Arbitrage Trading Currency Risk by mikeholy

VIEWS: 231 PAGES: 21

									                                 Example
Arbitrage Trading

On international markets there is a constant need to borrow securities for various reasons:
       to cover short positions in trading activities,
       to cover for settlement failures,
       to provide otherwise illiquid stocks.
As in all markets, it is important to match offer and demand. This is where Fortis Bank Information
Banking' GSLA (Global Securities Lending and Arbitrage) division can help. GSLA is a net
borrower, which means that they always have the need to borrow securities in order to support
their trading activities. GSLA always acts as a principal, which means that the counterparty risk is
with the Fortis Group.
GSLA conducts it's arbitrage on a global basis. Arbitrage trading strategies enable us to fully
utilise the securities borrowing and lending and structured product competences. Our own
proprietary business enables us to offer better returns on our securities lending program. GSLA
also acts as an end user of it's own securities borrowing and lending business. The following
arbitrage trading strategies can be defined:
       Reversals and Conversions
       Option strategies like Volatility Arbitrage
       Time Spreads
       Index Arbitrage using futures and / or options
       Warrant Arbitrage
       Convertible Arbitrage
       Statistical Arbitrage

Structured Products

Through our wide experience in the securities borrowing and lending industry, GSLA set up a
group specialising in legal, and regulatory issues. Structured Products enhances our securities
borrowing and lending capabilities and adds to the succes of it. The two competences mutually
influence and benefit from each other.
Examples of Structured Products are:
      Index Loans
      OTC Derivatives
      Equity Swaps
      Sell and Buy Backs
      Equity Repos and other balance sheet related transactions




Evolution GSLA

In the late 1970's, securities lending was a rather esoteric business even to some of the people
running the business for their firms. The concept was a secret of which the full potential was
understood by few. These handful of people had the vision to see what the future could hold for
the borrowing and lending of securities. They understood that as the business grew it would
spread into trading, treasury and custody areas of the industry and become one of the vital
organs of the financial world.
In this day and age almost every sizeable institutional investor directly or indirectly engages in
securities lending. Custody banks have come to appreciate that this activity is not just value-
added to these clients, but is essential to competitively pricing services overall.
The enhancing effect on portfolio returns, the potential lowering of custody costs and a better
understanding of the risks involved in the securities borrowing and lending market have led more
clients into the worldwide pool of lendable assets. This increase in the supply side, or loanable
securities, has dramatically impacted pricing. Alternatively, it has made the demand side, or
borrowers, more important.
GSLA was established in 1995. The business was established from the beginning as a front office
business. This enabled the business to grow rapidly and create a high level of quality service and
profitability. Within our business unit, there are several specialised teams from around the globe,
working together as one unit.
GSLA is a net borrower of securities. This is because it combines securities borrowing and
lending with it's own proprietary trading activities. GSLA borrows and lends securities from and to
the Netherlands, Germany, Belgium, Denmark, Sweden, Norway, Finland, France, Switzerland,
Italy, Spain, USA, Hong Kong, Singapore, Malaysia, Thailand, Japan, Indonesia, Philippines and
Australia. Since 1992, GSLA's daily outstanding loan portfolio has grown 20 times and is
expanding rapidly. During the high season in 2000, GSLA had a book of between EUR 30- and
EUR 40 billion.
The terms which apply to securities lending transactions are set out in a master agreement based
on the agreement drawn up by the International Securities Lending Association (ISLA). After
securities lending began to develop further, it rapidly became clear that suitable legislation and
regulation were either lacking entirely, or in need of adjustment. ISLA was set up in England by
market participants in order to standardise the international legislation, regulation and contracts.
Fortis Bank was the first foreign (non-UK) organisation to join ISLA.
The details of each individual loan transaction are set down in a confirmation and faxed to the
client / counterpart. This confirmation, together with the master agreement, is the specific
agreement between the parties concerned.
Fortis Bank (Nederland) N.V. acts as principal in all securities lending transactions (Our Hong
Kong office operates on behalf of Fortis Bank (Nederland) N.V.). This means that Fortis Bank
assumes all counterparty risks arising from borrowing and lending.
GSLA only deals with top-rated institutions, and has more than 250 securities lending agreements
in place. Each client / counterpart is treated individually in order to reflect not only the total
portfolio available for lending, but also the overall relationship between Fortis Bank and the client
/ counterpart. Before any loan transactions can be executed with any counterpart, a number of
conditions must be satisfied.
Client / counterparts, and the credit limits assigned to them, must be approved in advance by our
Credit Committee. Approval of the counterpart and the credit limit are based on their net equity
and risk weighting. In addition to this, general guidelines have been established for how GSLA
conducts its business. These guidelines are based on daily turnover of the underlying securities
involved and the total quantity of the securities which are freely available. When we lend to third
parties we accept collateral in the form of cash, fixed income securities, other securities and/or
letters of credits. At present however, letters of credit are accepted only on a limited basis.



