fair curves at the end of November together with the most                  Exhibit 217
recent curve. As can be observed from exhibits, the model                  Back-Testing for 10y Israel Rate – Number of
did not perform too well in Czech Republic (at least so far),              Correct Predictions = 10, Number of Incorrect
did an outstanding job in South Africa, a good job in Hungary              Number = 10
while Israel and Poland curves have not really moved much.                                                          Predicted   Actual
                                                                                Date         Market          Fair    Change     Change
We view the back-testing results as quite promising. They
                                                                             31-Dec-09        5.55           5.45
give us confidence that this model can be applied for real                   31-Jan-10        5.37           5.30       -9       -18     Correct
world strategy. Of course, at the end of the day, one has to                 28-Feb-10        5.12           5.01       -7       -25     Correct
always reserve the ultimate judgment on trade                                31-Mar-10        5.12           4.88      -11         0          -
recommendations.                                                              30-Apr-10       5.02           4.88      -24       -10     Correct
                                                                             31-May-10        4.81           4.64      -15       -21     Correct
Exhibit 215                                                                  30-Jun-10        4.67           4.61      -17       -14     Correct
                                                                               31-Jul-10      4.48           4.44       -6       -19     Correct
Back-Testing for 10y Czech Rate – Number of
                                                                             31-Aug-10        4.21           4.00       -4       -27     Correct
Correct Predictions = 8, Number of Incorrect                                 30-Sep-10        4.42           4.32      -21        21     Incorrect
Predictions = 6                                                               31-Oct-10       4.49           4.40      -11         7     Incorrect
                                         Predicted   Actual                  30-Nov-10        4.61           4.52      -10        11     Incorrect
     Date         Market          Fair    Change     Change                  31-Dec-10        4.81           4.66       -9        21     Incorrect
   31-Oct-10       2.77           2.79                                       31-Jan-11        5.32           5.22      -15        51     Incorrect
  30-Nov-10        2.98           3.09       2        21      Correct        28-Feb-11        5.37           5.38       -9         5     Incorrect
  31-Dec-10        3.06           3.31      11         8      Correct        31-Mar-11        5.36           5.59        1        -1          -
  31-Jan-11        3.30           3.47      25        24      Correct         30-Apr-11       5.35           5.34      23         -1     Incorrect
  28-Feb-11        3.20           3.41       17       -10     Incorrect      31-May-11        5.14           5.20       -1       -21          -
  31-Mar-11        3.37           3.39      21        17      Correct        30-Jun-11        5.24           5.19        7        11     Correct
   30-Apr-11       3.23           3.32       2        -14     Incorrect        31-Jul-11      5.11           4.73       -5       -13     Correct
  31-May-11        3.06           3.21       9        -17     Incorrect      31-Aug-11        4.42           4.48      -39       -70     Correct
  30-Jun-11        3.09           3.18      15         3      Correct        30-Sep-11        4.39           4.24        7        -2     Incorrect
    31-Jul-11      2.76           2.42       9        -33     Incorrect       31-Oct-11       4.48           4.30      -16         9     Incorrect
  31-Aug-11        2.42           2.61      -34       -34     Correct        30-Nov-11        4.47           4.56      -18        -1          -
  30-Sep-11        2.16           2.47       19       -26     Incorrect      12-Dec-11        4.46           4.39        9        -1     Incorrect
                                                                           Source: Morgan Stanley Research
   31-Oct-11       2.28           2.46      31        12      Correct
  30-Nov-11        2.38           2.61      18        10      Correct
  12-Dec-11        2.29           2.42       23        -9     Incorrect
Source: Morgan Stanley Research




                                                                                                                                              120
                                                                          MORGAN STANLEY RESEARCH

                                                                          March, 2012
                                                                          EM Local Markets Guide




Exhibit 218                                                                Exhibit 220
Back-Testing for 10y Poland Rate – Number of                               Czech Republic Rates Fair Value Assessments
Correct Predictions = 9, Number of Incorrect                               3.0
Number = 6
                                         Predicted   Actual                2.5
     Date         Market          Fair    Change     Change
  31-Aug-10        4.90           5.05
                                                                           2.0
  30-Sep-10        4.91           5.05      15         1      Correct
   31-Oct-10       5.19           5.24      14         28     Correct
  30-Nov-10        5.53           5.58        5        33     Correct      1.5
  31-Dec-10        5.63           5.71       6         11     Correct
  31-Jan-11        5.76           5.89        8        13     Correct      1.0
  28-Feb-11        5.60           5.71      13        -17     Incorrect
  31-Mar-11        5.70           5.74      11         11     Correct
   30-Apr-11       5.67           5.88       4         -3     Incorrect    0.5
                                                                                 0       2     4        6      8        10     12      14     16        18    20
  31-May-11        5.50           5.65      21        -17     Incorrect
                                                                                                   31-Nov Market         31-Nov Fair        12-Dec Market
  30-Jun-11        5.37           5.56      14        -13     Incorrect
    31-Jul-11      5.37           5.36      19          0          -       Source: Morgan Stanley Research, Bloomberg

  31-Aug-11        4.82           4.64       -1       -55     Correct
                                                                           Exhibit 221
  30-Sep-11        5.01           4.85      -18        19     Incorrect
   31-Oct-11       4.94           4.78      -16        -7     Correct      Hungary Rates Fair Value Assessments
  30-Nov-11        5.09           4.98      -16        15     Incorrect    8.0
  12-Dec-11        5.03           4.85      -11        -6     Correct
Source: Morgan Stanley Research
                                                                           7.5
Exhibit 219
Back-Testing for 10y South Africa Rate – Number
of Correct Predictions = 11, Number of Incorrect                           7.0

Number = 5
                                         Predicted   Actual                6.5
     Date         Market          Fair    Change     Change
  31-Aug-10        7.38           7.50
  30-Sep-10        7.49           7.44      12         11     Correct      6.0
   31-Oct-10       7.25           7.46       -5       -24     Correct            0       2     4        6      8        10     12      14     16        18    20
  30-Nov-10        7.84           8.03      21         59     Correct                         31-Nov Market        31-Nov Fair          12-Dec Market
  31-Dec-10        7.77           8.05       19        -7     Incorrect    Source: Morgan Stanley Research, Bloomberg
  31-Jan-11        8.57           8.78       28        80     Correct
  28-Feb-11        8.37           8.67       21       -20     Incorrect    Exhibit 222
  31-Mar-11        8.44           8.71      30         7      Correct      Israel Rates Fair Value Assessments
   30-Apr-11       8.15           8.37      27        -29     Incorrect
                                                                           6.0
  31-May-11        8.07           8.33       22        -8     Incorrect
  30-Jun-11        8.16           8.37       26         8     Correct
    31-Jul-11      7.97           7.91       22       -19     Incorrect
  31-Aug-11        7.20           7.26       -6       -77     Correct      5.0
  30-Sep-11        7.87           7.68       6         67     Correct
   31-Oct-11       7.51           7.67      -19       -36     Correct
  30-Nov-11        7.56           7.78      16         4      Correct      4.0
  12-Dec-11        7.67           7.80       23        12     Correct
Source: Morgan Stanley Research

                                                                           3.0



                                                                           2.0
                                                                                 0       2     4        6      8        10     12      14     16        18    20
                                                                                             31-Nov Market          31-Nov Fair             12-Dec Market
                                                                           Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                             121
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




Exhibit 223                                                                            Current Fair-Value Estimates
Poland Rates Fair Value Assessments                                                    Exhibit 225-Exhibit 234 display the current market prices and
5.5                                                                                    the model-generated fair value assessments under two
                                                                                       different assumptions. For each country, the first table shows
                                                                                       what happens if we use model-implied expectations for
5.2
                                                                                       future monetary policy and the second table uses instead the
                                                                                       current Morgan Stanley CEEMEA economists’ forecasts. It is
4.9
                                                                                       important to note we can use the model with any monetary
                                                                                       policy path of our choosing. Therefore, it is a valuable tool for
4.6                                                                                    scenario analysis.

                                                                                       If we compare market pricing and fair value estimates in
4.3
      0       2     4      6         8       10      12    14       16      18   20
                                                                                       Exhibit 225-Exhibit 234, at least the following trade ideas are
                     31-Nov Market           31-Nov Fair        12-Dec Market          suggested by both scenarios: (i) Czech 2s10s steepener
Source: Morgan Stanley Research, Bloomberg                                             which is also consistent with our view that the risk premium
                                                                                       in Czech Republic is too low (see GEM Credit Trade Idea:
Exhibit 224
                                                                                       Buy Czech 5Y CDS, November 9, and p. 41 in Global EM
South Africa Rates Fair Value Assessments
                                                                                       Investor: Trading Funding Stresses, Printing Presses,
8.5                                                                                    December 8); and (ii) paying Hungary rates which is also
                                                                                       consistent with our negative view Hungary rates as
7.5                                                                                    expressed in Global EM Investor: Trading Funding Stresses,
                                                                                       Printing Presses (December 8, see p. 40). In addition, given
                                                                                       the Morgan Stanley call of two 25bp cuts in Poland and one
6.5
                                                                                       25bp cut in Israel next year, (iii) receiving 2y Poland and (iv)
                                                                                       Israel rates would make sense as well.
5.5
                                                                                       Future Work
                                                                                       The logical next step is to see how well the model would
4.5                                                                                    work for AXJ rates. In principle, the same procedure can be
      0       2     4       6        8       10      12    14       16      18   20
                  31-Nov Market          31-Nov Fair            12-Dec Market          applied in Asia as we did for CEEMEA here. In practice, AXJ
Source: Morgan Stanley Research, Bloomberg                                             rates probably have some idiosyncrasies (such as following
                                                                                       closely US rates) which may require additional submodels.




                                                                                                                                                   122
                                                                        MORGAN STANLEY RESEARCH

                                                                        March, 2012
                                                                        EM Local Markets Guide




Exhibit 225                                                              Exhibit 226
Czech Republic Rates Fair Value Assessments                              Czech Republic Rates Fair Value Assessments
under Model-generated Expectations                                       under Morgan Stanley Forecasts (no cuts)
 Maturity       Market           Fair        Spread     Cheap/Rich        Maturity       Market           Fair        Spread     Cheap/Rich
   3M            1.16            1.35           19         Rich     ▲       3M            1.16            1.35           19         Rich     ▲
   6M            1.46            1.54           8     Fairly Valued        6M            1.46            1.54           8     Fairly Valued 
   1Y            1.08            1.38           30         Rich     ▲       1Y            1.08            1.30           22         Rich     ▲
   2Y            1.40            1.27          -13    Fairly Valued        2Y            1.40            0.98          -42        Cheap     ▼
   3Y            1.51            1.50           -1    Fairly Valued        3Y            1.51            1.32          -19        Cheap     ▼
   4Y            1.63            1.71           8     Fairly Valued        4Y            1.63            1.61           -2    Fairly Valued 
   5Y            1.76            1.91           15         Rich     ▲       5Y            1.76            1.84           8     Fairly Valued 
   7Y            2.02            2.27           25         Rich     ▲       7Y            2.02            2.21           19         Rich     ▲
  10Y            2.29            2.46           17         Rich     ▲      10Y            2.29            2.42           13    Fairly Valued 
  15Y            2.59            2.69           10    Fairly Valued       15Y            2.59            2.66           7     Fairly Valued 
  20Y            2.64            2.82           18         Rich     ▲      20Y            2.64            2.79           15         Rich     ▲
Source: Morgan Stanley Research, Bloomberg                               Source: Morgan Stanley Research, Bloomberg

Exhibit 227                                                              Exhibit 228
Hungary Rates Fair Value Assessments under                               Hungary Rates Fair Value Assessments under
Model-generated Expectations                                             Morgan Stanley Forecasts (5*25bps hikes, followed
 Maturity       Market           Fair        Spread    Cheap/Rich        by 5*25bps cuts)
   3M            7.00            6.83          -17      Cheap     ▼       Maturity       Market           Fair        Spread    Cheap/Rich
   6M            7.24            7.56           32       Rich     ▲         3M            7.00            7.48          48       Rich      ▲
   1Y            7.47            7.65           18       Rich     ▲         6M            7.24            8.13          89       Rich      ▲
   2Y            7.37            7.62           25       Rich     ▲         1Y            7.47            8.25          77       Rich      ▲
   3Y            7.24            7.68           44       Rich     ▲         2Y            7.37            7.92          55       Rich      ▲
   4Y            7.23            7.67           44       Rich     ▲         3Y            7.24            7.93          69       Rich      ▲
   5Y            7.27            7.69           42       Rich     ▲         4Y            7.23            7.90          68       Rich      ▲
   7Y            7.40            7.72           32       Rich     ▲         5Y            7.27            7.88          61       Rich      ▲
  10Y            7.43            7.63           20       Rich     ▲         7Y            7.40            7.79          39       Rich      ▲
  15Y            7.23            7.57           34       Rich     ▲        10Y            7.43            7.68          26       Rich      ▲
  20Y            6.94            7.54           60       Rich     ▲        15Y            7.23            7.61          38       Rich      ▲
Source: Morgan Stanley Research, Bloomberg                                 20Y            6.94            7.58          64       Rich      ▲
                                                                         Source: Morgan Stanley Research, Bloomberg

Exhibit 229                                                              Exhibit 230
Israel Rates Fair Value Assessments under Model-                         Israel Rates Fair Value Assessments under Morgan
generated Expectations                                                   Stanley Forecasts (one 25bps cut)
 Maturity       Market           Fair        Spread     Cheap/Rich        Maturity       Market           Fair        Spread     Cheap/Rich
   3M            2.64            2.85           21         Rich     ▲       3M            2.64            2.81           17         Rich     ▲
   6M            2.57            2.76           19         Rich     ▲       6M            2.57            2.74           17         Rich     ▲
   1Y            2.54            2.60           6     Fairly Valued        1Y            2.54            2.52           -2    Fairly Valued 
   2Y            2.68            2.61           -7    Fairly Valued        2Y            2.68            2.39          -29        Cheap     ▼
   3Y            2.87            2.92           5     Fairly Valued        3Y            2.87            2.77          -10    Fairly Valued 
   4Y            3.13            3.21           8     Fairly Valued        4Y            3.13            3.13           0     Fairly Valued 
   5Y            3.41            3.51           9     Fairly Valued        5Y            3.41            3.44           3     Fairly Valued 
   7Y            3.98            4.03           5     Fairly Valued        7Y            3.98            3.98           0     Fairly Valued 
  10Y            4.46            4.43           -3    Fairly Valued       10Y            4.46            4.39           -7    Fairly Valued 
  15Y            4.91            4.80          -11    Fairly Valued       15Y            4.91            4.78          -13    Fairly Valued 
  20Y            5.07            4.99           -9    Fairly Valued       20Y            5.07            4.97          -11    Fairly Valued 
Source: Morgan Stanley Research, Bloomberg                               Source: Morgan Stanley Research, Bloomberg




                                                                                                                                          123
                                                                        MORGAN STANLEY RESEARCH

                                                                        March, 2012
                                                                        EM Local Markets Guide




Exhibit 231                                                              Exhibit 232
Poland Rates Fair Value Assessments under                                Poland Rates Fair Value Assessments under
Model-generated Expectations                                             Morgan Stanley Forecasts (two 25bps cuts)
 Maturity       Market           Fair        Spread     Cheap/Rich        Maturity       Market           Fair        Spread     Cheap/Rich
   3M            4.88            4.85          -3     Fairly Valued        3M            4.88            4.84           -4    Fairly Valued 
   6M            4.89            5.12          23          Rich     ▲       6M            4.89            4.99           10    Fairly Valued 
   1Y            4.94            4.99           5     Fairly Valued        1Y            4.94            4.79          -15        Cheap     ▼
   2Y            4.82            4.82           1     Fairly Valued        2Y            4.82            4.54          -28        Cheap     ▼
   3Y            4.81            4.96          15          Rich     ▲       3Y            4.81            4.78           -3    Fairly Valued 
   4Y            4.84            5.00          16          Rich     ▲       4Y            4.84            4.91           7     Fairly Valued 
   5Y            4.89            5.05          16          Rich     ▲       5Y            4.89            4.97           8     Fairly Valued 
   7Y            4.99            5.17          18          Rich     ▲       7Y            4.99            4.88          -11    Fairly Valued 
  10Y            5.04            5.07           3     Fairly Valued       10Y            5.04            4.85          -19        Cheap     ▼
  15Y            4.86            5.13          27          Rich     ▲      15Y            4.86            4.98           12    Fairly Valued 
  20Y            4.63            5.18          55          Rich     ▲      20Y            4.63            5.05           42         Rich     ▲
Source: Morgan Stanley Research, Bloomberg                               Source: Morgan Stanley Research, Bloomberg

Exhibit 233                                                              Exhibit 234
South Africa Rates Fair Value Assessments under                          South Africa Rates Fair Value Assessments under
Model-generated Expectations                                             Morgan Stanley Forecasts (no cuts)
 Maturity       Market           Fair        Spread     Cheap/Rich        Maturity       Market           Fair        Spread     Cheap/Rich
   3M            5.58            5.68          10     Fairly Valued        3M            5.58            5.68          10     Fairly Valued 
   6M            5.82            5.78          -4     Fairly Valued        6M            5.82            5.78          -4     Fairly Valued 
   1Y            5.56            5.81          25          Rich     ▲       1Y            5.56            5.73          17          Rich     ▲
   2Y            5.82            6.05          22          Rich     ▲       2Y            5.82            5.79          -3     Fairly Valued 
   3Y            6.27            6.45          19          Rich     ▲       3Y            6.27            6.31           5     Fairly Valued 
   4Y            6.60            6.79          19          Rich     ▲       4Y            6.60            6.70          10     Fairly Valued 
   5Y            6.89            7.10          21          Rich     ▲       5Y            6.89            7.03          14     Fairly Valued 
   7Y            7.36            7.56          20          Rich     ▲       7Y            7.36            7.45           9     Fairly Valued 
  10Y            7.67            7.88          21          Rich     ▲      10Y            7.67            7.80          13     Fairly Valued 
  15Y            7.77            8.13          36          Rich     ▲      15Y            7.77            8.06          29          Rich     ▲
  20Y            7.77            8.24          48          Rich     ▲      20Y            7.77            8.18          42          Rich     ▲
Source: Morgan Stanley Research, Bloomberg                               Source: Morgan Stanley Research, Bloomberg




                                                                                                                                          124
                                                                          MORGAN STANLEY RESEARCH

                                                                          March, 2012
                                                                          EM Local Markets Guide




