International Fixed Income
Topic VB:
Emerging Markets-Description
Review from last time:
Numerical Example
• Consider a 5.5% 2-yr semi-annual coupon bond.
• Now suppose that this bond has the following
characteristics:
– guaranteed principal
– nonguaranteed interest, with default probability each
6-mth period of P=.15
– First, price the guaranteed part, and then the
nonguaranteed component.
Recall the Data from Class
r.5 .9730, r1 .9476, r1.5 .9222, r2 .8972
First, the guaranteed part:
PG 100 d 2 89.72
Second, the Brady Bond:
The way to value this bond is to realize today’s
value is the discounted value of all future expected
cash flows. These cash flows only occur if there is
no default, i.e., if (1-p) occurs.
Brady Bond Mathematics
T
PNG E[C ] d
t .5
t t
T
Ct (1 p ) 2 t d t
t .5
T
2.75(1 .15)
t .5
2t
dt
( 2.27 1.88 1.56 1.28)
6.99
Thus, the price of the Brady Bond is
89.72+6.99=96.71
What About the Strip Spread?
T
PNG
t .5
1
Ct
rt s 2 t
2
T
6.99
t .5
1
2.7 5
rt s 2 t
2
Given the interest rates of 5.54%, 5.45%, 5.47% and 5.5%;
solve for the strip spread s. Note that this is one equation
and one unknown, but needs to be done on a computer.
What is S? S=36.42%!
Outline
• Emerging Markets
• Stylized Facts
• Case Study
I. 1997 Share of World Economy
GDP ($ trillions)
6.44 Developed
21.56 Emerging
Pops. (billions) Land Area (mm. km)
0.855 31.9
Developed Developed
Emerging Emerging
4.85 101
Emerging Markets
• Only a subset of these emerging market
countries offer investable debt securities.
• In fixed income, emerging debt markets
refer to this subset:
(see next page)
Types of Debt (1997)
Bradys
Other $dn.
Local
Loans
$-corp
Lcl-corp.
Who Holds This Debt? (1996)
Banks
Mutual Fds.
He. Fds.
Dealers
Pension
Insurance
Debt Volume (1997)
Bradys
Other $dn.
Local
Loans
options
1997 Brady Bond Universe
$ vs. non-$
7%
$
Non-$
93%
Types Guarantees
12%
36% 38
Fixed
Collateralized
Floating
Uncollateralized
Non-perf.
62%
52%
Types of Risks for $-denominated Bonds
• Interest rate risk: pays $ $ risk
• Credit risk (recall the strip spread)
– Economic fundamentals (e.g., GNP)
– Solvency (e.g., meeting debt obligations)
– Serviceability (e.g., foreign exchange reserves)
– Political considerations
– Willingness to pay
Bond Price Sensitivity
Recall that $-investments in international fixed income
had three components:
[yield] - [dur x (Dr)] - %DS(Fn/$)
There many not be foreign exchange risk now, but
sovereign risk implies
[yield] - [dur x (Dr$)] - [durs x (DSpd)]
Durs represents the duration with respect to a credit
spread change.
II. Stylized Facts about Emerging
Debt Market Returns
• Look at JP Morgan Emerging Market
Bond Indices (see handout)
• Monthly data, 1/1993 - 3/2000
• Estimate
– means
– volatilities
– durations & credit risk duration
Emerging Market Debt: Means
Annualized
0.16
0.155
0.15
mean
0.145
0.14
0.135
Fix Fltr Latin Total
Emerging Market Debt: Vol
Annualized
0.184
0.182
0.18
0.178
0.176
0.174 volatility
0.172
0.17
0.168
0.166
Fix Fltr Latin Total
Emerging Market Debt: Vol
Annualized
0.2
0.18
0.16
0.14
0.12
0.1 volatility
0.08 int.rate.vol
0.06
0.04
0.02
0
Fix Fltr Latin Total
Selective Correlations
0.9
0.8
0.7
0.6
0.5
0.4 correlation
0.3
0.2
0.1
0
Fix-Flt. Brz.-Mex Brz.-Nigeria
Conclusions
• Emerging markets trade much more like
equity returns in terms of returns/risk.
Why?
• Most of the volatility is due to credit risk,
i.e., $ interest rate risk plays only a small
role. Why?
• Given this, correlations seem to be
particularly high. Why?
III. Case Study
• Did investors foresee the collapse of the
Mexican peso in 1994?
• Look at short-term debt instruments over
1993-94 time period:
– cetes (23% of mkt): peso-denominated
– tesobonos (55% of mkt): $-denominated albeit
w/ capital controls
Mexican Bond Premiums
Let T be the Tesobono rate, C be the Cetes rate and r be
the US rate. It’s possible to show that, in $ terms,
Default Pm T rUS
Curncy Pm C T
Premiums (91 days)
0.2
0.18
0.16
0.14
0.12
0.1 Default
0.08 Currency
0.06
0.04
0.02
0
1993 1994 Nov-94 Dec-94 Jan-95
What about the 182-day Cetes and
Tesebonos?
• Assuming the expectations hypothesis,
could we have looked prior to September-
November of 1994, and inferred
devaluation risk from the longer maturity
bonds?
• The answer is no - there was no sign of a
currency premium on longer versus
shorter bonds.
Cetes & Currency
90 9
80 8
70 7
60 6 Cetes-91
50 5
Cetes-182
40 4
30 3 MP/$
20 2
10 1
0 0
N Jan- M M Jul- Sep- N Jan-
ov- 95 ar- ay- 95 95 ov- 96
94 95 95 95