Settlement

In light of changing market conditions, procedures for stock borrowing and lending are discussed
with the client / counterpart in close cooperation with its appointed custody account manager.
Unless otherwise agreed, all securities that are borrowed by GSLA from the client / counterpart
can be recalled at any time subject to the standard settlement period in the country concerned.
This guarantees the proper settlement of any sale of the securities by the client.


Technology

Advanced technology is used for the administration of securities borrowed and loaned, and for
the calculation and settlement of income derived.
GSLA uses state-of-the-art technology including the number one system in the securities lending
world, K-Tek's Global One.
This system is fully computerised and handles the most advanced information flow in the field of
securities lending. It is useful both internally and externally.
It looks after the maintenance of positions, daily "mark to market" calculations, monitoring of risk
management, internal securities lending parameters, limits, collateral, etc ...


Administration

Our securities lending administration is independent of our custody administration. The positions
of Fortis Bank's custody clients, who have agreed to participate, are made available to GSLA via
their inclusion in a lending pool.
This pool is viewed daily by us to see how it can be optimised and utilised. Loaned securities are
transferred into a separate account in the name of the lender, marked "loan account", in order to
reflect their status. Collateral is posted when required.
Custody clients are often prepared to give Fortis Bank an uncollateralized line for securities
lending activities based on the financial strength / soundness of Fortis Bank with its 16 Billion
Euro (Tier I and II) regulatory capital. Whenever collateral is not required, the possibilities of doing
securities lending transactions improve.

Transactions

Just as any other loan, a loan of securities will be backed by some form of collateral. The amount
of collateral pledged is equivalent to the market value of the securities borrowed by GSLA, plus
an excess margin required by the lender (typically 5%). Collateral and loaned securities are
marked to market daily. GSLA acts as principal in all securities lending transactions.
In a cash collateralised stock borrow transaction, securities are borrowed against cash collateral.
In a non-cash collateralised stock borrow transaction, securities are borrowed against non cash
collateral.

Financing and Leverage

Due to the financial strength of Fortis (AA- rated by Standard & Poors), solid financing is offered
even in the most difficult of markets. We assess possibilities for leveraging various strategies.
For example, Equities or Bonds Long/Short, Merger Arbitrage, Warrant/Convertible Bond
Arbitrage, Bridging/Redemption Financing, Fund-of-Fund Financing.
Recently developed risk models enable us to consider all strategies allowing us the flexibility to
remain focused during all market conditions.

Central bank watching out for arbitrage

    Taiwan's Central Bank of China will check large foreign exchange settlement cases to see if
    local firms have remitted the funds abroad and then sent them back to Taiwan for forex
    arbitrage amid the devaluation of the New Taiwan dollar.

    The CBC has asked designated forex banks to immediately report to CBC's Foreign
    Exchange Department the names of local enterprises whenever any single forex settlement
    case exceeds US$1 million, and check the usage of the fund, in a bid to prevent local firms
    from remitting money to their overseas subsidiaries and then sending it back to Taiwan to
    arbitrage from the devaluation of the local currency.