Appendix: The Expectation-Generation Model
We start from Taylor Rule, which is most widely used in the                Exhibit 235
literature of monetary policy. The rule can be written as                  Estimated Model Coefficients (standard errors)
                                                                                                   1y          18m         2y        3y        4y
             it   t  rt   ( t   * )   ( yt  y * )               Constant               0.28         0.79       1.27      1.00      0.63
                                                                                                 (0.063)      (0.056)    (0.068)   (0.268)   (0.309)
where   it is target policy rate,  t is inflation rate,  * is the        Real Rate              -0.09        -0.20      -0.30     -0.39     -0.41
                                                                                                 (0.011)      (0.009)    (0.010)   (0.031)   (0.035)
central bank’s inflation target, rt is equilibrium real rate, and
                                                                           Excess Inflaton        -0.14        -0.30      -0.43     -0.44     -0.50
yt  y* is an output gap.                                                                        (0.019)      (0.018)    (0.020)   (0.038)   (0.040)
                                                                           3m forecast            -1.05        -0.87      -0.74     -0.79     -0.63
However, the Taylor rule has several problems when using                                         (0.157)      (0.177)    (0.194)   (0.227)   (0.194)
directly to estimate policy rate expectation. First and                    6m forecast            1.79         1.48       1.20      0.91      0.74
foremost, the equation includes three unobservable                                               (0.106)      (0.124)    (0.142)   (0.174)   (0.148)


variables:  , t , and
                                                                           R2                      0.89        0.93       0.95      0.96        0.96
                 *   r         y*
                               . These are hard to estimate and            Source: Bloomberg, Morgan Stanley Research.

tend to change over time. Finally, the regression coefficient
on output gap (maybe due to estimation problems) typically                      Excess Inflation:
                                                                                                           t   *
is not statistically significant. Since our purpose is to find a
general pattern from Brazil/Chile/Mexico consensus data and                      o       The primary objective of monetary policy is to keep
apply it to other emerging market countries, we need to                                  inflation under control. Assuming that the inflation
remove all country-specific unobservable variables.                                      rate follows a mean-reverting process, we can
                                                                                         expect a temporary change of policy rate caused by
Thus, we derive our model by looking at the expected                                     fluctuating inflation to be cancelled out later.
change and adding more explanatory variables. We can                                     Therefore, the current policy rate should be
write our model as                                                                       negatively correlated with policy rate expectation.

             ite  it   0  1 (it   t )   2 ( t   * )                                                   ft 3m  it and ft 6m  it
                                                                                3m and 6m Forecasts:
                           3 ( ft 3m  it )   4 ( ft 6m  it )  òt
                                                                                 o       This variable is intended to capture a forward-
          e                                                                              looking component. Using 3m and 6m forecast
where
         i
         t    is market expectation of nominal policy rate in                          numbers has two benefits. First, these numbers
                               3m
                                                  f 6m                                   contain information most directly related to
 periods in the future and ft     and t are 3m and 6m                                   prospective future monetary policy. Second, they
forecasts of monetary policy rate. Our estimation is shown in                            are easily available for all emerging markets from,
Exhibit 235. The last two terms dominate short-term                                      say, Morgan Stanley publications.
expectations (  1 year), meanwhile the first two terms
                                                                           We tested the regressions by including a fixed-effect term for
turn out dominant for long-end expectations (   2 year).                 each country. The coefficients were robust in this alteration.
Explanatory power represented by R2 varies from 89% to                     Of course, we cannot include a country-specific fixed effect
96%, depending on expectation horizon. We explain                          in the final regression as we have no way of estimating that
rationales of each control variable below.                                 for CEEMEA countries.

    Real Rate: it   t                                                   Finally, we assume that the long-term expectations converge
                                                                           to a predetermined equilibrium short rate. The rate of
     o       This variable is supposed to capture the cross-               convergence and the equilibrium rate have been calibrated
             sectional difference between Brazil and Chile.                to each country separately.




                                                                                                                                                  125
                                                                                            MORGAN STANLEY RESEARCH

                                                                                            March, 2012
                                                                                            EM Local Markets Guide




A Guide to Latin America Interest Rate Options
Juha Seppala
                                    th
First Published November 4 , 2010: A Guide to Latin American Interest Rate Options

We present a brief guide for clients interested in options                                   Exhibit 236
over the local Brazilian interest rate DI curve and over                                     IDI Options and DI Swaptions
Mexican TIIE rates.
                                                                                                               IDI Option
In Brazil, two types of options are available: DI
swaptions and IDI options. A typical DI swaption would be
                                                                                                               3 months                 12 months
Jan 11 versus Jan12: an option to enter into the Jan12 DI
swap on the first business day of 2011. These options are
typically European type options (can only be exercised at                                      Today                           Jan-11               Jan-12
expiry date) and can be payer or receiver. The relevant
underlying rate for this derivative would be the (offshore)                                                   DI Swaption                FRA
Jan11 versus Jan12 forward rate (FRA). 25 This swaption
would be suitable for investors who have a view on what                                      Source: Morgan Stanley Research

shape the DI curve will have on January 2011 or what the
monetary policy path will be throughout 2011.
                                                                                             Brazil: DI Curve and DI Swaptions
                                                                                             Before describing and giving examples of DI swaption, it is
IDI options are very different instruments and unique to
                                                                                             useful to elaborate on the underlying assets.
Brazil’s local rates curve. This derivative is an option over
the accumulated overnight rate (CDI) until maturity. IDI                                     DI swaps are among the most liquid instruments in LatAm.
options are very suitable instruments for clients to trade pure                              They trade in the local Brazilian market as interest rate
monetary policy views when these differ from what is implied                                 futures settled through BM&F (Bolsa de Mercadorias e
by the DI curve.                                                                             Futuros). However, the payoff of DI futures mimics the payoff
                                                                                             of zero coupon swaps where the floating leg is the overnight
Turning to Mexico, aside from the payment convention of
                                                                                             CDI rate. The fact that they are traded as futures in the local
the underlying swaps and the resetting of the floating rates
                                                                                             market is only relevant in that the mark-to-market is done
leg (28 days instead of 6 months), swaptions, caps and
                                                                                             daily by the BM&F exchange, which mitigates counterparty
floors are very similar to the ones traded in more developed
                                                                                             risks somewhat. DI swaps are traded over-the-counter (OTC)
markets.
                                                                                             offshore as zero coupon swaps.
An example of a TIIE swaption would be a 3M10Y
                                                                                             Swaps from interest rate curves of other countries are
receiver (or payer), which gives the owner of the swaption
                                                                                             usually available in fixed tenors (1y, 5y, 10y, etc.). However
the right to receive (or pay) 10y TIIE fixed rate (the strike of
                                                                                             DI futures are available in fixed expiry dates (January 2011,
the swaption) on the maturity date of swaption, which in this
                                                                                             July 2011, January 2012, etc.).
example is 3m in the future.
                                                                                             A typical DI swaption would be Jan 11 versus Jan12, which
Caps and floors protect against movements in floating
                                                                                             is an option to enter into the Jan12 DI swap on the first
rate. For example, in a swap receiver position, an investor
                                                                                             business day of January 2011. Liquidity is concentrated on
receives the fixed rate and pays the floating rate. Caps limit
                                                                                             swaptions with strikes close to Jan11 versus Jan12 forward
each one of the payments (in the case of TIIE, payments are
                                                                                             rate (FRA), although in-the-money and out-of-the-money
every 28 days) to be no more than the rate predetermined by
                                                                                             (OTM) strikes are not uncommon.
the cap.




25
  Most of the examples in this paper are based on local markets. Investors should be
aware that occasionally there is an offshore-onshore basis as local BRL nominal rates are
derived form the DI curve and offshore BRL rates are related to the NDF implied yields.


                                                                                                                                                         126
                                                                                    MORGAN STANLEY RESEARCH

                                                                                    March, 2012
                                                                                    EM Local Markets Guide




Exhibit 237                                                                          contract, whereas in the offshore market, it is the offshore
DI Information                                                                       PRExDI swaps. These underlying instruments are usually
 Tenor                             Interest rate (%)                     Bus days    closely related but investors should be aware that
 Jan-11                                        10.67                           67    occasionally there are offshore-onshore basis differences
 Jan-12                                        11.53                          318    between the two. As the rates oscillate, the liquidity
Source: Bloomberg                                                                    gravitates towards the swaptions with the strike closest to the
                                                                                     current FRA level.
Using the numerical example from
                                                                                     Exhibit 240
                                                                                     DI Jan11 x Jan12 Swaptions

                                                                                      Type                        Strike          Contracts outstanding
Exhibit 237, the computation of the at-the-money forward
                                                                                      Call                       11.50%                         34,000
strike is                                                                             Call                       11.75%                         82,000
                                       252                                            Call                       12.00%                         87,000
                                                                                      Call                       12.50%                        138,500
      
        1  11.53% 252 
                     318 318  67
                                                                                      Call                       13.50%                         91,000
FRA                             1  FRA  11.76%                                  Put                        12.00%                          7,150
       1  10.67% 252 
                     67
                                                                                      Put                        11.75%                         31,000
                                                                                    Put                        11.50%                         97,800
                                                                                      Put                        11.25%                         74,500
                                                                                     Source: BMF&BOVESPA
Exhibit 238                                                                          In the OTC offshore market, the notional, the tenor, the
Forward Rate Diagram                                                                 expiry, and the strikes of the DI swaptions can be tailored to
                                                                                     client requests, but liquidity and bid-ask spreads are
                     3 months                            12 months
                                                                                     generally better for strikes and maturity that are similar to the
                                                                                     ones found in the onshore market.

    Today                                  Jan-11                        Jan-12      Payers and Receivers
                                                                                     DI swaptions can be payers (call options on interest rates) or
                                                          FRA                        receivers (put option on interest rates).
Source: Morgan Stanley
In this example, an at-the-money-forward DI swaption would
                                                                                     Continuing the example above, the purchase of a Jan11
have a strike associated with the FRA level.
                                                                                     versus Jan12 payer DI swaption with strike 11.75% (hence
Exhibit 239                                                                          slightly ITM) would give the holder of the option the right to
FRA DI Jan’11 x Jan’12                                                               enter into a payer exposure in the Jan12 DI swap at the
                                                                                     swaption maturity, which is the first business day of January
  14.00%                                                                             2011.
  13.50%                                                                             Just as an example, assume the trajectory of the Jan12 DI
  13.00%
                                                                                     rate follows what is currently priced into the forward curve
                                                                                     and achieves 11.90% at the swaption maturity date in early
  12.50%
                                                                                     January 2011. In this case, payer DI swaptions with a strike
  12.00%                                                                             of 12.00% or higher would expire worthless. However, the
  11.50%
                                                                                     call option with a strike of 11.75% would have a final intrinsic
                                                                                     value as descried below.
  11.00%
       Jan-09   Apr-09   Jul-09   Oct-09     Feb-10    May-10   Aug-10     Dec-10

Source: Bloomberg                                                                    Payoff of Call DI Swaptions
DI swaptions trade in the local markets through the BM&F
                                                                                     The payoff of each contract of a call DI swaption at the
exchange with a wide range of standardized size strikes
                                                                                     expiration date is given by:
(each swap has notional of R$ 100k) in increments of 25bp.
In the local market, the underlying instrument is the DI future

                                                                                                                                                   127
                                                                                                                          MORGAN STANLEY RESEARCH

                                                                                                                          March, 2012
                                                                                                                          EM Local Markets Guide




                                                                                                                           However, the receiver with a strike of 12.00% would have
                                                               du Long  du Short            du Long  du Short
                                                                                                                    
                                                 max  1  it  252                1  ik  252
                      Notional
Payoff                     du Long  du Short
                                                                                                                 ,0       payout
                1  iT          252
                                                                                                                   
                                                                                                                                                                               du Long  du Short            du Long  du Short
                                                                                                                                                                                                                                    
                                                                                                                                                                 max  1  ik  252                1  it  252
                                                                                                                                          Notional
                                                                                                                             Payoff                                                                                             ,0 
iT : Offshore Pre x CDI rate associated to the swap                                                                                           du Long  du Short
                                                                                                                                                                     
                                                                                                                           iT : Offshore iT  associated to the option expiration date.                                            
expiration date.
                                                                                                                                      1 rate 252

                                                                                                                                 12.00% , it  11.90% , iT  8.00%
                                                                                                                           With ik                                                                                 ,
To illustrate the payoff, the above numerical example gives
                                                                                                                           Notional  1,000,000.00 BRL
ik  11.75% , it  11.90% , iT  8.00%                                                      ,

Notional  1,000,000.00 BRL                                                                                                Payoff 
                                                                                                                                                 1,000,000                                318  67            318  67
                                                                                                                                                                        max  1  12.00%  252  1  11.90%  252 ,0 
                                                                                                                                                                                                                        
                                                                                                                                                             318  67
                                                                                                                                              1  8.00%      252
                                                                                                                                                                                                                       
                                                                    318  67            318  67
                                                                                                  
                                                  max  1  11.90%  252  1  11.75%  252 ,0 
                      1,000,000
Payoff                               318  67
                                                                                                                         Payoff  922 BRL
                   1  8.00%          252

                                                                                                                           Exhibit 242
                                                                                                                           Receiver Swaptions (Put Option)
Payoff  1,384.00 BRL
If we assume different end value rates (on January 2011) for                                                                      BRL 150.00
                                                                                                                                                           Theoretical Prices          Strike         FRA              Intrisic Value
the Jan12 DI contract, the call payoff for different rates can                                                                    BRL 100.00

be seen in Exhibit 241.
                                                                                                                                   BRL 50.00
                                                                                                                            P&L




Exhibit 241
Payer Swaptions (Call Option)                                                                                                       BRL 0.00


                                                                                                                                  -BRL 50.00

       BRL 150.00
                               Theoretical Prices                  Strike          FRA            Intrisic Value               -BRL 100.00

       BRL 100.00
                                                                                                                               -BRL 150.00
                                                                                                                                         0%




                                                                                                                                                         0%




                                                                                                                                                                             0%




                                                                                                                                                                                             0%




                                                                                                                                                                                                              0%




                                                                                                                                                                                                                                  0%
        BRL 50.00
                                                                                                                                       .5




                                                                                                                                                       .6




                                                                                                                                                                           .7




                                                                                                                                                                                           .8




                                                                                                                                                                                                            .9




                                                                                                                                                                                                                                .0
 P&L




                                                                                                                                      11




                                                                                                                                                      11




                                                                                                                                                                          11




                                                                                                                                                                                         11




                                                                                                                                                                                                           11




                                                                                                                                                                                                                               12
         BRL 0.00
                                                                                                                                                                        DI Jan12 at January 2011
                                                                                                                           Source: Morgan Stanley
       -BRL 50.00                                                                                                          The put payoff for different end DI Jan12 rates can be seen
                                                                                                                           in Exhibit 242.
   -BRL 100.00


   -BRL 150.00
                                                                                                                           Other Option Strategies
                                                                                                                           Investors may use a combination of swaptions to achieve
              0%




                                   0%




                                                          0%




                                                                              0%




                                                                                                  0%




                                                                                                                     0%
            .5




                                 .6




                                                        .7




                                                                            .8




                                                                                                .9




                                                                                                                   .0
           11




                               11




                                                      11




                                                                         11




                                                                                             11




                                                                                                                12




                                                   DI Jan12 at January 2011                                                specific payoffs and often lower the cost of the option
Source: Morgan Stanley                                                                                                     structure.
Let’s further assume that the cost of this call DI swaption had
been 1000 BRL and the USD/BRL FX spot rates were 1.75                                                                      For example, if the investor has a view that the interest rate
and 1.80 at the date of purchase and at maturity of the                                                                    is going to rise modestly, one simple way to open a position
swaption, respectively. In this case, the return on the                                                                    in the interest rates market is to buy a call swaption.
investment would have been:                                                                                                However, a call spread where the investor purchases an
                                                                                                                           ATM payer and sells an OTM payer (see Exhibit 243) will
 Payoff / 1 . 80  1000 / 1 . 75                                                                                           require less upfront cash might achieve better risk-rewards.
                                  35 %
        1000 / 1 . 75                                                                                                      A butterfly (see Exhibit 245), might further reduce the cost (to
                                                                                                                           zero or even negative) but at the risk of losing money in the
Payoff of Put DI Swaption                                                                                                  scenario where rates increase substantially.

In the above numerical example, all the put DI swaptions
with strikes below 11.90% would expire out of the money.


                                                                                                                                                                                                                             128
                                                                                                 MORGAN STANLEY RESEARCH

                                                                                                 March, 2012
                                                                                                 EM Local Markets Guide




Exhibit 243                                                                                       Exhibit 245
Payer/Call Spread                                                                                 Butterfly Theoretical Prices

   BRL 200.00                                                                                        BRL 200.00

                         Theoretical Prices            Intrisic Value                                                      Theoretical Prices      Intrisic Value
   BRL 150.00                                                                                        BRL 150.00


   BRL 100.00                                                                                        BRL 100.00
 P&L




                                                                                                   P&L
       BRL 50.00                                                                                         BRL 50.00


        BRL 0.00                                                                                          BRL 0.00


       -BRL 50.00                                                                                        -BRL 50.00


   -BRL 100.00                                                                                       -BRL 100.00
                5%




                                  5%




                                                                 5%




                                                                                            5%




                                                                                                                  5%




                                                                                                                                5%




                                                                                                                                                     5%




                                                                                                                                                                         5%
              .2




                                .7




                                                               .2




                                                                                          .7




                                                                                                                .2




                                                                                                                              .7




                                                                                                                                                   .2




                                                                                                                                                                       .7
           11




                             11




                                                            12




                                                                                       12




                                                                                                             11




                                                                                                                           11




                                                                                                                                                12




                                                                                                                                                                    12
                              DI Jan12 at January 2011                                                                      DI Jan12 at January 2011
Source: Morgan Stanley                                                                            Source: Morgan Stanley

Exhibit 244
                                                                                                  Trading, Market Size and Liquidity
Receiver/Put Spread
                                                                                                  DI swaptions are flexible instruments that allow investors to
   BRL 150.00
                                                                                                  hedge or take outright exposure to Brazil’s interest rate
                                       Theoretical Prices             Intrisic Value
                                                                                                  market with limited downside risk. This derivative is a
   BRL 100.00
                                                                                                  European option and does not depend only on the monetary
       BRL 50.00
                                                                                                  policy decisions taken during the swaption’s life, but also on
 P&L




                                                                                                  the different scenarios of Copom (Comitê de Política
        BRL 0.00                                                                                  Monetária) expectations after the maturity of the swaption.
                                                                                                  The IDI option, which we will describe later in this document,
       -BRL 50.00
                                                                                                  only depends on monetary policy decisions taken up to the
                                                                                                  maturity of the option.
   -BRL 100.00

                                                                                                  DI swaptions are materially less liquid than the underlying DI
                5%




                                  5%




                                                                 5%




                                                                                            5%
              .2




                                .7




                                                               .2




                                                                                          .7
           11




                             11




                                                            12




                                                                                       12




                              DI Jan12 at January 2011                                            swaps, which are among the most liquid EM assets.
Source: Morgan Stanley                                                                            Nonetheless, liquidity in DI swaptions is not poor and has
On the other hand, put spreads (see Exhibit 244) may be                                           been improving. Exhibit 246 shows that the accumulated
appealing for investors who believe that rates are biased for                                     number of DI future contracts traded up to July since January
a modest decline. Also, selling a call spread should be                                           2010 was about 5 million contracts (each contract has about
considered by investors who believe that rates increases are                                      BRL100k of notional). A typical day would have liquidity of
extremely unlikely. Finally, selling straddles (sell ATM payer                                    around 50k contracts.
and ATM receiver) would profit in scenario of range-bound                                         DI swaptions can have several combinations of expiry and DI
rates.                                                                                            tenors with pivots on January, April, July and October.
                                                                                                  Currently, the most liquid DI swaption contracts have expiry
                                                                                                  at Jan11 and have as an underlying the DI rates Apr11,
                                                                                                  Jul11 and Jan12. These tend to have 0.5 to 1 vol of bid-ask
                                                                                                  spread. Swaption maturities on Jul11 and Jan12 are also
                                                                                                  liquid with about 2 vols of bid-as spread, except for the
                                                                                                  longest swaption that is currently trading Jan12 versus
                                                                                                  Jan17, which can have 3 vols of transaction cost.




                                                                                                                                                                       129
                                                                                       MORGAN STANLEY RESEARCH

                                                                                       March, 2012
                                                                                       EM Local Markets Guide




DI swaptions trade on the over-the-counter (OTC) market                                 IDI Options
and can be quoted in terms of BRL premium or implied
                                                                                        The Underlying Asset: IDI
volatility. The liquidity tends to be focused on the call and put
                                                                                        IDI stands for Interbank Deposit Index; it is an index that
swaptions with at-the-money-forward (ATMF) and deltas
                                                                                        accrues every day by the average one-day interbank lending
near 10% and 25%; tailor-made strikes can also be quoted.
                                                                                        rate (CDI). The IDI option is a good financial instrument to
The expiration date of this derivative is the first business day                        trade monetary policy expectations, because IDI options are
of the corresponding month of expiration. The convention is                             options on the forward value of this index. The premium of
to quote price in BRL or as a percentage of the notional. The                           these options reflects the future behavior of the interest rate
derivative price must be converted to USD based on the spot                             from today to the expiration date of the product.
value of USD/BRL.
                                                                                        Currently, there are two IDI indexes open, the first was set at
Exhibit 246                                                                             100,000.00 points on 02-Jan-2003 and options with
Contracts Traded January-July 2010                                                      expiration date until Jan-12 have this index as the underlying
                                                                                        asset. From Feb-12 onward, the IDI index was set at 100k on
                                                                                        02-Jan-2009.

                                                                                        In order to compute the IDI forward, we use DI swap rates to
                                                                                        compound IDI spot. Based on the example in Exhibit 248 we
Source: BMF&Bovespa
                                                                                        compute the IDI forward for Jan-11 as:

Realized Volatility Level                                                               IDI spot  286,345.59
As an example of FRA realized volatility, see Exhibit 247,                              IDI Jan 11  IDI spot * 1  10.68%  252  IDI Jan 11  297,152.40
                                                                                                                                   92

which shows the annualized volatility of the log-return on the
                                                                                        Exhibit 248
continuous FRA rate for two different window size: 30 days
and 60 days.                                                                            DI Information

Exhibit 247
Realized Volatility FRA DI Jan11 x Jan17                                                Source: Bloomberg

 40.00%                                                                                 Exhibit 249
                                                              vol 30d   vol 60d         IDI Forward Diagram
 35.00%

                                                                                          IDI spot                                                 IDI forward
 30.00%


 25.00%
                                                                                        Source: Morgan Stanley
                                                                                          Today                                                        Jan-11
 20.00%                                                                                 In the onshore market, BM&F contracts have IDI strikes
                                                                                        quoted as rounded numbers, so in the previous example,
 15.00%
                                                                                        with the Jan-11 DI rate at 10.68% and the IDI forward at
 10.00%                                                                                 297,152.40, typical Jan-11 swaps would have a strike of
  5.00%
                                                                                        297,000, which could be associated with a rate strike of
      Jan-09     Apr-09   Jul-09   Oct-09   Feb-10   May-10        Aug-10     Dec-10    10.52%:
Source: Morgan Stanley
                                                                                                                       252
DI Swaption Vols in Bloomberg                                                                           IDI strike    du
                                                                                                                                                         252

                                                                                                                        1  10 .52 %   297 ,000   1
                                                                                                                                                          92
                                                                                        Rate strike                                                  
Morgan Stanley DI swaption vols for different                                                           IDI                             286 ,345 .59 
                                                                                                             spot    
delta/tenor/expiry combinations can be found in Bloomberg’s
page MSBZ <GO>.
                                                                                        or a strike of 297,500.00, which could be associated with a
                                                                                        rate strike of 11.03%:
                                                                                                                             252
                                                                                                    297 ,500  92
                                                                                        11 .03 %                 1
                                                                                                    286 ,345 .59 


                                                                                                                                                               130
                                                                                                  MORGAN STANLEY RESEARCH

                                                                                                  March, 2012
                                                                                                  EM Local Markets Guide




In the offshore market there is more flexibility and the strike                                    Nonetheless, we believe that liquidity of IDI options will pick
can be chosen to match ATMF, 25 delta, etc.                                                        up as investors realize the benefits of trading them.

In Exhibit 250, we show the historical values of DI Jan11 and                                      Exhibit 252
in Exhibit 251 we have the IDI spot level.                                                         Contracts Traded January-July 2010
Exhibit 250
DI Jan11
  12.50%
                                                                                                   Source: BMF&Bovespa
  12.00%
                                                                                                   Transaction costs vary by strikes, and liquidity tends to be
  11.50%
                                                                                                   better for close to ATMF IDI options. Low-delta options
                                                                                                   usually have higher bid-ask spreads. Deltas are computed
  11.00%                                                                                           using the Black-Scholes formula described in the technical
  10.50%
                                                                                                   appendix. Transaction costs also depend on the expiration
                                                                                                   date and generally increase with the maturity of the option.
  10.00%                                                                                           The expiration date of this derivative is the first business day
   9.50%
                                                                                                   of the corresponding month of expiration.
        Jan-09      Apr-09      Jul-09        Oct-09       Feb-10     May-10    Aug-10   Dec-10
                                                                                                   In the local market, all months are authorized to be traded on
Source: Bloomberg
                                                                                                   BM&F, but strikes are usually available only on round
Exhibit 251
                                                                                                   numbers. In the offshore market there is more flexibility for
IDI Spot
                                                                                                   tailor-made strikes and maturities but, as usual, liquidity is
  300,000.00                                                                                       better and bid-ask spreads tighter from structures that are
  280,000.00
                                                                                                   readily available in the local market.
  260,000.00

  240,000.00                                                                                       Why Trade IDI Options?
  220,000.00

  200,000.00                                                                                       We believe that IDI options are the best instrument to
  180,000.00                                                                                       implement a view on the monetary policy, as the option is
  160,000.00                                                                                       only related to the section of the interest rate curve from the
  140,000.00                                                                                       trade date to the maturity of the option. On the other hand, DI
  120,000.00
                                                                                                   swaptions have as underlying the section of the curve
  100,000.00
           Jan-03      May-04        Sep-05            Feb-07       Jun-08     Nov-09    Mar-11
                                                                                                   between the expiry of the option and the expiry of the
Source: Bloomberg                                                                                  contract and, hence, are more appropriate for investors with
                                                                                                   a view on the future shape and level of the curve.
Trading, Market Size and Liquidity
As mentioned before, the liquidity of the offshore market
generally mirrors the liquidity of the onshore BM&F market.
Exhibit 252 shows that the total volume negotiated in 2010
up to July was 170 million for DI contracts and 64 million for
IDI contracts.

Exhibit 246 and Exhibit 252 highlight the fact that IDI options
are materially more liquid in the onshore market than DI
swaptions. This is often a surprise to foreign investors as the
offshore IDI market is only in its infancy, whereas liquidity for
offshore DI swaptions is quite decent.

In our view, the reason is that IDI options have unique
payoffs that are not found in any other derivative markets.



                                                                                                                                                               131
                                                                                      MORGAN STANLEY RESEARCH

                                                                                      March, 2012
                                                                                      EM Local Markets Guide




Exhibit 253                                                                            Payoff for Call Contracts
IDI Option and DI Swaptions
                                                                                       The payout of a IDI call option is the maximum value
                                                                                       between the difference of the IDI forward against the strike
                         IDI Option                                                    and zero, or

                                                                                                Payoff  maxIDI T  IDI k ,0
                         3 months                          12 months
                                                                                       IDI k : Strike, IDI T : IDI for expiration date T.
      Today                                    Jan-11                   Jan-12         To illustrate the payoff regarding the Jan-11 call option:

                    DI Swaption                             FRA
                                                                                       IDI k  297,000.00 , IDI T  297,152.40
                                                                                       =>
Source: Morgan Stanley
DI rates and their derivatives are computed from the floating                          Payoff  max 297,152.40  297,000.00 , 0 
overnight CDI. On the other hand, the monetary policy is
conducted by the central Bank using Selic (see Brazil and
                                                                                       Payoff  152.40 BRL
Mexico Local – Market Cheat Sheet, September 16, 2010).                                The call payoff for different IDI forwards can be seen in
                                                                                       Exhibit 256.
Exhibit 254
Selic Rate                                                                             Exhibit 256
  21.50%
                                                                                       IDI Call Options
  19.50%
                                                                                              150.00
  17.50%
                                                                                                            Theoretical Prices           Strike           IDI Fwd           Intrisic Value
                                                                                              100.00
  15.50%

                                                                                               50.00
  13.50%

                                                                                                 0.00
  11.50%
                                                                                        P&L




                                                                                               -50.00
   9.50%

                                                                                              -100.00
   7.50%
       Jan-05       May-06            Sep-07      Feb-09      Jun-10       Nov-11
                                                                                              -150.00
Source: Bloomberg
However, Selic and the CDI are highly correlated and usually                                  -200.00

differ by just a few basis points. Hence, it is reasonable to                                 -250.00

trade changes on the Selic value through IDI options.
                                                                                                    00




                                                                                                                  00




                                                                                                                                    00




                                                                                                                                                     00




                                                                                                                                                                       00




                                                                                                                                                                                         00
                                                                                                 0.




                                                                                                               0.




                                                                                                                                 0.




                                                                                                                                                  0.




                                                                                                                                                                    0.




                                                                                                                                                                                      0.
                                                                                                75




                                                                                                               85




                                                                                                                              95




                                                                                                                                                  05




                                                                                                                                                                   15




                                                                                                                                                                                      25
                                                                                              6,




                                                                                                             6,




                                                                                                                            6,




                                                                                                                                                7,




                                                                                                                                                                 7,




                                                                                                                                                                                    7,
                                                                                            29




                                                                                                           29




                                                                                                                          29




                                                                                                                                              29




                                                                                                                                                               29




Exhibit 255                                                                                                                         IDI fwd                                       29

Historical Spread between CDI and SELIC                                                Source: Morgan Stanley


  0.80%
                                                                                       Payoff for Put Contracts
  0.70%

  0.60%
                                                                                       The purchaser of a put option has a view that the IDI forward
  0.50%
                                                                                       will be less than the strike price on the expiration date. In
  0.40%
                                                                                       other words, the average of the CDI between the trade date
  0.30%
                                                                                       and expiration date of the option has to be less than the CDI
  0.20%
                                                                                       expectation to achieve the strike level.
  0.10%
                                                                                       The put payoff for the expiration date is given by:

                                                                                                         Payoff  maxIDI k  IDI T ,0
  0.00%
      Jan-05        May-06            Sep-07      Feb-09       Jun-10        Nov-11

Source: Bloomberg
                                                                                       Exhibit 257 shows the put payoff for different IDI forwards.




                                                                                                                                                                                       132
                                                                                                     MORGAN STANLEY RESEARCH

                                                                                                     March, 2012
                                                                                                     EM Local Markets Guide




Exhibit 257                                                                                           TIIE swaps are the most liquid OTC interest rate swaps (IRS)
IDI Put Options                                                                                       in Mexico and one of the most liquid ones in Latin America.
       250.00                                                                                         The 28-day TIIE rate (similar to Libor in other markets) is the
                   Theoretical Prices           Strike           IDI Fwd           Intrisic Value     floating benchmark for TIIE swaps. The curve is well
       200.00
                                                                                                      developed and ranges from 3 months (3x1, or three periods
       150.00                                                                                         of 28 days) to 30 years (390x1), but the best liquidity is
                                                                                                      between the 2- and 10-year (26x1 and 130x1).
       100.00
 P&L




                                                                                                      Typical trade sizes are MXN100-250 million in 10-year and
        50.00
                                                                                                      MXN250-500 million in 2-year rates. However, the curve is
          0.00                                                                                        most liquid from the 1y to the 10y sector. The TIIE market
                                                                                                      allows investors to sell rates, a prospect that is more difficult
        -50.00
                                                                                                      through bonos given the underdevelopment of the local repo
       -100.00                                                                                        market.
             00




                         00




                                           00




                                                            00




                                                                              00




                                                                                                00




                                                                                                      The Mexican swaption market allows counterparties to buy
          0.




                      0.




                                        0.




                                                         0.




                                                                           0.




                                                                                             0.
       75




                      85




                                     95




                                                         05




                                                                          15




                                                                                             25
     6,




                    6,




                                   6,




                                                       7,




                                                                        7,




                                                                                           7,
   29




                  29




                                 29




                                                     29




                                                                      29




                                                                                         29




                                           IDI fwd                                                    or sell European style options of various maturities on the
Source: Morgan Stanley                                                                                MXN fixed/float TIIE interest rate swap market. Options
The convention is to quote price in BRL or the notional                                               expire from one month to five years on an underlying of 1-
percentage. The derivative price must be converted to USD                                             month up to 20-year TIIE rates. These options offer option
based on the spot value of USD/BRL.                                                                   buyers the right but not the obligation to pay or receive a
                                                                                                      fixed interest rate at a given strike price on a given date in an
How to Trade Monetary Policy?                                                                         agreed upon amount for a given swap maturity. Caps/floors
                                                                                                      of up to five years on 20-year rates are also quoted.
Due to the close link between the Selic rate and the CDI rate,
                                                                                                      Swaptions have become more liquid over the past year and
the investor can trade monetary policy through vanilla
                                                                                                      are traded in sizes similar to TIIE swaps. Delta neutral
options or using different strategies, as butterflies, call
                                                                                                      strategies are for the most part quoted as Black-Scholes
spreads, put spreads or any other payoff can be made by
                                                                                                      volatility expressed as a percentage of yield/underlying swap
mixing calls/puts.
                                                                                                      rate.
If an investor believes that O/N interest rate cuts will be
higher than expected (Selic rate will be lower than the market                                        Payoff of Payer Swaptions (Call on Rates)
expectation), the trader would buy a put option on the IDI
                                                                                                      Suppose TIIE swap (fixed) rate at the expiry of swaption
index, because the adjusted IDI will be lower than expected.
                                                                                                      happens to be R and the swaption strike rate is Rk . Given
As we have mentioned, the main difference between the DI                                              that TIIE swaps have coupon payments every 28 days or 13
swaptions and the IDI options is the section of the interest                                          times per year, the payoff from the payer swaption consists
rate curve that is relevant to the option. IDI options are                                            of a series of cash flows equal to

                                                                                                                Payoff  maxR  Rk ,0  / 13
suitable for trades and views on the monetary policy, while in
using DI swaptions options, expected curve shapes and
levels are more important.                                                                            The real value of this payoff depends on how these payoffs
                                                                                                      will be discounted. Assume we have a 3m ATM swaption
Mexico: TIIE Swaptions                                                                                contract on the 10y TIIE swap contract, the current 10y TIIE
                                                                                                      rate is 6.55%, and the size of the notional is MXN100,000.
The Underlying Asset: TIIE
                                                                                                      Exhibit 258 displays the payoff as a function on different 10y
The TIIE (Tasa de Interes Interbancaria de Equilibrio) is the                                         TIIE rates 3 x 28 days later with the future cash flows
benchmark interbank interest rate that Mexican banks use to                                           discounted using the future 10y TIIE rate (which should be a
borrow or lend from the Bank of Mexico. It is calculated daily                                        good approximation to the actual discount rates at the expiry
by the central bank by the arithmetic mean of quotes                                                  of the swaption).
submitted by at least six commercial banks.




                                                                                                                                                                   133
                                                                    MORGAN STANLEY RESEARCH

                                                                    March, 2012
                                                                    EM Local Markets Guide




Exhibit 258                                                          each payment can be thought of as a separate call option,
Call Theoretical Prices                                              called caplet, with

                                                                               Payoff  ndays * maxR  Rk ,0  / 360,
MXN 12,000
                                         Intrisic Value


MXN 10,000                                                           where R is the current floating rate in the swap receiver
                                                                     position and ndays is the number of actual days between
 MXN 8,000
                                                                     payments.
 MXN 6,000                                                           Similarly, in a swap payer position, an investor pays the fixed
                                                                     rate and receives the floating rate. Floor guarantees the
 MXN 4,000
                                                                     received floating rate payments to be at least the
 MXN 2,000
                                                                     predetermined rate, Rk .Hence, for the owner of the floor,
                                                                     each received payment can be thought of as a separate put
     MXN 0                                                           option, called floorlet, with

                                                                               Payoff  ndays * maxRk  R,0  / 360,
         6.0%            6.5%     7.0%              7.5%     8.0%
Source: Morgan Stanley


                                                                     where R is the current floating rate in the swap receiver
Payoff of Receiver Swaptions (Put on Rates)
                                                                     position and ndays is the number of actual days between
Similarly to payer swaptions, the payoff from the receiver           payments. Please note that buying and selling both caplets
swaption consists of a series of cash flows equal to                 and floorlets means taking a position on actual monetary

              Payoff  MaxRk  R,0 / 13
                                                                     policy which determines the 28D rate, and not on current
                                                                     market views of what this policy will be in the future.
Using the same numbers as in Exhibit 258, Exhibit 259
                                                                     MXN TIIE caps and floors have the same payment calendar
displays the payoff as a function of different 10Y TIIE rates
                                                                     as TIIE swaps except for the initial TIIE reference. Caps and
3*28 days later with the future cash flows discounted again
                                                                     floors initially reference TIIE on trade date + 28 days, hence
using the future 10y TIIE rate.
                                                                     having one less calculation period than the corresponding
Exhibit 259                                                          interest rate swap maturity. Intrinsic value, if any, for each
Put Theoretical Prices                                               caplet or floorlet is paid by option writer to option buyer in
MXN 14,000                                                           arrears.
                                         Intrisic Value

MXN 12,000
                                                                     Valuation in Bloomberg
MXN 10,000
                                                                     Bloomberg function VCUB<GO> allows one to view current
 MXN 8,000                                                           market ARM, ITM, and OTM volatility quotes for different
                                                                     TIIE cap, floor, and swaption contract tenors and expiries.
 MXN 6,000
                                                                     One can view these in a table format or as 2D and 3D
 MXN 4,000                                                           graphs.
 MXN 2,000                                                           In addition, it is possible to see the histories of many of the
     MXN 0
                                                                     volatilities. By selecting 91) Actions -> Export to Excel ->
         5.0%            5.5%     6.0%              6.5%     7.0%    Tickers, one can find tickers for different contracts and type
Source: Morgan Stanley                                               TICKER Curncy GP<GO> in Bloomberg gives the history.