    A top executive with a large forex bank said that one of his bank's major corporate customers
    recently remitted a large amount of money abroad, and remitted it back to Taiwan a few days
    later, attracting grave concern from the CBC. The central bank asked the bank to check the
    usage of the fund written by the corporate customer when conducting the forex settlement, to
    judge whether the customer was engaged in forex arbitrage amid the fluctuations of the local
    currency.

    The executive said that if a local firm settles US$1 million in forex exchange at the NT$-US$
    parity rate of 30.5:1, the company will spend NT$30.5 million for the settlement. If the firm
    sends the money back to Taiwan at a time when the N.T. dollar depreciates to 31.5:1, then
    the company can easily enjoy a forex gain of NT$1 million. This is why the CBC is worried
    about such practice by local enterprises, now that the local currency has declined steadily to
    a seven-month low of 31.014 as of Wednesday, the executive added.

    Meanwhile, executives with most forex banks shared the view that the local currency is
    expected to rebound later, and the CBC is expected to intervene to prop up the local unit.
    Accordingly, they urged local firms holding U.S. dollars to convert their greenback on a
    gradual basis to secure maximum profit amid the fluctuation of the local currency.



Funding arbitrage

Funding arbitrage is another method through which financial institutions can control risk.
Suppose bank A borrows at 5% to buy a risky asset which yields 8%. This could be for a variety
of reasons, this bank might be involved in a underwriting syndicate where it, while receiving fees
for helping underwrite a debt issue, might wish to reduce the risk it now holds in the form of these
newly issued bonds. To reduce risk this bank could enter into the buy side of a total-rate-of-
returns swap as follows: the 8% rate of return on this risky asset is transferred to another bank,
say Bank B, which has a weaker credit rating, in return for payments of 6%. Due to its lower credit
rating Bank B’s borrowing costs are higher then Bank A’s, say 6.5% and were it to borrow to buy
this risky asset on its own we can see its net rate of return would be 1.5% (8%-6.5%). Through
this transaction Bank B’s net rate of return is now 2% (8% rate of return it is receiving from
BankA’s risky asset minus 6% payment it is making to Bank A). Bank A meanwhile has reduced
its exposure to risk as it is receiving payments from Bank B and not the risky asset and
nevertheless still has a net rate of return of 1% (6% it is receiving from Bank B minus 5%
borrowing costs).

Yield Curves
Construction of yield curves is sometimes neccessary to determine rates for future dates for
which no quoted rate exists. This can be accomplished in the following manner: Suppose the 91
day rate is 4.5% while the 121 day rate is 5% and we wish to know the rate for 111 days. First
calculate the difference between the two rates, 5%-4.5%=0.5%, now the period over which this
increase occurs, 121-91=30, which, assuming the increases take place at a constant rate, yields
a per day rate increase of 0.5/30=0.01667%. Then to calculate the rate at 111 days, [(0.01667% *
20) + 4.5%]=4.8334% at 111 days. (where 20 equals the # of days into the 30 day period we wish
to know the rate for). This type of extrapolation can be undertaken over a wide-variety of periods
to construct a yield curve.

Future Rates
It is often neccessary to establish a forward rate when conducting a transaction or pricing a
contract, the following methodology is used to reach a solution:
Suppose currently 6 month bid/offer rates are 9% and 10% respectively while the 12 month
bid/offer rates are 11% and 12%. To calculate the rate for a future period of, say, 6 months which
will start 6 months from now one would proceed as follows - you can borrow $5million at 12% for
365 days with the ensuing interest costs being $600,000. You can further lend $5million at 9% for
6 months with the interest earned over this period being $225,000 and therefore in 182 days you
will have $5,225,000 available to lend for another 183 days. This will have to earn ($600,000-
$225,000) =$375,000 for you to break even and cover the interest costs of having borrowed
$5million over a 365 day period. So the current offer rate for a 6 month period starting 6 months
from now must be, $375000/$5225000 = (0.072*2)*100 =14.4%
This must be the interest rate, given the current bid/offer rates for 6mnth and 12mnth periods,
because a higher rate would create an arbitrage opportunity whereby one could borrow over a
12mnth period at 12%, simultaneously lend out for a 6 month period 6 months from now at a
higher rate (say 11% instead of our calculated 9.18%), and since this transaction will take place 6
months from now lend out the amount currently held at the current 6 month bid rate of 9%. In the
end this would generate a profit and thus enacted several times over by market participants
would bring about a lower current bid rate (or a lower forward rate) to bring the market back into
parity over the segmented time periods.