Caps and Floors
Caps and floors protect against movements in a swap
contract’s floating rate (28D TIIE). In a swap receiver
position, an investor receives the fixed rate and pays the
floating rate. Cap limits the payments to be no more than the
predetermined rate Rk . Hence, for the owner of the cap,



                                                                                                                                 134
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




                                                                    Delta:
Technical Appendix
In the DI futures market we have:                                              call  N d1
                   100000                                                      put  N d1  1
          PU 
                                                                                      F      
                           du                                                                 2
                  (1  i ) 252                                                     ln    * t 
                                                                                     X  2 
i : Interest Rate in percentage                                               d1 
du: days between the trade date and the day preceding the
                                                                                        * t
                                                                                                               du
expiration date
                                                                                                 DI  252
                                                                              F  IDI spot * 1     
                                                                                              100 
For FRA, we have:
                                                                                 du
                                                   252
                                                                              t
                                                                                 252
                                          dulong  du short
                   1  ilong  252
                               du long

          FRA                                              1
                                         
                 1  ishort  252
                               du short
                                                                    F: IDI for the expiration date T.
                                         
                                                                   IDI spot: IDI index for the current day.

                                                                    DI: DI Rate for the same expiration date of the option.
ilong : long term Interest Rate in percentage
                                                                    du: Number of business days until the expiration date.
dulong: days between the trade date and the day preceding
the expiration date for the long term contract                      X: Strike Price.

ishort : short term Interest Rate in percentage                     σ: Volatility of the underlying asset, based on the day before
                                                                    the trade date.
dushort: days between the trade date and the day preceding
the expiration date for the short term contract.



For the IDI options:
                                               1
                                 i  252
          IDI t  IDI t 1 * 1  t 1 
                              100 
i t-1 : CDI daily rate, computed by the CETIP (OTC clearing
house) .

IDI t-1 : IDI index for the previous day.

The IDI spot is published everyday by BMF&BOVESPA
before the market starts.




                                                                                                                              135
                                                                          MORGAN STANLEY RESEARCH

                                                                          March, 2012
                                                                          EM Local Markets Guide




Hiking Cycles and Term Spreads in Latin America
Juha Seppala

First Published March 26th, 2010: Hiking Cycles and Term Spreads in Latin America

We study the dynamics of Latin America interest rate                      Monetary Policy in Latin America
curves around the inflection period when the monetary
                                                                          The end of the easing cycle and the beginning of a hiking
authorities start a tightening cycle. We find some
                                                                          cycle has either arrived or will arrive soon to the three major
evidence which suggests curve flatteners as good
                                                                          Latin American economies. Brazil’s central bank raised its
risk/reward exposures near the infection points.
                                                                          overnight Selic rate by 75bp at the end of April and Morgan
In this piece, we take a look at the behavior of term spreads of          Stanley expects Chile and Mexico to start raising overnight
some of the Latin America local interest rates curves when                rates in June 2010 and March 2011, respectively. Given that
monetary policy shifts from neutral to hiking mode.                       the interest rate curves can be decomposed as the sum of
                                                                          expected future short-term interest rates and risk premia
A serious challenge for any strong conclusion is the region’s             (minus a small option value of convexity, which we ignore in
relatively recent macro stabilization and use of monetary                 this article), it is valuable to understand what this view of
policy tools. Indeed, the data contain only two full hiking               future monetary policy paths implies to the curve trades.
cycles in Brazil and only one hiking cycle in Chile and Mexico.
The current exchange rate regime in Brazil dates from 1999,               The biggest problem in analyzing the relationship between
Chile started using the nominal overnight rate as a monetary              monetary policy and term spreads in Latin American countries is
policy instrument only in 2001 and Mexico formally                        the relatively recent stabilization of markets and policy in these
implemented its inflation-targeting strategy only in 1999. Also,          countries. The current exchange rate regime in Brazil dates from
developed interest rate curves for all these countries only date          1999. Similarly, Chile started using the nominal overnight rate as
from the early 2000s. Hence, given the short time series of               a monetary policy instrument in 2001 and Mexico formally
relevant data in Latin American countries, we also look at the            implemented its inflation-targeting strategy in 1999. Finally, and
                                                                          most importantly, the Camara swap market in Chile and the TIIE
empirical and theoretical evidence concerning the US to
                                                                          swap market in Mexico only date from early 2000s.
suggest patterns.
                                                                          In our analysis, we use monthly data from January 2004.
Although inconclusive from a strict econometric point of view,
                                                                          Given the short time series of relevant data in Latin American
the results from the historical data analysis suggest that
                                                                          countries, it is worthwhile to discuss the empirical and
curve-flattener exposures tend to exhibit solid risk/rewards
                                                                          theoretical evidence concerning the US.
near the infection points.
                                                                          Monetary Policy and the Yield Curve in the US
Taking as reference the 1s10s slope, Brazil’s DI curve
experienced on average about 10bp of flattening per week in               Every recession in the US since the early 1950s (that is, since
the months before and after the hiking cycle started. The                 the establishment of well-functioning government bond
developments are a bit more mixed and less pronounced for                 markets) has been predicted by an inverted (downward-
the other countries, with the Chile Camara curve materially               sloping) yield curve with only one false signal. 26 The yield
flattening for several months after the first hike and the slope          curve is the steepest at the bottom of the business cycle and
of the Mexican TIIE curve gradually reducing as the tightening            the most inverted at the top of the business cycle.
cycle approached. The differences between the countries in                The precise mechanism which causes this regularity is not
how fast the curve responds to anticipated changes in                     known, but clearly the Fed’s policy plays an important role. At
monetary policy can probably be accounted by differences in               the top of the cycle, the Fed is concerned about an
the central banks’ credibility and differences in response to             overheated economy, which results in slowing activity, which
outside influences (i.e., Fed policy).                                    results in lower expectations of future inflation and hence
                                                                          lower expected future short-term interest rates.


                                                                          26
                                                                               See The Yield Curve as a Leading Indicator: Frequently Asked Questions by Arturo
                                                                               Estrella, October 2005, available at
                                                                               www.newyorkfed.org/research/capital_markets/ycfaq.pdf for a long list of references.



                                                                                                                                                                  136
                                                                                                      MORGAN STANLEY RESEARCH

                                                                                                      March, 2012
                                                                                                      EM Local Markets Guide




Market expectations about future economic activity and the                                            Exhibit 261
risk premium play a role as well. 27 It is worth emphasizing that                                     Brazil: Cumulative Change in Term Spreads
a large part of risk premium is compensation for the                                                   2.5

uncertainty about the future direction of monetary policy. As                                               2

long the markets are uncertain when the hiking cycle is going                                          1.5
                                                                                                                                                                  1s5s           2s5s
to start, a non-trivial part of the high long-term rates is due to                                          1
                                                                                                                                                                  1s10s          2s10s
risk premia. When the consensus is formed about the                                                    0.5

beginning of the hiking cycle, the risk premia can begin to                                                 0
                                                                                                                -5      -4      -3      -2          -1     0       1        2       3           4      5
                                                                                                       -0.5
collapse.
                                                                                                        -1             Months from the beg. of hiking cycle

Summarizing, based on the US evidence, we should expect                                                -1.5

the approaching hiking cycle to signal flattening term spreads.                                         -2

Testing this hypothesis empirically is not easy as the data                                            -2.5

contain only two full hiking cycles in Brazil and only one hiking                                       -3
                                                                                                      Source: Bloomberg, Morgan Stanley Research
cycle in Chile and Mexico. Therefore, in the subsequent
analysis, we limit ourselves to graphical evidence, which turns
                                                                                                      Exhibit 261 clearly documents how the curve begins to flatten
out to be quite revealing.
                                                                                                      before the hikes begin and continues to do so for several
Brazil                                                                                                months afterwards. Typically 1s5s and 1s10s sectors flatten
                                                                                                      more. Moreover, these same sectors still have some upside
In Exhibit 260 we present PRExDI swap curve term spreads
                                                                                                      left as far as flatteners are concerned.
for different maturities and the overnight Selic target rate.
Simple eyeballing the data reveals significant negative                                               Chile
correlation between the two series. As a matter of fact, the
correlations vary between -0.67 for the 1s5s spread and -0.51                                         In Exhibit 262, we present CLPxCamara swap curve term
for the 2s10s spread. Moreover, it is very informative to look                                        spreads for different maturities and the overnight central bank
at cumulative changes in term spreads relative to the level at                                        target rate.
the end the first hiking month. Exhibit 261 shows the average                                         Exhibit 262
differences for each month five months before to five months
                                                                                                      Chile: Term Spreads and O/N
after the beginning of the hiking cycle relative to the level at
                                                                                                       6                                                                                                   9
the beginning of the cycle. 28                                                                                                       1s5s                  2s5s
                                                                                                       5                                                                                                   8
                                                                                                                                     1s10s                 2s10s
Exhibit 260                                                                                                                                                                                                7
                                                                                                       4                             O/N (RHS)
Brazil: Term Spreads and Selic                                                                                                                                                                             6
                                                                                                       3
     5                                                                                           22
                   1s5s          2s5s          1s10s           2s10s          O/N (RHS)                                                                                                                    5
     4                                                                                                 2
                                                                                                 20                                                                                                        4
     3                                                                                                 1
                                                                                                 18                                                                                                        3
     2
                                                                                                        0
                                                                                                                                                                                                           2
     1                                                                                           16      Jan-   Jul-   Jan-   Jul-   Jan-    Jul-   Jan-   Jul-   Jan-    Jul-   Jan-    Jul-   Jan-
                                                                                                       -1 04     04     05     05     06      06     07     07     08     08      09      09     10        1
  0                                                                                              14
   Jan-     Jul-   Jan-   Jul-   Jan-   Jul-   Jan-    Jul-   Jan-   Jul-   Jan-   Jul-   Jan-         -2                                                                                                  0
 -1 04       04     05     05     06     06     07      07     08     08     09     09     10
                                                                                                 12
                                                                                                      Source: Morgan Stanley Research, Bloomberg
 -2
                                                                                                 10
 -3
                                                                                                      The correlations are even more remarkable than in the case
 -4                                                                                              8
                                                                                                      of Selic – all of them are at least -0.96. Exhibit 263 shows the
Source: Bloomberg, Morgan Stanley Research                                                            differences for each month five months before to five months
                                                                                                      after the beginning of the hiking cycle relative to the level at
27
     See Monetary Policy and Rejections of the Expectations Hypothesis by Federico Ravenna and
     Juha Seppala, May 2007, available at                                                             the beginning of the cycle during the only hiking cycle in our
     http://papers.ssrn.com/sol3/papers.cfm?abstract_id=923779 for a theoretical model able to        sample, starting in July 2007.
     capture this mechanism.
28
     The hiking cycles begin in November 2004, April 2008, and April 2010. The contribution of
     the last hike is used only for the months before the beginning of the cycle.



                                                                                                                                                                                                       137
                                                                                                               MORGAN STANLEY RESEARCH

                                                                                                               March, 2012
                                                                                                               EM Local Markets Guide




Exhibit 263                                                                                                    Exhibit 265
Chile: Cumulative Change in Term Spreads                                                                       Mexico: Cumulative Change in Term Spreads
 0.40                                                                                                           0.60
                                                                                                                                                     1s5s      2s5s
 0.20                                                                                                           0.50
                                                                                                                                                     1s10s     2s10s
                                                                                                                0.40
 0.00
           -5       -4          -3          -2       -1      0       1           2          3      4      5     0.30
 -0.20
                 Months from the beg. of hiking cycle                                                           0.20
 -0.40
                                                                                                                0.10
 -0.60
                                1s5s                2s5s                                                        0.00
 -0.80                          1s10s               2s10s                                                                -5       -4     -3     -2      -1    0        1   2   3   4   5
                                                                                                                -0.10

 -1.00                                                                                                                        Months from the beg. of hiking cycle
                                                                                                                -0.20

 -1.20                                                                                                          -0.30
Source: Morgan Stanley Research, Bloomberg                                                                     Source: Morgan Stanley Research, Bloomberg


Exhibit 263 shows how the curve begins to flatten one month                                                    Mexico differs from Brazil and Chile as in the previous hiking
before the hikes begin and continue to do so for several                                                       cycle flattening began 12 months before the beginning of the
months afterwards. The difference in timing compared to                                                        cycle and ended already one month after the beginning of the
Brazil can probably be contributed to Brazil’s central bank’s                                                  tightening cycle. Drawing conclusions about Mexico is
efforts to maintain open communication with the market.                                                        complicated by the fact that Mexico’s rate moves have
Again, 1s5s and 1s10s sectors flatten more. Moreover, these                                                    traditionally been correlated with the moves by the Fed in the
same sectors are still very close to historical highs.                                                         US. EM Profile: Hiking Cycles and Flattening Trends in
                                                                                                               CEEMEA
Mexico
In Exhibit 264 we present TIIE swap curve term spreads for
different maturities and the benchmark overnight rate (bank
funding rate until 1/18/2008 and the target rate after that). The
correlations are between the levels in Brazil and Chile: they
vary from -0.67 for 2s10s to -0.8 for 1s5s. Exhibit 265 shows
the differences for each month 5 months before to 5 months
after the beginning of the hiking cycle relative to the level at
the beginning of the cycle during the April 2007 hiking cycle

Exhibit 264
Mexico: Term Spreads and O/N
 4                                                                                                        11
                 1s5s                2s5s            1s10s          2s10s              O/N (RHS)
 3                                                                                                        10

 3
                                                                                                          9

 2
                                                                                                          8
 2
                                                                                                          7
 1

                                                                                                          6
 1

  0                                                                                                       5
   Jan-   Jul-   Jan-    Jul-    Jan-        Jul-   Jan-    Jul-   Jan-   Jul-       Jan-   Jul-   Jan-
 -1 04     04     05      05      06          06     07      07     08     08         09     09     10    4

Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                                       138
                                                                          MORGAN STANLEY RESEARCH

                                                                          March, 2012
                                                                          EM Local Markets Guide




Hiking Cycles and Flattening Trends in CEEMEA
                     th
First Published July 14 , 2010: Hiking Cycles and Flattening Trends in CEEMEA

The Next Big Trend – Curve Flattening?                                    Exhibit 266
                                                                          CEEMEA Policy Rate Forecasts
Since 2008, as most countries in CEEMEA were experiencing                                   Current    3Q10      4Q10               1Q11     2Q11       3Q11       4Q11
growth slowdown and recession, central banks in this region               Czech Republic      0.75      0.75      0.75              0.75     1.00       1.25       1.50
                                                                          Hungary             5.25      5.25      5.25              5.25     5.25       5.50       5.75
have embarked on deep rate-cutting cycles. Yield curves bull-             Israel              1.50      1.75      2.00              2.25     2.75       3.25       3.50
                                                                          Poland              3.50      3.75      4.00              4.25     4.50       4.50       4.50
steepened on the back of that, and our economists now think               South Africa        6.50      6.50      6.50              6.50     7.00       7.50       7.50
that the policy rates have most likely bottomed in CEEMEA. In             Turkey              6.50      6.50      6.50              7.25     8.00       8.50       9.00
                                                                          Source: Morgan Stanley Research estimates
fact, the Bank of Israel (BoI) started hiking rates back in July
2009, as part of its rate normalisation process.
                                                                          Czech Republic
Most central banks are likely to remain on hold over the next
                                                                          There were two hiking cycles in Czech since 2004 – the first
few months, according to our economists’ forecasts. In this
                                                                          (mini) cycle was between April and September 2004, when
environment, the curve shape tends to be dictated by fund
                                                                          the CNB hiked just 50bp. A more prolonged hiking cycle was
flows and risk sentiment: in a positive risk environment,
                                                                          seen between October 2005 and February 2008, with a
inflows are normally directed to the mid-to-long end of the
                                                                          cumulative rate hike of 200bp over 28 months. In Exhibit 267
curve, particularly if there is no major mispricing in the front
                                                                          and Exhibit 268, we show the policy rate paths during the
end of the curves. The opposite generally holds true in a risk-
                                                                          hiking cycles, as well as the evolution of CZK 2s5s and 2s10s
off environment as well.
                                                                          spread.
As we approach the start of the hiking cycles in CEEMEA,
                                                                          Unsurprisingly, given the shallow hiking cycle in 2004, the
the yield curves tend to flatten in anticipation of a) higher
                                                                          curve flattening was fairly modest as well – 2s5s and 2s10s
policy (and hence front-end) rates; as well as b) lower
                                                                          flattened by 50bp and 80bp (peak to trough), respectively.
inflation expectations in the future, and hence lower
                                                                          More interestingly, the flattening started almost 5 months
expected future short-term interest rates.
                                                                          before the first rate hike (see Exhibit 267).
Market expectations about future economic activity and risk
                                                                          Exhibit 267
premium play a role as well. A large part of risk premium is
                                                                          Czech Republic Hiking Cycle (I)
compensation for the uncertainty about the future direction of
                                                                                2.6                                                                                     200
monetary policy. As long as the markets are uncertain when
                                                                                2.5                                                                                     180
the hiking cycle is going to commence, a non-trivial part of the                                                                                                        160
                                                                                2.4
high long-term rates is due to risk premia. When the                                             CZ Repo                                                                140
                                                                                2.3
consensus is formed about the start of the hiking cycle, or                                      CZK 2s5s
                                                                                                 CZK 2s10s
                                                                                                                                                                        120
                                                                                2.2
when the central banks signal that the tightening cycle is                                                                                                              100
                                                                                2.1
imminent, the risk premia can thus begin to collapse.                                                                                                                   80
                                                                                2.0
                                                                                                                                                                        60

In fact, looking back at history, the swap curves generally                     1.9                                                                                     40

started flattening before the delivery of the rate hikes,                       1.8                                                                                     20
                                                                                  02/04 03/04   04/04 05/04   06/04 07/04   08/04   09/04 10/04   11/04 12/04   01/05
and most continued doing so through the hiking cycle.
                                                                          Source: Morgan Stanley Research, Bloomberg
Indeed, this is already happening in Latin America, with the              In the second hiking cycle, 2s5s and 2s10s flattened by
BRL DI curve flattening as hiking cycles began in this market.            almost 80bp and 150bp, respectively, on the back of 200bp of
A similar analysis has been done by our LatAm colleagues –                tightening by the CNB. The flattening trend started around
see Hiking Cycles and Term Spreads in Latin America,                      7 months before the first rate hike. In fact, the curve was
Rogerio Oliveira and Juha Seppala, May 26, 2010.                          already flattening even before the CNB delivered the final
In order to help us time our entry into flattening trades, we             rate cut, as the market prepared for the reversal of the
attempt to investigate a) how long before the delivery of the             monetary policy cycle.
first hike that the swap curve started its flattening trend; and
b) how long the flattening trend lasted for.