Net Present Value
Net Present Value criterion is an important assessment which calculates the current value of a
future cash flow. NPV is a very useful tool for corporations and governments alike in that it allows
for a comparison of current costs to undertake a project versus the potentials benefits, in this
case revenues, that the project will yield sometime in the future. The formula for calculating NPV
is:
           n            i
 NPV=-I+∑ CFi/(1+r)
where: I= initial investment, CFi=cashflow in year i, r=discount rate, n=time horizon of the project.
If the NPV is greater then 0 then the current value of the cash flow that will be generated by the
project or business is positive. A simple example follows: Suppose the cost of project XYZ is
$200,000 which is expected to lose money for the first 2 years before beginning to turn a profit.
Lets denote these losses as $75,000 in year 1 and $10,000 in year 2 before a profit of $60,000 in
year 3, and $200,000 in year 4. The discount rate is the rate of return which could be had on the
$200,000 if it is not invested in project XYZ, lets suppose this is 5%. Then:
                                          1                          2                         3
NPV=-200000 + (-75000/(1+0.05) ) + (-10000/(1+0.05) ) +                       (60000/(1+0.05) )     +
                   4
(200000/1+0.05) ) = -64128.20
Therefore the NPV of project XYZ is currently -64,128.20 or to state it another way, the current
value of the cash flows generated by this project in the future is -$64,128.20. Therefore on a
cost/value basis it should not be initiated as the negative Internal rate of return during the period
when the project is losing money is greater, cumulatively, then the corresponding positive IRR
when the project begins to generate positive cash flow.

Forecasting Models
The following is an example of a simple OLS regression, with some non-linearity added for
demonstrative purposes. Such regressions allow a user to determine what effect some set of
variables, the independant variables, have on a dependant variable. In this case the dependant
variable is the price of the 30yr. T-Bill and we are attempting to find what explanatory power the
value of commercial bank loans, the Prime Rate, and the Average Yields on Moody’s Aaa rated
bonds have in explaining the 30yr. T-Bill’s price movements. 30yr T-Bill regressed on commercial
bank loans (variable X1), Prime Rate (X2), Avg. yields of selected Moody’s Aaa rated corporate
bonds(X3), and the treasury secondary market bid-rate (6mnth bill) (X4). Using monthly data,
1980-2001. Yields the equation:
Yt=B1+B2(X1/100)+B3X2+B4X3+B5lnX4+et
where X1,X2,...,Xn respectively take on the values annotated above. This regressed yields the
coefficients:
tcm30t=0.66512-0.14363X1-0.023172X2+0.13024X3+0.94X4
with a variance of estimate (sigma squared) of 0.074323. Note however this is simply a random
equation and used only for example purposes, its results most likely are the result of high
correlations present with (and within) the independent variables upon the dependent (very strong
ones in the case of the Prime Rate), thus there is little in the way of actual use for forecasting to
obtain profit. However this is the shape OLS regressions take, while they are not ideal for
forecasting in complex scenarios involving financial markets they can be incorporated into ANN’s
(Automated Neural Networks) to yield more accurate results in these cases. These simply models
can also serve some purposes when analyzing demand, supply, effects of interest rates on
various factors (ex. housing starts), and other such scenarios. Even in more complex situations,
say forex movements, such models can give indications as to whether explanatory variables
being tested are significant and what their effects are on the dependent variable (ex. do capital
flows into US equity markets have a significant negative effect on the price of the euro in US$’s?)