                                                                                                                                                                    139
                                                                                                 MORGAN STANLEY RESEARCH

                                                                                                 March, 2012
                                                                                                 EM Local Markets Guide




It is worth highlighting, however, that the flattening trend was                                 the back of one 250bp hike, but steepened up almost
not entirely one-directional. There were three episodes of                                       immediately after that.
steepening during the hiking cycle – the first two instances
                                                                                                 During the first hiking cycle, 2s5s and 2s10s flattened by 70bp
happened when the CNB held the policy rate unchanged for 9                                       and 120bp, respectively. 2s5s started flattening approximately
months and 7 months, respectively, while the third instance                                      4 months before the first rate hike, but 2s10s only started
was on the back of the EUR curve steepening.                                                     flattening convincingly around 2 months before the first hike.
Exhibit 268                                                                                      Exhibit 270
Czech Republic Hiking Cycle (II) and Three                                                       Hungary Hiking Cycle (I)
Episodes of Steepening                                                                                  8.5                            HU Repo                                          40
                                                                                                                                       HUF 2s5s
  4.0                                                                                      160          8.0                                                                             20
                                                                                                                                       HUF 2s10s
                                                   CZ Repo                                 140                                                                                          0
                                                                                                        7.5
  3.5                                              CZK 2s5s
                                                                                           120                                                                                          -20
                                                   CZK 2s10s                                            7.0
  3.0                                                                                      100                                                                                          -40
                                                                                                        6.5
                                                                                           80                                                                                           -60
  2.5                                                                                                   6.0
                                                                                           60                                                                                           -80

                                                                                                        5.5                                                                             -100
  2.0                                                                                      40
                                                                                                        5.0                                                                             -120
                                                                                           20
  1.5                                                                                                     02/06   03/06    04/06   05/06   06/06   07/06    08/06      09/06    10/06
                                                                                           0
                                                                                                 Source: Morgan Stanley Research, Bloomberg
  1.0                                                                                      -20
    03/05 06/05 09/05 12/05 03/06 06/06 09/06 12/06 03/07 06/07 09/07 12/07 03/08                In the second hiking cycle, 2s5s and 2s10s flattened by 45bp
                                                                                                 and 90bp, respectively. The flattening trend started 2 months
Source: Morgan Stanley Research, Bloomberg
                                                                                                 before the first rate hike, and was already running out of
Our economists are forecasting unchanged rates at 0.75%
over the next three quarters. The next 25bp hike is currently                                    steam even before the first rate hike.
forecast to be delivered in 2Q11. Assuming that the curve                                        Given the volatile nature of the Hungarian rates market,
behaves like it did during the second hiking cycle (i.e.,                                        coupled with the higher weight currency fluctuations
flattening 7 months before the first rate hike), investors                                       have in the NBH’s decision-making process, it is perhaps
should consider putting on flatteners towards the end of                                         not surprising that the curve was only able to flatten
this year.                                                                                       convincingly when it was fairly close to the first rate hike.
Exhibit 269                                                                                      Exhibit 271
Czech Republic: Cumulative Change in Curve                                                       Hungary Hiking Cycle (II)
Spreads                                                                                                 8.6                                                        HU Repo              0
                                                                                                                                                                   HUF 2s5s
                         80                                                                             8.4                                                        HUF 2s10s
                                                CZK 2s5s (I)         CZK 2s10s (I)                                                                                                      -20
                         60                     CZK 2s5s (II)        CZK 2s10s (II)
                                                                                                        8.2
                                                                                                                                                                                        -40
                         40                                                                             8.0
                                                                                                                                                                                        -60
                         20                                                                             7.8
                                                                                                                                                                                        -80
                          0                                                                             7.6
           -10     -5         0    5       10       15          20    25       30     35
                        -20                                                                                                                                                             -100
                                                                                                        7.4

                        -40                                                                             7.2                                                                             -120

                        -60
                                                                                                        7.0                                                                             -140
                                                                                                          02/08           03/08            04/08           05/08               06/08
                        -80


Y-axis shows the change in curve spreads from the date of the first hike (set to be zero)        Source: Morgan Stanley Research, Bloomberg
Source: Morgan Stanley Research, Bloomberg


Hungary
There were also two hiking cycles in Hungary – the first was
from June to November 2006 (200bp hike), and the second
was a much shallower hiking cycle (100bp), from March to
June 2008. We have excluded the emergency hiking cycle in
September 2008, where the curve flattened significantly on


                                                                                                                                                                                               140
                                                                                             MORGAN STANLEY RESEARCH

                                                                                             March, 2012
                                                                                             EM Local Markets Guide




Exhibit 272                                                                                  Exhibit 273
Hungary: Cumulative Change in Curve Spreads                                                  Israel Hiking Cycle (I)
                                            80
                                                                                                  4.50                                                          IS Repo                  100
                                            60         HUF 2s5s (I)      HUF 2s10s (I)                                                                          ILS 2s5s
                                                                                                                                                                                         90
                                                       HUF 2s5s (II)     HUF 2s10s (II)                                                                         ILS 2s10s
                                            40                                                    4.25
                                                                                                                                                                                         80
                                            20
                                                                                                                                                                                         70
                                             0                                                    4.00
         -6          -4           -2              0         2            4               6
                                            -20                                                                                                                                          60

                                            -40                                                   3.75
                                                                                                                                                                                         50
                                            -60
                                                                                                                                                                                         40
                                                                                                  3.50
                                            -80
                                                                                                                                                                                         30
Y-axis shows the change in curve spreads from the date of the first hike (set to be zero)
Source: Morgan Stanley Research, Bloomberg                                                        3.25                                                                                   20
                                                                                                     06/07        07/07     08/07      09/07      10/07         11/07         12/07
Our economists are forecasting unchanged rates over the
next 12 months, and the first 25bp to be delivered in 3Q11. It                               Source: Morgan Stanley Research, Bloomberg

is therefore still too early to position for flatteners in                                   Exhibit 274
Hungary, in our view. We think that any weakness will be                                     Israel Hiking Cycle (II)
driven by the 5y sector, and therefore prefer paying HUF                                          4.50                                                          IS Repo                  200
2s5s10s butterfly.                                                                                                                                              ILS 2s5s                 180
                                                                                                  4.25                                                          ILS 2s10s
                                                                                                                                                                                         160
Israel                                                                                            4.00                                                                                   140

There were also two relatively short hiking cycles in Israel –                                                                                                                           120
                                                                                                  3.75
75bp hike between July 2007 and January 2008, and 100bp                                                                                                                                  100

hike between May and September 2008. More recently, the                                           3.50                                                                                   80

BoI started hiking rates since July 2009 (100bp so far), and                                                                                                                             60
                                                                                                  3.25
our economists are forecasting another 200bp of hikes over                                                                                                                               40
the next 18 months.                                                                               3.00                                                                                   20
                                                                                                     03/08 04/08 04/08 05/08 05/08 06/08 06/08 07/08 07/08 08/08 08/08 08/08
During both earlier hiking cycles, the curve started
flattening 1-2 months before the delivery of the first hike.                                 Source: Morgan Stanley Research, Bloomberg

2s5s flattened (peak to trough) by 30bp in the first hiking                                  Exhibit 275

cycle, and by 50bp in the second cycle – both by                                             Israel: Cumulative Change in Curve Spreads
approximately 50% of the cumulative amount of hikes. 2s10s                                                                100
                                                                                                                                                ILS 2s5s (I)            ILS 2s10s (I)
flattened by more – 40bp in the first hiking cycle and 70bp in                                                                                  ILS 2s5s (II)           ILS 2s10s (II)
                                                                                                                           80
the second hiking cycle, or approximately 70% of the
cumulative rate hikes.                                                                                                     60


What is interesting, however, is that most of the flattening                                                               40

happened before the delivery of the rate hikes (see Exhibit                                                                20
273-318). In fact, during hiking cycle (I), the curve retraced
                                                                                                                            0
almost all of the flattening after the delivery of the first rate hike.                          -3          -2      -1         0      1         2          3            4         5          6
This could be due to the efficient, forward-looking signaling from                                                        -20

the BoI, and therefore allowing the curve to adjust very quickly
                                                                                                                          -40
during the turns of the policy rate cycles.
                                                                                             Y-axis shows the change in curve spreads from the date of the first hike (set to be zero)
                                                                                             Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                                         141
                                                                                                  MORGAN STANLEY RESEARCH

                                                                                                  March, 2012
                                                                                                  EM Local Markets Guide




Poland                                                                                            Exhibit 278
                                                                                                  Poland: Cumulative Change in Curve Spreads
Since 2004, Poland has seen two hiking cycles as well. The
first one was relatively short (125bp over three months – June                                                               150
                                                                                                                                                    PLN 2s5s (I)            PLN 2s10s (I)
                                                                                                                                                    PLN 2s5s (II)           PLN 2s10s (II)
to September 2004), with peak-to-trough flattening of 105bp                                                                  100

for 2s5s and 130bp for 2s10s. The curve started flattening
                                                                                                                              50
around 5 months before the first hike, and continued to do so
until the last hike was delivered.                                                                                             0
                                                                                                    -10            -5              0           5               10                15            20
One possible reason why the curve flattened way before the                                                                   -50

first hike was due to the sell-off in 2003/early 2004, pushing
                                                                                                                            -100
the curve steeper. As risk sentiments improved, the curve
also started to flatten, even before the NBP started hiking                                                                 -150

rates in June 2004.Exhibit 276                                                                    Y-axis shows the change in curve spreads from the date of the first hike (set to be zero)
                                                                                                  Source: Morgan Stanley Research, Bloomberg
Poland Hiking Cycle (I)
  6.75                                                                                      100   South Africa
                                                         PD Repo                            80
  6.50
                                                         PLN 2s5s
                                                                                            60    During the first hiking cycle (January-September 2002), the
  6.25                                                   PLN 2s10s
                                                                                            40    SARB hiked 400bp over 8 months, at 100bp per meeting. The
  6.00                                                                                      20    curve started to flatten convincingly a month prior to the first
  5.75                                                                                      0
                                                                                                  hike, with a peak-to-trough flattening of 170bp and 250bp for
  5.50
                                                                                            -20
                                                                                                  2s5s and 2s10s, respectively. On average, the flattening was
                                                                                            -40
  5.25
                                                                                                  around 20bp/month for 2s5s, and 27bp/month for 2s10s.
                                                                                            -60

  5.00                                                                                      -80   Exhibit 279
     02/04      03/04        04/04    05/04      06/04        07/04     08/04       09/04
                                                                                                  South Africa Hiking Cycle (I)
Source: Morgan Stanley Research, Bloomberg
                                                                                                    14.0                                                                                            150
The second hiking cycle lasted for 15 months (April 2007 to
                                                                                                    13.5
July 2008), where the NBP hiked 200bp. Despite the deeper                                           13.0
                                                                                                                                                                                                    100


hiking cycle, 2s5s and 2s10s flattened by the same amount as                                        12.5                                                                                            50

the first hiking cycle. The pace of the flattening was much                                         12.0
                                                                                                                                                                                                    0

more gradual than in the first hiking cycle, starting at                                            11.5
                                                                                                                                                                                                    -50
                                                                                                    11.0
around 3 months before the first hike.
                                                                                                    10.5                                           SA Repo                                          -100
                                                                                                    10.0                                           ZAR 2s5s
Exhibit 277                                                                                                                                                                                         -150
                                                                                                      9.5                                          ZAR 2s10s
Poland Hiking Cycle (II)                                                                              9.0                                                                                           -200
                                                                                                        12/01   01/02   02/02 03/02    04/02   05/02      06/02     07/02     08/02    09/02
  6.5                                                                                       60

                                                                                            40    Source: Morgan Stanley Research, Bloomberg
  6.0
                                                                                                  The second hiking cycle was much longer, spanning over 25
                                                                                            20
  5.5                                                                                             months (June 2006 to July 2008), with a more gradual, 50bp-
  5.0
               PD Repo
                                                                                            0
                                                                                                  per-meeting hikes. The flattening trend started much earlier,
               PLN 2s5s
                                                                                            -20   almost 13 months before the first hike, and continued until
               PLN 2s10s
  4.5
                                                                                            -40   after the penultimate hike in November 2007 (possibly
  4.0                                                                                             because the market thought that was the last hike). 2s5s and
                                                                                            -60
                                                                                                  2s10s flattened by 180bp and 290bp, or roughly 6bp/month
  3.5                                                                                       -80
    01/07    03/07   05/07    07/07   09/07   11/07   01/08    03/08   05/08    07/08   09/08
                                                                                                  and 10bp/month, respectively.

Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                                                142
                                                                                                   MORGAN STANLEY RESEARCH

                                                                                                   March, 2012
                                                                                                   EM Local Markets Guide




Exhibit 280                                                                                        Turkey
South Africa Hiking Cycle (II)
                                                                                                   The previous two hiking cycles only lasted for 2-3 months.
  12                                              SA Repo                                   150
                                                                                                   The first was in fact an emergency rate hike in June 2006,
                                                  ZAR 2s5s
  11                                              ZAR 2s10s
                                                                                            100
                                                                                                   when the market was selling off (USD/TRY spiked from 1.32
                                                                                            50     to 1.70 within a month). The CBT reacted by hiking 175bp and
  10
                                                                                            0      225bp in two meetings within the same month. When the
    9                                                                                       -50    market settled down thereafter, the CBT hiked by another
                                                                                            -100   25bp in July 2006, and that marked the end of the hiking
    8
                                                                                            -150   cycle. Unsurprisingly, the curve started flattening just before
    7
                                                                                            -200   the rate hikes, with 2s5s and 2s10s flattening by 230bp and
    6                                                                             -250
                                                                                                   400bp, respectively.
    05/05 08/05 11/05 02/06 05/06 08/06 11/06 02/07 05/07 08/07 11/07 02/08 05/08
                                                                                                   Exhibit 282
Source: Morgan Stanley Research, Bloomberg
                                                                                                   Turkey Hiking Cycle (I) – Emergency Hikes
Given the huge differences between the two hiking
cycles, timing the flattening trend is harder to call in                                             18.0                                                                            50
                                                                                                                                                                                     0
South Africa. Our economists are forecasting 100bp of hikes                                          17.5
                                                                                                                                                                                     -50
in 2011, with the first 50bp hike in 2Q11. Unlike the previous                                       17.0
                                                                                                                                                                                     -100
two cycles, the SARB is unlikely to embark on a prolonged                                            16.5
                                                                                                                                                                                     -150
monetary tightening cycle, we believe, especially with inflation                                     16.0
                                                                                                                   TU Repo                                                           -200
expected to stay within the 3-6% target range over the next 18                                       15.5
                                                                                                                   TRY 2s5s                                                          -250
months. We therefore think that it is still too early to be                                          15.0          TRY 2s10s
                                                                                                                                                                                     -300
positioned in flatteners now. Any flattening in the near                                             14.5
                                                                                                                                                                                     -350
term will most likely be driven by positive risk sentiments                                          14.0                                                                            -400

driving duration-buying at the mid-to-long end of the                                                13.5                                                                            -450

curve.                                                                                               13.0                                                                             -500
                                                                                                         05/06   05/06   06/06   06/06   06/06   06/06   06/06   07/06   07/06   07/06

Exhibit 281
                                                                                                   Source: Morgan Stanley Research, Bloomberg
South Africa: Cumulative Change in Term Spreads                                                    The second hiking cycle started in May 2008, when the CBT
                        150
                                                                                                   hiked 50bp. This was followed by another two 50bp hikes in
                        100                     ZAR 2s5s (I)         ZAR 2s10s (I)                 the two subsequent months, and the hiking cycle was cut
                                                ZAR 2s5s (II)        ZAR 2s10s (II)
                         50                                                                        short by the sell-off in September 2008. Interestingly, despite
                          0                                                                        USD/TRY moving from 1.15 to 1.69 during the sell-off, the
  -15     -10      -5          0   5       10       15          20      25       30
                         -50                                                                       CBT kept rates on hold, unlike the previous emergency hiking
                        -100                                                                       cycle in 2006.
                        -150
                                                                                                   The curve started flattening 1 month before the first hike, with
                        -200
                                                                                                   peak-to-trough flattening of 145bp and 235bp for 2s5s and
                        -250                                                                       2s10s, respectively. Another point worth highlighting is that
Y-axis shows the change in curve spreads from the date of the first hike (set to be zero)          most of the flattening this time occurred prior to the first hike
Source: Morgan Stanley Research, Bloomberg
                                                                                                   (see Exhibit 282-6), as the market priced in significant
                                                                                                   amounts of rate hikes very quickly even before the CBT
                                                                                                   started hiking rates.




                                                                                                                                                                                         143
                                                                                                          MORGAN STANLEY RESEARCH

                                                                                                          March, 2012
                                                                                                          EM Local Markets Guide




Exhibit 283                                                                                               Based purely on the timing of the rate hiking cycle, we may
Turkey Hiking Cycle (II)                                                                                  not be far from initiating flattening trades in Turkey, as our
  17.00                                                                                        0
                                                                                                          economists are expecting the rate hike to start in 1Q11. The
                                                   TU Repo
                                                   TRY 2s5s
                                                                                                          recent flattening happened because the market removed
  16.75
                                                   TRY 2s10s                                   -50        some inflation premium further out on the curve, as inflation
  16.50
                                                                                                          expectations continued to inch lower. At the same time, the
                                                                                               -100
  16.25                                                                                                   continued improvement in Turkey’s fiscal outlook also
  16.00                                                                                        -150       supports the argument for a flatter curve.
  15.75
                                                                                               -200       Given the sound fundamental picture and light foreign
  15.50                                                                                                   positioning in local bonds, coupled with TRY real rates being
                                                                                               -250
  15.25                                                                                                   among the highest in the CEEMEA region, we think it is
  15.00                                                                                        -300       possible that the recent flattening momentum could continue
      03/08    03/08   04/08   04/08   05/08       05/08   06/08   06/08   07/08       07/08              in the near term. Going forward, we would recommend
Source: Morgan Stanley Research, Bloomberg
                                                                                                          fading any steepening moves, and starting to accumulate
Exhibit 284                                                                                               flattener positions as we approach the start of the hiking
Turkey: Cumulative Change in Term Spreads                                                                 cycle, possibly in 1Q11.
                                          200                 TRY 2s5s (I)             TRY 2s10s (I)
                                          150                 TRY 2s5s (II)            TRY 2s10s (II)

                                          100

                                           50

                                               0
  -3             -2             -1                 0               1               2                  3
                                          -50

                                         -100

                                         -150

                                         -200

                                         -250

                                         -300


Y-axis shows the change in curve spreads from the date of the first hike (set to be zero)
Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                     144
                                                                         MORGAN STANLEY RESEARCH

                                                                         March, 2012
                                                                         EM Local Markets Guide




How to Value EM Breakeven Inflation?
Juha Seppala
                           th
First Published May 27 , 2011: How to Value EM Breakeven Inflation

In a previous focus piece, see EM Quantitative Strategy                   Exhibit 286
Update: Global EM Inflation and Asset Prices (March 16,                   Mexico: Market vs. Fair BE (R2=0.713)
2011), we studied how global EM inflation trends translate                 8
from asset class to asset class. We showed how the best                                                                    Market BE          Fair BE

predictor of future inflation in EM countries is what we called
                                                                           7
the external versus local rate difference – the relative value
of local debt rates versus external rates for issuers of the
                                                                           6
same credit risk. We also showed how EM inflation moves
from oil prices first to the external vs local debt spread, and
then to different asset classes. Finally, we used these results            5
to construct a relative-value model for EM breakeven
inflation.                                                                 4

In this part two of the exercise, we have extended our simple
                                                                           3
model by selecting the optimal lag length based on the best
fit in the historical data. We also did more in-depth analysis
                                                                           2
of the statistical significance of explanatory variables and                Jun-07   Dec-07   Jun-08   Dec-08   Jun-09   Dec-09   Jun-10   Dec-10   Jun-11
extensive back-testing. Exhibit 347 uses Brazil as a baseline             Source: Morgan Stanley Research, Bloomberg
example and lists the best set of regressors found: the
lagged level of 5y breakeven rate, Brazil CPI, US CPI, the                In Exhibit 286 we display the model fair value and the market
difference between local and external Brazil yields, and oil              5y breakeven inflation rate. Upon examining the picture, the
price. Unlike in the previous version, we now use the same                model seems to have some predictive power: When the fair
lag for oil price as for other variables.                                 value line is above the market line, breakeven inflation tends
                                                                          to go up during the next 2-3 months, and when it is below,
We ran independent regressions using weekly data for
                                                                          the opposite tends to happen. However, to formally assess
Poland, South Africa, Brazil, Chile, Mexico, and Korea. The
                                                                          the how accurate the model is in generating
chosen specifications were all very similar to each other. The
                                                                          recommendations, in the next section we study the results
optimal lag length was 13 weeks on average, and for most
                                                                          through an extensive back-testing exercise that we applied to
countries the chosen explanatory variables were the same
                                                                          the model.
as for Brazil (that is, compared to the previous analysis, the
US breakeven rate was omitted).                                           Back-Testing the Model

Exhibit 285                                                               In order to test our model recommendations, we applied the
Brazil: Regression of 5y BE (t)-BE (t-13 weeks)                           following scheme. We first calculate the spread difference
(R2=0.542)                                                                between our fair value estimate and the current value of the
Variable                        Coefficient   White S.E.   t-Statistic    5y breakeven inflation rate. Next we calculate the 90-day
Constant                               1.57         0.20         8.02     rolling standard deviation of this spread. A buy (sell) signal is
BE(t-13w)                             -0.58         0.06        -9.97     triggered whenever the spread goes above (below) one unit
CPI(t-13w)                             0.19         0.03         6.35     of standard deviation. The trade is closed when the spread
USCPI(t-13w)                          -0.18         0.01       -16.66     crosses through zero and changes sign. Exhibit 287
diffRates(t-13w)                      -0.04         0.02        -1.59
                                                                          presents a diagram to show how this automated strategy
Oil(t-13w)                             0.01         0.00         6.79
                                                                          works, and in Exhibit 288 we show trades generated in this
Source: Morgan Stanley Research, Bloomberg
                                                                          manner for South Africa. The hit ratio is remarkable 100%.




                                                                                                                                                        145
                                                                                         MORGAN STANLEY RESEARCH

                                                                                         March, 2012
                                                                                         EM Local Markets Guide




Exhibit 287                                                                               Exhibit 289
Trading System Diagram                                                                    South Africa 5y BE Spread and Current Value
                                                                                          10.00             Std range              Spread             BE              0.60
              Mar ket BE

                                                                                                                                                                      0.50
                                                                                           9.00
                                    Spread     Spre ad           12 w rolling
               Difference                                                                                                                                             0.40
                                                                   Std Dev
                                                                                           8.00
                                                                                                                                                                      0.30

                Fair BE       Spread>1*Stdev                                               7.00                                                                       0.20

              Buy Tr ig ger                     Spread
                                                                                                                                                                      0.10
                                                                                           6.00
                                                        Spread<-1 *Stdev
                                                                                                                                                                      0.00
                                             Sell T rigger                                 5.00
                                                                 Trigger Generator                                                                                    -0.10

                                                                                                               2-Feb-07                          5-Sep-08
                                                                                           4.00                                                                       -0.20
                                                                                              Jan-07       May-07         Sep-07        Jan-08      May-08   Sep-08
                                   Spread< 0       Spread>0
                    Long                                              Short               Source: Morgan Stanley Research, Bloomberg

                                                                                          As mentioned, in Exhibit 288 we listed all model-generated
                 Buy                                                 Sell                 trades for South Africa. In Exhibit 289 we concentrate on one
                                             Neu tral
                                                                                          South Africa trade, the long position from February 2, 2007,
                                                             Position Management
                                                                                          to September 5, 2008, to illustrate how the system works. It
Source: Morgan Stanley Research
                                                                                          shows how the system opened a long position when the dark
Exhibit 288                                                                               blue line (spread) rose above the gray area. This trade was
South Africa: 5y BE Trades (Sharpe Ratio = 1.04)                                          closed 25 weeks later when the blue line fell below the zero
Open                      Close                         Long/Short              Return    line, resulting in a profit of 6.08%. Please note that had the
3-Jun-05                  17-Jun-05                       Short                 0.40%     trade been closed on June 28, 2008, at the peak of BE rate,
29-Jul-05                 9-Sep-05                         Long                 1.28%     the profit would have been 14.9%. In other words, while our
30-Sep-05                 4-Nov-05                         Long                 0.30%
                                                                                          simple trading system is very illustrative, one can be
11-Nov-05                 6-Jan-06                        Short                 2.44%
13-Jan-06                 3-Mar-06                         Long                 0.69%     definitely improve its results by making sharp judgment calls.
10-Mar-06                 7-Apr-06                        Short                 0.47%
                                                                                          Exhibit 290 shows the statistics for the current trade returns
14-Apr-06                 19-May-06                        Long                 1.36%
2-Jun-06                  16-Jun-06                        Long                 1.82%
                                                                                          for all countries to which we have applied our methodology.
23-Jun-06                 14-Jul-06                       Short                 0.36%     Based on these results, the model and trading strategy
25-Aug-06                 1-Dec-06                        Short                 1.75%     seems to work quite nicely for all countries with the exception
12-Jan-07                 26-Jan-07                       Short                 0.16%     of Korea.
2-Feb-07                  5-Sep-08                         Long                 6.08%
10-Oct-08                 31-Oct-08                       Short                 0.27%     Exhibit 290
7-Nov-08                  23-Jan-09                       Short                 4.89%     Historical Recommendations – Statistics
3-Apr-09                  25-Sep-09                        Long                 2.23%
                                                                                                         Total Returns             Sharpe         Sortino Hit Ratio
30-Oct-09                 13-Nov-09                        Long                 0.20%
4-Dec-09                  2-Apr-10                        Short                 2.34%
                                                                                          Poland                 24.70               0.61            2.12      76%
7-May-10                  10-Sep-10                       Short                 4.09%     S Africa               35.47               1.04              Inf    100%
17-Sep-10                 8-Oct-10                        Short                 0.62%     Brazil                 25.50               0.75            7.05      87%
15-Oct-10                 13-May-11                        Long                 4.25%     Chile                  26.82               0.57            2.02      77%
Source: Morgan Stanley Research, Bloomberg
                                                                                          Mexico                 36.30               0.70            5.92      94%
                                                                                          Korea                  11.93               0.40            1.21      50%
                                                                                          Source: Morgan Stanley Research, Bloomberg




                                                                                                                                                                  146
                                                                               MORGAN STANLEY RESEARCH

                                                                               March, 2012
                                                                               EM Local Markets Guide




Exhibit 291                                                                     Current Fair – Value Estimates
Strategy vs. Market – Weekly Returns
                                                                                In Exhibit 9 we summarize the latest fair value assessments
                              Average                       Hit         Max
                              Returns    Sharpe Sortino   Ratio   Drawdown      for all countries. According to the model, BE rates in Latin
Poland          Strategy         5.88      1.62    3.14    59%         -1.67    America are currently undervalued while Poland is slightly
              Buy n' Hold        1.26      0.33    0.43    53%         -3.72    overvalued. South Africa and Korea appear fairly valued.
S Africa        Strategy         7.47      1.50    2.50    61%         -3.35
              Buy n' Hold        0.89      0.18    0.25    48%         -3.35    Exhibit 292
Brazil          Strategy         6.29      1.37    2.35    59%         -1.78
                                                                                Exhibit Headline
              Buy n' Hold        0.27      0.06    0.08    51%         -4.09
Chile           Strategy         5.38      1.10    1.81    56%         -2.13                                   5y BE          Fair Value   Difference
              Buy n' Hold        0.34      0.07    0.10    50%         -2.13
Mexico          Strategy        11.93      1.46    5.42    60%         -1.17    Poland                          2.92                2.79          -13
              Buy n' Hold        0.08      0.01    0.01    50%        -10.97    S Africa                        6.31                6.31            0
Korea           Strategy         3.44      0.95    1.64    58%         -1.73
              Buy n' Hold        1.69      0.47    0.64    57%         -3.09    Brazil                          5.58                5.89           31
Source. Morgan Stanley Research, Bloomberg
                                                                                Chile                           3.46                3.85           39
To show that the these results are not driven by long-term                      Mexico                          3.71                4.29           57
trends in EM inflation, in Exhibit 8 we display weekly returns                  Korea                           2.52                2.52            0
                                                                                Source.: Morgan Stanley Research, Bloomberg
obtained by following the strategy and by a simple buy-and-
hold strategy. The advantage of our strategy over the market
is very significant.




                                                                                                                                                 147
                                                                                     MORGAN STANLEY RESEARCH

                                                                                     March, 2012
                                                                                     EM Local Markets Guide




SGD: The New S$NEER Model
Yee Wai Chong
                                         th
First Published February 24 , 2012: The New S$NEER Model

Building a S$NEER Model with Limited Information                                      The rival candidates are shown in Exhibit 293 and their
                                                                                      component currency weights are shown in Exhibit 294. The
Constructing a S$NEER model is a challenge. The problem
                                                                                      key distinction between the two candidates is that the USD
is a lack of public information about this MAS policy tool. For
                                                                                      weight in S$NEER(Regression) (27.6%) is significantly larger
example, the MAS does not publish the currencies, nor their
                                                                                      than that in S$NEER(TWI) (11.6%).
weights, in the index. Admittedly, the MAS releases its index
to the market, but it does so with a lag of up to six months,                         Exhibit 294
which makes tracking the S$NEER in real time challenging.                             Top Five Currencies in of the S$NEER Candidates
Moreover, there is even greater uncertainty about the policy                               (Weight, %)
                                                                                      80
band the S$NEER trades within; for the MAS states its
S$NEER policy in qualitative terms only (e.g., a “modest and                          70
gradual appreciation” slope). Accurately replicating the                                                                                              IDR


MAS’s S$NEER index and its band is therefore problematic                              60
                                                                                                             IDR                                      EUR
as there are too many unknowns for market participants to                             50
model the MAS’s S$NEER policy perfectly. Our solution to                                                     EUR


this limited-information problem: reverse-engineering and                             40
                                                                                                                                                     USD

optimization.                                                                                                USD

                                                                                      30

Candidates for the S$NEER Index                                                                              CNY
                                                                                      20

To determine a suitable proxy for the MAS’s S$NEER index,
                                                                                                                                                      CNY


                         29                                                           10
we tested two candidates :                                                                                   MYR
                                                                                                                                                     MYR

                                                                                       0
1) S$NEER(TWI): A trade-weighted index composed of the                                                   S$NEER (TWI)                         S$NEER (Regression)

currencies of Singapore’s leading trading partners;                                   Source: Bloomberg, Morgan Stanley Research


2) S$NEER(Regression): An index composed of currencies,                               Testing the S$NEERs
whose weights in the index were determined by reverse-
engineering (i.e., from regressing the official S$NEER on a                           We tested the candidate S$NEERs to find which best fits the
range of currencies).                                                                 MAS S$NEER out of sample. Our criteria for fit were
                                                                                      correlation and Root Mean Square Forecast Error (RMSFE).
Exhibit 293
                                                                                      Exhibit 295 shows the summary results. We find that the
The S$NEER Model Candidates                                                           S$NEER (Regression) model most closely fitted the
 120
         (S$NEER, Jan 2000 = 100)
                                                                                      S$NEER on both measures and is therefore chosen as our
                S$NEER (TWI)

                S$NEER (MAS)
                                          S$NEER (Regression)
                                                                                      S$NEER proxy.
 115
                                                                                      Exhibit 295

 110                                                                                  Table of Test Results for the S$NEER Candidates
                                                                                            Sample period:              Weekly Correlation with     Root Mean Square
 105                                                                                          2007-2011                     MAS S$NEER               Forecast Error
                                                                                             S$NEER (TWI)                        0.80                        0.95
 100                                                                                    S$NEER (Regression)                      0.85                        0.77
                                                                                      Source: Bloomberg, Morgan Stanley Research
     95
      Jan-99      Jan-01        Jan-03        Jan-05      Jan-07   Jan-09   Jan-11

Source: Bloomberg, Morgan Stanley Research


29
 See Appendix A and B for greater detail on their construction.


                                                                                                                                                                    148
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




The Band Parameters                                                 Given these problems, we choose another methodological
                                                                    approach. Rather than let EE models determine the trend,
The MAS can adjust its S$NEER policy in three ways, via: 1)
                                                                    we prefer to let the data speak to us more directly: We use a
the band slope; 2) the band width; and 3) re-centering the
                                                                    moving average (MAV). Such a method is flexible enough to
band. Since the MAS has re-centered the S$NEER band no
                                                                    avoid the problems the EE models have as the MAV will pick
more than six times since it started publishing this
                                                                    up the appropriate slope from the S$NEER’s price action
information in 2001, determining the band slope (1) and the
                                                                    over time, constantly updating itself with the latest
band width (2) are arguably the most important parameters
                                                                    information. An extra benefit is that the MAV will be able to
to get right for our proxy model.
                                                                    smooth out over time the effects of any error in our
(1) Constructing the S$NEER Slope                                   calculation of band re-centering (if the band is not re-
                                                                    centered at the prevailing spot).
How do you quantify the gradient of the S$NEER’s slope
with only qualitative information given by the MAS? Some            The Optimal MAV
have advocated using econometric methods to determine the           The question remains as to what is the optimal MAV
underlying slope trend from previous price action and then          window 30 to use to determine the MAS’s policy slope. A MAV
extrapolating the trend forwards. However, such an                  with a small window will be quick to pick up changes in the
econometric/extrapolation (EE) approach is flawed, in our           MAS’s policy slope (indeed, it may be able to pick up
view, for two reasons:                                              potential inter-meeting changes if they exist). However, a
                                                                    MAV window that is too small in size would swing the MAV
Reason 1: The Mapping Problem
                                                                    too much with the gyrations of the S$NEER, making the
It cannot be certain that the MAS always chooses the same
                                                                    slope trend noisy and unstable. A larger MAV window avoids
exact slope angle for each qualitative slope description (e.g.,
                                                                    this issue, producing a steady trend. However, it could be
the MAS may not always selects a slope, say, of 2%
                                                                    slow to adapt to any new slope regime.
annualized S$NEER appreciation when it states to the
market that it will pursue a “policy of modest and gradual          Exhibit 296
appreciation”). It could well be that the MAS selects its slope     Determining the Optimal MAV Window
from a range of gradient possible angles, but which are all                   (S$NEER, Jan 2000 = 100)
                                                                     125
still consistent with its qualitative description (e.g., the MAS
selects a particular slope from a range of 1.5-2.5%                  120
annualized S$NEER appreciation, according to the degree of                            Increasing moving average window
                                                                     115
inflation/growth risks at the time of its policy decision). EE
methods break down in such situations as they cannot                 110
functionally deal with ‘mapping’ a range of slope angles to a
                                                                     105
single qualitative description. One could argue that the
standard error of a regression slope could provide such a            100
range, but then that leaves the modeller faced with the
                                                                         95
problem of arbitrarily choosing a particular slope candidate.
                                                                         90
Reason 2: The Problem of Unprecedented                                    Jan-99       Jan-01        Jan-03   Jan-05     Jan-07      Jan-09      Jan-11
Announcements                                                        Source: Bloomberg, Morgan Stanley Research
EE methods also cannot cope with determining the correct
slope angle if the MAS states a new, unprecedented                  (2) Constructing the S$NEER Policy Bands
qualitative slope angle in its policy statement. For example,
                                                                    The MAS directs the S$NEER in a policy band. However, it
on October 14, 2010, the MAS announced for the first time
                                                                    does not reveal the width of the band to the market. The only
ever that it would increase the slope beyond the modest and
                                                                    information it does reveal is when it has narrowed or
gradual appreciation trend. With no precedent or data to
                                                                    widened the band and whether it is in the upper or lower half
model such a new slope regime, how can the EE method
                                                                    of the band. With so little information, we tested a large
determine the new slope ex ante? As with reason 1, the
                                                                    range of band widths to find an optimal band that:
modeler is forced to choose arbitrarily.

Listening to the Data
                                                                    30
                                                                         The MAV window refers to the number of observations included in the moving average.


                                                                                                                                                       149
                                                                                                   MORGAN STANLEY RESEARCH

                                                                                                   March, 2012
                                                                                                   EM Local Markets Guide




a) Minimized the number of times the S$NEER breached the                                            optimization of the S$NEER’s band slope and band width.
band.                                                                                               We did this with a grid search in which for every MAV
                                                                                                    window for the S$NEER band slope that we selected, we
b) Maximized the number of times the S$NEER was close to
                                            31                                                      tested it with a variety of band widths, incremental band
the policy bands whenever the MAS intervened .
                                                                                                    widening and re-centering parameters. The merits (or score)
c) Best fitted the MAS’s statements that the S$NEER was in                                          for each slope/band/band increment/re-centering
                                                                                                                                               34
the upper or lower half of the band.                                                                combination were based on four criteria :

Incremental Band Widening                                                                           1. Slope: The slope gradient corresponded with the MAS’s
The testing is complicated by the fact that the band width is                                       qualitative guidance outlined in its policy statements.
not constant over time. Occasionally, the MAS states that it
                                                                                                    Exhibit 298
has widened or narrowed the band. We determine this
                                                   32                                               Slope Test
incremental band change using the above criteria .