Forecasting Models: Part 2
Models as the one presented in the previous example can take many forms some which will be
presented next.
                                         2
Take for ex.:Eurot= B1 + B2Oilt + B3Oilt + B4GDPt + et
herein Oil prices and the avg. GDP of eurozone economies is used to predict movements in the
                            2
Euro but an extra term,Oil has been added. This is used to capture the effect a unit increase in
the price of Oil on the Euro, so if it was hypothesized that Oil prices do not have a significant
effect on the Euro until they transcend, say, $25 a barrel the squared term would capture this
effect and its coefficient would be larger if periods where oil was higher then $25 a barrel were
taken in isolation and used as a separate data set to run the above regression
For example suppose we have a production function where Q t depends upon Et(energy),
Lt(labor), Kt(capital) where if one of these variables is equal to zero there should be no output (Q).
Herein you could log each variable to yield the equation, lnQ t = B1 + B2lnEt + B3lnLt + B4lnKt + et.
We could further take elasticities (this can be done regardless of whether the models is log-log or
not) to determine whether there are increasing, decreasing, or constant returns to scale. This
would be accomplished by taking elasticities at various points to determine whether they were
increasing, decreasing, or constant over time.

Automated Neural Networks
Automated neural networks can be constructed in a variety of ways, below is a simple flowchart
explaining the basic structure of these models:




The X’s in this case are the explanatory variables which will be used to predict Y, the dependent
variable. N1, N2, N3 represent nodes, each explanatory variable inputs into each node and each
node performs a specific instruction with the set of explanatory variables it receives. In the case
                                                                                                    -Z
of discrete choice models the structure of each node would generally take the form, N1=1/1+e t
                                                                2
where Z is equal to some functional form, say a1+a2X1+a2X2 . Nodes can be structured in a wide
variety of ways depending upon how the explanatory variables are to be interpreted. The nodes
can also feed into each other, so if N1 gives a result of less then 7 this could feed into node 2 and
instruct node 2 to alter its functional form or to not perform a calculation at all. Finally the nodes
feed into the output layer where the beta’s (B’s) denote the weight attached to each node and
whether the effect on the explanatory variable is “+” (so positive beta) or “-” (negative beta). The
arrow at the bottom of the picture indicates that the model flows upwards although many times
the output will feed back into the nodes which can restructure depending upon what output was
last period. To find the optimal number of nodes to be used and to determine what structure they
should take ANN’s are usually tested over historical data. The models which generate the most
accurate forecasts are saved, modified, and retested over historical data until the most efficient
model with respect to, usually, accuracy of forecasts is obtained. It can take many of these
“epochs” before a usable model is found, especially with respect to financial markets data.

Alpha, Sharpe Ratio, and Sortino Ratio
Alpha: Two methods for evaluating the risk of an asset are Alpha and the Sharpe Ratio. Alpha is
the risk-adjusted rate of return on an asset, for example suppose here we are considering a
bond. The equation for expected rate of return for this bond is: E(Rb)=R+Bb[E(Rm)-R]
here: R=the risk free rate of return, Rb=rate of return on the bond, Rm= the rate of return for the a
relevant bond index (of which either this bond is a part of or could be a part of), and
B(beta)b=Cov(Rb,Rm)/Var(Rm). Herein Beta is the coefficient which tells us to what extent this
particular bond moves with relation to some relevant bond index. Say B=1.5, then for every one
unit uptick (downtick) in the bond index this particular bond moves up (down) 1.5. Now if the
return we expect for this bond over the next period, say one year, exceeds E(R b) then this asset
is underpriced because E(Rb) is the expected rate of return in equilibrium. While we cannot know
what the expected rate of return is on our bond over the next year we can infer this from the
average return on the bond from previous periods. Suppose historically this bond has returned
15% while the average interest rate over this period has been 5% and thus the excess return of
the bond, E(Rm)-R, has been 10%. Now suppose Rm, the relevant bond index, has returned an
average of 8% over this same period so E(Rm)-R=3% and the beta (B) of our bond is 1.5. Then
alpha, the risk-adjusted return, is equal to: [E(Rb)-R]-Bb[E(Rm)-R]. Substituting in the above
values gives an alpha of 5.5%.