Exhibit 297
                                                                                                                           Slope according
Determining the Optimal Band Width                                                                                         to MAS
                                                                                                                           statement
          (S$NEER, Jan 2000 = 100)
 125

 120
                       Increasing initial band width
 115
                                                                                                                                                                MS moving average slope

 110

 105                                                                                                Source: Bloomberg, Morgan Stanley Research


 100                                                                                                2. Breaches: The number of times S$NEER breached the
                                                                                                    band.
     95
                                                                                                    Exhibit 299
     90                                                                                             Breach Test
      Jan-99       Jan-01        Jan-03   Jan-05       Jan-07       Jan-09        Jan-11
Source: Bloomberg, Morgan Stanley Research


(3) Re-centering
The MAS occasionally re-centers the S$NEER band. When it
does this, the MAS usually moves the band to the prevailing
S$NEER rate, which does not present much of problem for
our testing. However, the MAS has on one occasion not re-
centered at the prevailing rate. We therefore test for the size
of this incremental re-centering 33.
                                                                                                    Source: Bloomberg, Morgan Stanley Researc


Grid Search and the Quadruple Optimization                                                          3. Intervention: The closeness of the S$NEER to the band
Problem                                                                                             during periods of MAS intervention.

Solving for the optimal band slope and the optimal band
width can be done separately, of course. However, the
optimal S$NEER model is best solved with the simultaneous


31
   After all, the market participant is only aware that the S$NEER is close to the band when
the MAS intervenes.
32
   See Appendix C.
33
   With a sample of one, our results can only be tentative. However, as such re-centerings
are rare, getting this parameter right does little to impact the efficiency of our modeling. See
                                                                                                    34
Appendix C for the test.                                                                                 See Appendix C for more details of the score system for the grid search.


                                                                                                                                                                                     150
                                                                                            MORGAN STANLEY RESEARCH

                                                                                            March, 2012
                                                                                            EM Local Markets Guide




Exhibit 300                                                                                  Results of the Grid Search
Intervention Tests
                                                                                             The results for our grid search for the S$NEER(Regression)
                                                                                             are shown in Exhibit 302. The darker zones represent the
                                False negative
                                (far from the
                                                                                             band slope/width combination with the higher optimal score.
         Close to the
         band,
                                band,
                                intervention)
                                                                                             The optimal setting is found with a S$NEER slope defined by
         intervention
                                                                                             a 18M MAV window, with +/-3% initial 35 policy band, +/-
                                                                                             0.25% band widening/narrowing increment and a 75% re-
                                                                                                       36
                                                                                             centering . Owing to the incremental band changes over the
                                   Heavy intervention
                                                                                             years, the current policy band width is 3.25%.
                                                              False positive
                                                              (close to the band,
                                                              no intervention)               The Grid Search for the S$NEER(TWI) Model
                                                                                                                                           37
                                                                                             For completeness, Exhibit 303 shows the grid search for
Source: Bloomberg, Morgan Stanley Research
                                                                                             the S$NEER(TWI) model. In this case, the optimal band has
                                                                                             a smaller MAV window for its slope (12M) and a narrower
4. Upper bound/lower bound: The S$NEER’s location in                                         initial band (+/-2.5%) than that for S$NEER(Regression), but
the S$NEER band corresponding to the MAS’s qualitative                                       with a similar band widening/narrowing increment (+/-0.25%)
guidance outlined in its policy statements.                                                  and re-centering (75%).
Exhibit 301
                                                                                             Exhibit 303
Upper/Lower Band Test                                                                        The Grid Search for the S$NEER(TWI)
                                                                                                                                                                          Initial band w idth
                                                                                                                                                                                6.0%


                                                                                                                                                                                5.5%


                                                                                                                                                                                5.0%


                                                                                                                                                                                4.5%



                                                                                                                                                                                4.0%
         MAS statement says
         SNEER in upper half      MAS statement says           SNEER in wrong half
         of band                  SNEER in lower half          of band                                                                                                          3.5%
                                  of band
                                                                                                                                                                                3.0%
                                                                                                                              optimal model
Source: Bloomberg, Morgan Stanley Research
                                                                                                                                                                                2.5%


Exhibit 302                                                                                                                                                                     2.0%

The Grid Search for the S$NEER(Regression)
                                                                      Initial band w idth                         old model                                                     1.5%
                                                                               6.0%

                                                                                                                                                                                1.0%
                                                                              5.5%                   2m      3m     6m      9m      1y     18m      2y     30m     3y      4y
                                                                                                  MA w indow


                                                                              5.0%
                                                                                             Source: Morgan Stanley Research; Note: For this heat map, the width increment and re-
                                                                                             centering parameters are already set at the optimal levels.
                                                                              4.5%


                                                                              4.0%


                                                                              3.5%
                                   optimal model
                                                                              3.0%


                                                                              2.5%


                                                                              2.0%
                                                                                             35
                                                                                                We choose the “initial policy band” width at the start of the data sample (January 2001)
                                                                              1.5%           as our parameter in our grid search. This initial band width then changes with increments of
                                                                                             0.25% according to our optimization analysis. The current band width is 3.25%.
                                                                                             36
                                                                              1.0%              That is, when the MAS does not re-center the band to the prevailing spot, we find that the
      2m      3m    6m     9m       1y    18m    2y     30m      3y      4y                  MAS moves the center of the band so that it closes 75% of the gap between the prevailing
   MA w indow                                                                                spot and the old band center. We test the stability of this optimal point over varying time
Source: Morgan Stanley Research; Note: For this heat map, the width increment and re-        periods in Appendix D.
                                                                                             37
centering parameters are already set at the optimal levels.                                     The Exhibit also shows the suboptimality of our old S$NEER model, which used a TWI
                                                                                             construct and had a 6M MAV for its slope and +/-1.75% for its initial bandwidth.


                                                                                                                                                                                       151
                                                                               MORGAN STANLEY RESEARCH

                                                                               March, 2012
                                                                               EM Local Markets Guide




Decomposing the Errors                                                          Bottom Line
While these S$NEER(TWI) and S$NEER(Regression)                                  We introduce our new S$NEER model. It includes a new
models have been optimized, they are not without errors with                    S$NEER index and an optimized band with a 18M MAV
regards slope angle, breaches etc. Exhibit 304 shows the                        slope gradient, a +/-3% initial policy band, +/-0.25% band
decomposition of these errors for these models along with                       widening/narrowing increment and a 75% re-centering. The
our old S$NEER model. We can see that both optimal                              current policy band width is 3.25%
models have substantially fewer errors than the old model.
However, the optimal S$NEER(Regression) model was a
further 57% more efficient than the S$NEER(TWI) model in
terms of error reduction.

Exhibit 304
Decomposition of Model Errors
     (Error Score)
20                               Intervention errors, false negatives
18                               Intervention errors, false positives
16                                Slope errors
14                               Upper/lower band errors
12                                  Breaches
10
 8
 6
 4
 2
 0
          Old S$NEER (TWI)      Optimal S$NEER (TWI)          Optimal S$NEER
                                                               (Regression)
Source: Bloomberg, Morgan Stanley Research




                                                                                                                                       152
                                                                              MORGAN STANLEY RESEARCH

                                                                              March, 2012
                                                                              EM Local Markets Guide




A Primer in the EM FX Probability Analyzer
Yee Wai Chong
                                         th
First Published September 12 , 2011: EM - The FX Probability Analyzer

We introduce the ‘FX Probability Analyzer’ (FXPA) for the EM                   extract from FX option prices are the ‘risk-neutral’
currencies – a model that backs out the risk-neutral                           probabilities – i.e., the perceived distribution of market
probability distribution of exchange rates implied by the FX                   outcomes held by a risk-neutral investor rather than a risk-
options market. The EM FXPA is the expansion of our FXPA                       averse or risk-loving investor.
for the AXJ currencies, which we launched in 2009 (see AXJ:
                                                                               Comparison of the Risk-Neutral and Actual
The FX Probability Analyzer, June 26, 2009).
                                                                               Probability Distribution
Assessing the risk implied in option prices can be highly                      A key attribute of the risk-neutral probability distribution is
enlightening. Indeed, such forward-looking information can                     that it is centered around the ATMF. This may not be the
help us identify where there is opportunity and where there is                 case for the distribution of risk-averse or risk-loving investors
danger in the markets, in our view.                                            and so interpretation of our probability results will have to be
The concept of extracting probability distributions implied in                 taken with this caveat in mind. That said, the higher
option prices goes back to the early 1990s 38. The level of                    ‘moments’ of the risk-neutral distribution – the degree of tail
sophistication of modeling these probabilities has been                        fatness and skew – mirror that of the actual probability
refined over time. But in spite of the ongoing improvements,                   distribution.
backing out the probabilities is still hampered by the                         How to Use the FX Probability Analyzer
distortions created by risk-aversion.
                                                                               We believe that the FX Probability Analyzer will be very useful
The Problem of Risk-Aversion                                                   in several respects. First, it can provide a timely and accurate
Investors’ trading decisions are not just shaped by their                      gauge of market sentiment. The analyzer not only shows
expectation of future returns, but also by their risk appetite.                graphically the degree of price dispersion and skew over a 1M-
Most market participants are risk-averse and so demand to                      12M period, but also displays the risk-neutral probabilities of
be compensated with a premium for bearing uncertainty. On                      the market attaining certain key market levels. For example, if
the other hand, a few participants are ‘risk-seeking’ and are                  we sense a certain exchange rate level is being protected
even prepared to pay a premium to bear uncertainty. Risk-                      because there is an option barrier there or that the central
neutral participants are in between, requiring no premium to                   bank is intervening at that level, we can assess the market’s
bear risk. An added complication is that many investors can                    probability that spot will break through this level. Alternatively,
swing from being risk-averse to risk-neutral or risk-seeking,                  the probability grid may help in providing some guidance of the
depending on the direction of markets. Unfortunately, unless                   profit potential of certain trades.
the current degree of market risk-aversion is known, an                        Exhibit 305
observer of the option price action is unable to distinguish                   USD/MXN Call Spread Strategies
between changes in risk-aversion and changes in markets’
                                                                                     (Payoff at Maturity, %)
actual distribution of future market outcomes.                                   9
                                                                                 8                             P(Spot > 13.5 = 12.5%)
The Solution: The Risk-Neutral Probability                                       7
Risk-aversion factors thus prevent us from ‘backing out’ the                     6
markets’ actual distribution of future market outcomes.                          5                  P(Spot > 13.0 = 18.6%)

However, a partial way around the problem is to impose the                       4

assumption that all market participants are risk-neutral. It so                  3

happens that the pricing of derivative products makes the                        2

same assumption. (Quantitative analysts impose the                               1
                                                                                 0
assumption in order to establish the condition of no-arbitrage
                                                                                -111.0       11.5       12.0        12.5       13.0     13.5   14.0   14.5    15.0
in their pricing models.) Thus, the probability distribution we
                                                                                -2
                                                                                -3
38
     See Bates (1991, 1996) and Malz (1996, 1997) in the reference section.    Source: Morgan Stanley, FXPA


                                                                                                                                                             153
                                                                                     MORGAN STANLEY RESEARCH

                                                                                     March, 2012
                                                                                     EM Local Markets Guide




Furthermore, this approach can be extended to the options                             The Probability Graph (Bottom Right Quadrant)
market. For example, the FXPA can be used to determine                                The bottom right-hand quadrant contains the implied
the appropriate strike levels in option spread strategies. For                        probability density function for the 1M, 3M and 12M horizons.
example, the FXPA can help us set the far strike of a 3M                              The density functions of the longer tenors tend to have lower
USD/MXN call spread. A 12.295/13.00 call spread costs                                 peaks and are more spread out than those functions of
1.62%, while a 12.295/13.50 call spread costs 2.26% (given                            shorter horizons (see Exhibit 307). Note also that when there
the higher potential for upside). However, our FXPA                                   is a heavy skew in one direction, more probability weight is
indicates that the probability of even reaching 13.00 (or                             allocated to that side of the tail. However, this has the
higher) is only 18.6% and the probability of 13.50 (or higher)                        tendency to tilt the peak of the distribution in the other
is 12.5%. This probability information may therefore help us                          direction of the skew.
decide on which spread we may prefer (see Exhibit 310).
                                                                                      Exhibit 307
                                                                                      USD/MXN: Implied Probability Density Function
Finding Your Way Round the Analyzer
                                                                                                 Density
For the discussion below, readers can refer to Exhibit 311 for                        1.00
a sample of the FXPA screen. The FXPA comprises four
quadrants. The top-left quadrant contains the input table for                         0.80

the FXPA. The FXPA uses values from Morgan Stanley’s
options database as default inputs, but they can be manually                          0.60

changed for scenario analysis.
                                                                                      0.40

The Probability Grid (Top Right Quadrant)
The top right-hand quadrant holds the FX probability grid.                            0.20

The first column in the grid displays a range of spot values.
                                                                                      0.00
The other columns contain the risk-neutral probability                                    6.00             8.00     10.00          12.00        14.00     16.00
                                                                                                                                                                      FX
                                                                                                                                                                   18.00
distribution for each 1M, 2M, 3M, 6M and 12M tenors. Thus,                                                                  1m          3m          12m



each cell grid holds the risk-neutral probability of spot
                                                                                      Source: Morgan Stanley, FXPA
exceeding the specified exchange level at each given tenor
horizon. For example, the market is pricing in the probability
                                                                                      The Volatility Graph (Bottom Left Quadrant)
of 18.6% that USD/MXN will exceed 13.00 in three months’
                                                                                      The bottom-left quadrant fits the spot’s 25-75 delta volatility
time (see Exhibit 306). The shaded gray area in the grid
                                                                                      cone onto the end of the exchange rate’s previous two-year
marks the range above the ATMF.
                                                                                      price history. The cone roughly equates to defining the
Exhibit 306                                                                           middle 50% of the exchange rate’s probability distribution
USD/MXN: Implied Probability Table                                                    (see Exhibit 308).
                                       Breakthrough Probability***
     Level          1m           2m                3m                  6m     12m     Exhibit 308
     18.00         0.0%         0.0%              0.1%                1.1%    4.2%
     17.50
     17.00
                   0.0%
                   0.0%
                                0.0%
                                0.0%
                                                  0.2%
                                                  0.3%
                                                                      1.6%
                                                                      2.2%
                                                                              5.0%
                                                                              6.0%
                                                                                      USD/MXN: 25-75 Delta Volatility Cone
     16.50         0.0%         0.1%              0.6%                3.0%    7.1%
     16.00         0.0%         0.3%              1.1%                4.1%    8.5%                (USD/MXN)
     15.50         0.0%         0.6%              1.8%                5.5%   10.0%    14.5
     15.00         0.1%         1.3%              3.1%                7.4%   11.7%
     14.50         0.5%         2.7%              5.1%                9.8%   13.6%
     14.00         1.6%         5.3%              8.1%               12.7%   16.2%    14.0
     13.50         4.7%         9.6%             12.5%               16.3%   20.7%
     13.00        11.7%        16.3%             18.6%               23.0%   27.9%
     12.50        27.5%        31.4%             32.8%               35.1%   61.7%    13.5
     12.00        27.3%        35.9%             40.3%               45.4%   47.1%
     11.50         2.5%         8.9%             13.4%               21.8%   29.4%
     11.00         0.0%         1.1%              2.9%                8.9%   16.6%
                                                                                      13.0
     10.50         0.0%         0.1%              0.3%                2.7%    9.0%
     10.00         0.0%         0.0%              0.0%                0.6%    4.1%
     9.50          0.0%         0.0%              0.0%                0.1%    1.6%
     9.00          0.0%         0.0%              0.0%                0.0%    0.5%
                                                                                      12.5
     8.50          0.0%         0.0%              0.0%                0.0%    0.1%
     8.00          0.0%         0.0%              0.0%                0.0%    0.0%
     7.50          0.0%         0.0%              0.0%                0.0%    0.0%    12.0
     7.00          0.0%         0.0%              0.0%                0.0%    0.0%
     6.50          0.0%         0.0%              0.0%                0.0%    0.0%
     6.00          0.0%         0.0%             0.0%                 0.0%    0.0%    11.5
Source: Morgan Stanley, FXPA
                                                                                      11.0
                                                                                        Sep-09             Mar-10   Aug-10         Mar-11       Aug-11    Feb-12   Aug-12
                                                                                                                        75 Delta     25 Delta   Forward

                                                                                      Source: Morgan Stanley, FXPA



                                                                                                                                                                    154
                                                                   MORGAN STANLEY RESEARCH

                                                                   March, 2012
                                                                   EM Local Markets Guide




Calculating the FXPA − A Six-Step Procedure
                                                                    Part 2: How Useful Is the FXPA?
We incorporate the work of Malz (1997) and Kahale (2005)
to model the volatility surface and then extract the implied        We test how accurate the EM FX options market is in
risk-neutral probability. There are six steps in the                predicting the future distribution of spot rates over various
procedure 39:                                                       tenors. If the options market is indeed a good forecaster,
                                                                    then it supports the usefulness of looking at such market
1. We select three screen quotes − ATM, 25 delta risk-              distributions with the aid of our EM FXPA. Interestingly, our
reversal and the 25 delta butterfly − which are used to             extensive testing indicates that the options market has a high
calculate the 25, 50 and 75 delta volatilities using standard       success rate in forecasting future spot distributions, thereby
formulas. These points provide the fixed points upon which          validating the usefulness of the FXPA. Moreover, our testing
we build our continuous volatility smile for each of the 1M,        highlights the areas of greatest perfection and imperfection in
2M, 3M, 6M and 12M tenors.                                          the EM options market, which we believe is useful in its own
                                                                    right.
2. We fit a quadratic curve through these volatility points to
build a volatility smile proposed by Malz (1997). From this
                                                                    Are the Markets Good at Probability?
smile, we calculate two extra implied volatility points – the 15
and 85 deltas.                                                      The EM FXPA is a model that extracts the market risk-
                                                                    neutral probability distribution of exchange rates from FX
3. We determine the corresponding strike prices for the five
                                                                    option quotes. As discussed in Part 1, such information can
delta points. That is, the five delta points are mapped from a
                                                                    be extremely useful in gauging market sentiment and setting
volatility-delta space into a volatility-strike price space.
                                                                    strike levels for option strategies. However, the usefulness of
4. These volatility/strike price reference points are then fed      this probability information implicitly rests on the assumption
into the standard Black-Scholes (B-S) equation to produce           that the options market is an accurate forecaster of the
the B-S option call prices.                                         realized distribution of spot exchange rates. In that regard,
                                                                    we test whether the probability distribution embedded in
5. An option price curve is then interpolated between the five      options market prices is statistically close to the realized
points using the method proposed by Kahale (2005). The              distribution.
curve is also extrapolated beyond the 15 delta points to
include all possible strike prices.                                 The Market Test – The Intuition
6. We then calculate the first and second derivatives of the        We adopt the testing methodology used by Christoffersen
interpolated option price curve, using the methodology based        and Mazzotta (2004). The intuition behind the test is that we
on Kahale (2005). The first derivative is equivalent to the         examine how the probability distribution embedded in option
implied cumulative probability distribution of the future           prices matches the shape of the realized distribution of spot.
exchange rate process. The second derivative is the                 An exact fit would indicate that the options market can
estimate of the implied probability density function.               perfectly anticipate the realized distribution of future spot
                                                                    rates. However, if the goodness-of-fit is poor, then this would
                                                                    infer that the options market is a less reliable forecaster. We
                                                                    define goodness-of-fit 40 in our test by determining whether
                                                                    the difference between the market-implied and realized
                                                                    probability distribution curves is statistically insignificant.




39
     See Appendix 1 for further details.
                                                                    40
                                                                         Please see Appendix 2 for a more rigorous explanation.


                                                                                                                                  155
                                                                                        MORGAN STANLEY RESEARCH

                                                                                        March, 2012
                                                                                        EM Local Markets Guide




Exhibit 309
                                                                                         Generating the Implied Probability Distributions
Comparing the Implied Market and Realized Probability
Distributions                                                                            Before testing can begin, the implied and realized data must
                                                                                         be collected. The market-implied probability distributions are
     Density
                                                                                         generated by our FXPA engine 41. The FXPA engine captures
       30%
                                                                                         a snapshot of the market’s implied probability distribution
       25%
                                     Market Implied Probability Distribution             every week starting from between Jan-99 to Jul-05 42, and
                                                                                         ending in Jun-11. Each snapshot includes the probability
       20%
                                                                                         distribution at various tenors (i.e., 1M, 2M, 3M, 6M and 12M).
                                                                                         Thus, for every weekly observation in our data set, the FXPA
       15%
                                              Realized Probability Distribution          creates the market-implied probability data for the 55
       10%
                                                                                         specified intervals.

        5%                                                                               Generating the Realized Probability Distributions
        0%                                                                               The data set for the realized probability distribution spans
                                                                               Strike
                                                                                         across Feb-99 to Jul-11 (i.e., it is phased slightly in advance
Source: Morgan Stanley
                                                                                         of the implied market data set since we have to compare the
                                                                                         market-implied probability distribution data at time ‘T’ with the
Making Testing Easier
                                                                                         realized probability distribution of T+1M, T+2M, …, T+12M.
Testing for goodness-of-fit is made easier if the distributions                          The construction of the realized probability distributions is a
are first of all transformed. That is, the probability                                   simple task of building histograms with similar intervals to our
distributions are reconstructed so that they no longer span                              market implied distributions (i.e., <5%, 5-15%, 15-25%, etc.)
across spot levels but across percentiles. Such a                                        and over the same tenors (i.e., 1M, 2M, 3M, 6M and 12M).
transformation changes the shape of the distribution from a
bell-like shape to a uniform-like shape, which is far easier to                          The Results
compare.
                                                                                         The results of our tests reflect favorably on our FXPA model,
                                                                                         in our view. Of the 770 probability intervals examined across
Breaking Up the Task into 55 Mini-Tests
                                                                                         14 EM currencies, 78% passed our goodness-of-fit tests 43.
We do not apply a single goodness-of-fit test for the whole                              As shown in Exhibit 310, nearly all currencies have passing
distribution. Instead, we divide the distribution into a series of                       rates above 70%. In particular, USD/CLP, USD/TRY and
intervals (<5%, 5-15%, 15-25%, …, 85-95%, >95%) and then                                 EUR/HUF have passing rates above 90%.
measure the goodness-of-fit for each of these sub-intervals.
                                                                                         Exhibit 312to 312 further show the graphical breakdown of
In addition, we apply this interval test procedure to probability
                                                                                         the results for every tenor of each currency. The graphs
distributions of various tenors (1M, 2M, 3M, 6M and 12M).
                                                                                         show the difference between the market-implied and realized
Thus, we have 55 intervals (11 intervals across five tenors)
                                                                                         probabilities within each interval of the distribution and their
to test the overall goodness-of-fit between market implied
                                                                                         respective 95% confidence interval bands. Thus, whenever
distributions and the realized spot distributions.
                                                                                         the confidence interval spans the zero x-axis, the difference
                                                                                         in market-implied and realized probabilities is statistically
                                                                                         insignificant 44 .



                                                                                         41
                                                                                            Based on the work of Malz (1997) and Kahale (2005).
                                                                                         42
                                                                                            The start dates are: USD/BRL (Jan-99), USD/MXN (Jan-99), USD/ARS (Sep-01),
                                                                                         USD/CLP (Jul-99), USD/COP (Jul-04), USD/PEN (Nov-04), USD/ZAR (Jan-99), USD/TRY
                                                                                         (Dec-04), USD/ILS (Jun-99), USD/RUB (Jun-03), EUR/PLN (Apr-00), EUR/CZK (Jan-99),
                                                                                         EUR/HUF (May-01), EUR/RON (Jul-05).
                                                                                         43
                                                                                            These results are particularly satisfactory, given that theoretically there is a systematic
                                                                                         bias to implied market forecasts. This is attributed to the problem of comparing risk-neutral
                                                                                         probabilities and physical realized probabilities. Please refer to Christoffersen and Mazzotta
                                                                                         (2004).
                                                                                         44
                                                                                            To be more rigorous, the null hypothesis that the market-implied and realized probabilities
                                                                                         are the same cannot be rejected.


                                                                                                                                                                               156
                                                                 MORGAN STANLEY RESEARCH

                                                                 March, 2012
                                                                 EM Local Markets Guide




Exhibit 310
                                                                  Bottom Line
Table of Results: Percentage of Total Tests Satisfying
the Null Hypothesis That the Market-Implied and                   Our new FX Probability Analyzer for EM currencies backs out
Realized Probability Distributions Are Indistinguishable          the risk-neutral probability distribution of exchange rates
                                                                  implied by the FX options market. We believe that it will
 Rank                          Currency                 Score
                                                                  prove useful for assessing general market sentiment as well
 1                             USD/CLP                  94.5%
                                                                  as determining appropriate exit strategies for spot as well as
 1                             USD/TRY                  94.5%     setting strike levels and barriers in option strategies.
 3                             EUR/HUF                  92.7%
                                                                  Our testing shows that the EM options market in general has
 4                             USD/ZAR                  89.1%
                                                                  a good track record in forecasting the distribution of the
 5                             EUR/PLN                  81.8%     future spot exchange rates. This therefore makes analysis of
 5                             EUR/RON                  81.8%     implied market probabilities with our FXPA highly valid. We
 7                             USD/ARS                  78.2%     also note that several exchange rates have greater upside
                                                                  forecast imperfections than downside imperfections (e.g.,
 8                              USD/ILS                 76.4%
                                                                  BRL, MXN and RON), which may be related to the forward
 9                             USD/COP                  74.5%
                                                                  rate bias and investor hedging demand.
 9                             USD/MXN                  74.5%

 11                            USD/PEN                  72.7%     References
 12                            EUR/CZK                  70.9%
                                                                  Bates, D.S. (1991). “The Crash of ’87: Was it Expected? The
 13                            USD/RUB                  65.5%     Evidence from Options Markets”, Journal of Finance 46(3):
 14                            USD/BRL                  43.6%     1009-1044.
Source: Morgan Stanley, FXPA
                                                                  Bates, D.S. (1996). “Dollar Jump Fears, 1984-1992:
Asymmetric Accuracy                                               Distributional Abnormalities Implicit in Currency Futures
                                                                  Options”, Journal of international Money and Finance 15(1):
Despite the encouraging results, we note that the markets         65-93.
are less accurate when predicting upside USD/EM or
EUR/EM risks. For example, the results for USD/BRL,               Christoffersen, Peter and Mazzotta, Stefano, “The
USD/MXN and EUR/RON all tend to display this lopsided             Informational Content of Over-the-Counter Currency
inaccuracy with the markets consistently overpricing the risk     Options”, European Central Bank Working Paper Series No.
of upside moves. We believe that this may be partly               366, June 2004.
attributed to the forward rate bias – the tendency for the        Garman, M.B. and Kohlhagen, S.W. (1983). “Foreign
forward rates of high-interest currencies to over-estimate the    Currency Option Values”, Journal of international Money and
risks of future depreciation. Another partial reason may be       Finance, 2(3): 231-237.
the distortions created by investor hedging demand in illiquid
out-of-money options.                                             Malz, A. (1996). “Using Option Prices to Estimate
                                                                  Realignment Probabilities in the European Monetary System:
The asymmetric accuracy is most salient in USD/BRL. The           The Case of Sterling-Mark”, Journal of international Money
BRL has historically been one of the highest interest-bearing     and Finance, 15(5): 717-748.
currencies in the EM bloc. As a result, we believe the market
forecasts in USD/BRL are likely more impacted by the              Malz, A. (1997). “Option-Implied Probability Distributions and
forward rate bias, resulting in the low score in the goodness-    Currency Excess Returns”, Reserve Bank of New York, Staff
of-fit test (44% and lowest among EM currencies).                 Papers: Number 32, November 1997.

                                                                  Kahale, N. (2005). “Smile Interpolation and Calibration of the
                                                                  Local Volatility Model”, Working Paper, ESCP Europe, March
                                                                  2005.




                                                                                                                           157
                                                                                                                                         MORGAN STANLEY RESEARCH

                                                                                                                                         March, 2012
                                                                                                                                         EM Local Markets Guide




                                                                                                                                          In Step 5, we construct four piece-wise non-increasing,
Appendix 1 – Six-Step Procedure                                                                                                           twice-differentiable and convex splines between the five
of the FXPA                                                                                                                               input points ( K 1 , v1 )...( K 5 , v 5 ) based on a modified form of
                                                                                                                                          the Black-Scholes formula:
In Step 1 of our six-step procedure, the volatilities (  ) for
the 25 and 75 option deltas (  ) are calculated from the 25                                                                                                             fi  i2      fi  i2 
                                                                                                                                                                      ln  
                                                                                                                                                                        K 2            ln K 2 
delta butterfly ( bfly t ) and risk-reversal ( rrt ) through the
                                                                                                                                          ci (K; fi ,i ,ai ,bi )  fi          K          aKbi
standard equations:
                                                                                                                                                                           i          i         
                                                                                                                                                                                                        i


 t25  atmt  bflyt25  0.5rrt 25                                                                                                                                                             
                                                                                                                                                                                                  
 t75  atmt  bflyt75  0.5rrt75                                                                                                      Each spline i (i  1...4) has four curve-fitting parameters
                                                                                                                                          ( f i ,  i , ai , bi ) and connects input points ( K i , vi ) and
In Step 2, we determine the fitted volatilities for the 15 and
                                                                                                                                           ( K i 1 , vi 1 ) .
85 delta (   ) using the quadratic function 45:
            ˆ
                                                                                                                                          The option price curve is extrapolated beyond the input
  ( )  atmt  2rrt (  0.5)  16bflyt (  0.5) 2
 ˆ                                                                                                                                        points. For strikes below the 15 delta, we add two input
                                                                                                                                          points at the extreme strikes of zero and infinity, i.e.,
In Step 3, the five delta points are mapped from a volatility-                                                                            (0, e  r  St ) and (, 0) , with corresponding splines 0 and
                                                                                                                                                   *


delta space into a volatility-strike price space by implicitly                                                                            5 of the above functional form.
solving for K, the strike price, in the equation below:
                                                                                                                                          Using the iterative algorithm in Kahale (2005), we can
                                                                                                                                          numerically solve for all 24 spline parameters simultaneously
            St                                                2                                                                      so that when they are joined at the input points, the entire
           ln K                                   (r  r *    )
                                                                 2                                                                       option price curve is also non-increasing, twice-differentiable
  e  
      r                                                                                                                               and convex. As long as the input points are arbitrage-free,
                                                                                                                                      the slope of the interpolated option price curve is within the
                                                                                                                                        range [-1, 0] and hence also arbitrage-free 46. In addition,
                                                                     
                                                                                                                                          Kahale (2005) proves the existence (and postulates
                                                                                                       *
where ( S t ) is the current spot rate, ( r and r ) are the                                                                               uniqueness) of the arbitrage-free option price curve.
domestic and foreign risk-free (continuously compounded)
discount rates, (  ) is the investment horizon and (  ) is the                                                                          In Step 6, we make use of the fact that the cumulative
standard normal cumulative distribution function.                                                                                         distribution function can be obtained from the first derivative
                                                                                                                                          of the option call price. The first derivative is
In Step 4, the option call price, v ( S t , , K ,  , r , r ) , is
                                                                                                                   *

                                                                                                                                               v( S t , , K ,  , r , r * )
                                                                                                                                                                               e  r 1   ( K )
calculated by using the Garman-Kohlhagen (1983) extension
of the Black-Scholes function:
                                                                                                                                                           K
                                    St            2                                                                                  where  ( ST ) dST  P {S T  k } is the risk-neutral
                                                                                                                                                                                     *
                                    ln   (r  r*  ) 
                                   K                                                                                                     probability density function. The second derivative of the
v(St , , K, , r, r*)  er  St  
                            *                        2 
                                                                                                                                          option call price is equal to the risk-neutral probability density
                                                                                                                                      function with respect to the strike price, K.
                                                        
                                                        
                                                                                                                                            2 v ( S t , , K ,  , r , r * )
             St                2                                                                                                                                          e  r  ( K )
             ln   (r  r*  )                                                                                                                       K 2
            K
 er K  
                                    2 
                                     
                                       
                                       
                                                                                                                                               In theory, the option price at zero strike is e  r 
                                                                                                                                          46                                                      *
                                                                                                                                                                                                       St   . However, in practice (e.g., due to
45 We prefer to use the fitted rather than the actual volatilities for the 15 and 85 deltas because we are concerned that these actual
                                                                                                                                          data synchronization issues), adding this point sometimes results in arbitrageable inputs. In
volatilities are distorted by liquidity issues.
                                                                                                                                          those cases, the price is adjusted downwards to ensure arbitrage-free inputs.


                                                                                                                                                                                                                                          158
                                                                                                      MORGAN STANLEY RESEARCH

                                                                                                      March, 2012
                                                                                                      EM Local Markets Guide




Appendix 2 – FXPA Back-Test                                                                            Step 3: Testing Whether the Market Implied and Realized
                                                                                                       Probabilities Are Statistically Similar
Methodology
                                                                                                       We test whether the (pre-specified) implied and realized
We adopt our testing methodology from the interval analysis
                                                                                                       probabilities are statistically different by running the following
technique described by Christoffersen and Mazzotta (2004).
                                                                                                       regression 49:
There are three steps to the procedure:

Step 1: Defining the Market’s Prediction of the Range
                                                                                                                          ( pu  pl )  I T ,h, pu , pl  aT ,h, pu , pl   T ,h
that the Spot Is Likely to Lie within in the Future                                                                                                                                 T ,h
                                                                                                       where        aT ,h, pu , pl is the mean prediction bias and                          is the
We use our FXPA engine to calculate the particular strike                                              error term. If the options market is an accurate forecaster of
K T ,h , p at time T that corresponds to a specific percentile p                                       the future spot rate, then
(where p = 5%, 15%, …, 95%) 47 at each tenor h (h = 1M, 2M,                                                                E ( DT ,h, pu , pl )  pu  pl .
3M, 6M, 12M). We then construct the interval I T ,h , p , p , with
                                                                                   u   l
                                                                                                       This is equivalent to the null hypothesis that                         aT ,h, pu , pl  0
upper and lower bounds K T ,h , p and K T ,h , p , respectively,
                                                             u       l                                 (i.e., there is no systematic bias to the market’s forecasts).
(where        pu  pl ). This interval I T ,h, pu , pl           then defines the
range that the market implicitly predicts that the future spot
 ST h will lie within at time T+h with probability pu  pl 48.

Step 2: Calculating the Realized Probability of Spot
Lying in the Specified Interval

Next, we calculate the realized probability of spot staying
within IT , h , p , p at time T+h for each date T, tenor h and
                    u     l

interval I T ,h , p              . We then define the indicator
                        u , pl

variable DT ,h , p                  as:
                           u , pl




                       1 if ST h  I T ,h, pu , pl  ( K T ,h, p l , KT ,h, p u ]
      DT ,h, pu , pl  
                       0                  otherwise

E ( DT ,h, pu , pl ) is then the realized probability of the spot rate
staying within IT , h , p                , pl
                                                at time T+h, which we estimate with
                                     u

the sample mean of                        DT ,h, pu , pl .




47
     For illustration, on Sep 1, 2008, the FXPA predicts a 25% probability that the spot will be
below    K 9 /1/ 2008 ,3 M , 25% in 3 months’ time (Dec 1, 2008).
48
   For illustration, on Sep 1, 2008, the FXPA predicts that in 3 months’ time the spot will be
within the interval
                                                                                                       49
                                                                                                            Since the sample observations are overlapping, we correct for serial correlations in the
I 9 / 1 / 2008 ,3 M ,35%, 25% (  ( K 9 / 1 / 2008,3 M , 25% , K 9 / 1 / 2008,3 M ,35% ] ) with 10%
(= 35% - 25%) probability.
                                                                                                       error term    T ,h using the Newey-West kernel with size corresponding to h.

                                                                                                                                                                                                159
                                                                                                                                                                          MORGAN STANLEY RESEARCH

                                                                                                                                                                          March, 2012
                                                                                                                                                                          EM Local Markets Guide




Exhibit 311
USD/MXN: FX Probability Analyzer
                                                                                                                                                                                                              Breakthrough Probability***
       Currency                USDMXN                                                                                                                                     Level        1m               2m                 3m                 6m             12m
       As of Date              9/1/2011                9/6/2011              09062011              Cross PASS                                                             18.00       0.0%             0.0%               0.1%               1.1%            4.2%
         Spot                  12.2950                                                                                                                                    17.50       0.0%             0.0%               0.2%               1.6%            5.0%
                                                                                                                                                                          17.00       0.0%             0.0%               0.3%               2.2%            6.0%
         Quotes                   01M                    02M                    03M                     06M                    01Y                                        16.50       0.0%             0.1%               0.6%               3.0%            7.1%
           ATM                   14.60                  14.85                  15.00                   15.30                  15.70                                       16.00       0.0%             0.3%               1.1%               4.1%            8.5%
          25RR*                   4.75                   5.00                   5.50                    6.10                   6.50                                       15.50       0.0%             0.6%               1.8%               5.5%           10.0%
          25Bfly                  0.60                   0.70                   0.75                    0.85                   1.05                                       15.00       0.1%             1.3%               3.1%               7.4%           11.7%
     r (domestic)**              3.48%                  3.44%                  3.36%                   3.33%                  3.36%                                       14.50       0.5%             2.7%               5.1%               9.8%           13.6%
         r (FX)**                0.25%                  0.29%                  0.32%                   0.41%                  0.46%                                       14.00       1.6%             5.3%               8.1%              12.7%           16.2%
         Forward                 12.33                  12.36                  12.39                   12.48                  12.66                                       13.50       4.7%             9.6%              12.5%              16.3%           20.7%
         25 Delta                12.76                  12.99                  13.19                   13.68                  14.50                                       13.00      11.7%            16.3%              18.6%              23.0%           27.9%
         75 Delta                12.02                  11.92                  11.86                   11.72                  11.56                                       12.50      27.5%            31.4%              32.8%              35.1%           61.7%
                                                                                                                                                                          12.00      27.3%            35.9%              40.3%              45.4%           47.1%
                                                                                                                                                                          11.50       2.5%             8.9%              13.4%              21.8%           29.4%