Sharpe Ratio & the Sortino Ratio:
The Sharpe Ratio is very similar to Alpha as it is simply the return of an asset over a comparable
risk free rate of return divided by the assets standard deviation (how volatile the asset is): [E(R n)-
R]/se(Rn). So it follows then that if you had a stock give you a rate of return of 50% but its Sharpe
Ratio was very low, say 0.5, you might reconsider whether you wish to reinvest your profits in this
same asset. Unlike the Sharpe ratio, the Sortino ratio does not penalize a fund for its upside
volatility. It is defined as, [E(Rn)-Re]/v(Rn), where the numerator is an asset’s return minus minus
an investors target rate of return (Re). Unlike the Sharpe ratio, the denominator only consists of the
assets downside volatility. So the above stock could have a Sharpe Ratio of 0.5 but a Sortino Ratio of
say 2 or 3 if the bulk of the stock’s volatility rested in its upward movements.

Delta, Gamma, Theta, and Vega
Delta: ratio between the change in the price of a derivatives contract (ex. an option) and the
change in the price of the underlying asset. As options near expiration, in the money options
approach a delta of 1 while in the money put options approach a delta of -1. Delta is also used as
a reference to market direction (hence the term delta trading).
Gamma: the amount of change in the delta for, usually, a 1% rise in the underlying asset of the
derivatives contract.
Theta: the ratio of the change in the price of a derivative contract with respect to the decrease in
the amount of time left until expiration of the contract (time decay). Theta is also used to
determine whether market is range bound or looking to breakout.
Vega: the amount of change of an option with respect to a 1% change in volatility. Vega is also
used as a reference to volatility of the market as a whole

Hedging Bonds using Repos
While credit derivatives can help hedge default risk of a fixed income position there are other,
arguably more beneficial, ways in which to hedge risk in fixed income markets. One such way is
through repos, a repo is a customer borrowing a bond from a dealer with an obligation by that
dealer to repurchase this bond at an agreed upon price at the end of a specific period. Suppose
an investor holds bond XYZ in their fixed income portfolio and wishes to hedge against the risk of
the price of bond XYZ dropping, or this investor has no position in bond XYZ and simply wishes to
make a speculative profit. Here a repo transaction would occur as follows: investor buys either
bond XYZ or a bond very similar to XYZ from a dealer for price P1 with the condition that the
dealer will repurchase this bond in 6 months at a price of P 1*(1+Rm*t)         where: Rm=market
interest rate and t=180/360 (since this transaction will be completed at the end of 6months). The
investor would now sell this bond, which is very similar to bond XYZ, on the market at price P 1
and, assuming their prediction was correct, repurchase it at a cheaper price at the end of the 6
month period to sell back to the dealer.
numerical example: Bond XYZ is selling at parity, that is $100 per $100 face value and the
current market interest rate is 5%. It is expected bond XYZ’s price will fall over the next 2
months (suppose they are May and June). Bond ABC, issued by the same company but with a
longer maturity then Bond XYZ, is purchased from a dealer at $98 for $100 face value with the
obligation that the dealer will repurchase after 2 months at price [98*(1+0.05*62/360)] = $98.84
per $100 face value. This borrowed bond ABC is now sold on the market at $98/$100. Over the
next 2 months bond XYZ falls in price to $95 while the price of the longer dated bond ABC, since
it is issued by the same company, also falls in price to $93. Bond ABC is then repurchased at
$93 and resold, as per the repo contract, back to the dealer for $98.84 netting a profit of $5.84.
Hence the effect of the decrease in the price of bond XYZ has been successfully hedged and in
this case even manages to yield a moderate profit of $0.84.

Convertible Bonds and Arbitrage
Convertible Bonds are bonds which offer the holder an option to convert the debt held into equity.
An example would be a convertible bond of company X which can be converted into 5 shares of
company X’s stock, usually only when the stock price rises above a specific threshold. The three
main determinants of the price of convertible bonds are, investment value, conversion value, and
option value. Investment value refers to the price of the convertible if it were a straight bond, that
is, if it did not have an option whereby it could be converted into equity of the underlying
company. The convertible bond will never trade below this value because while the value of the
conversion option may drop if the underlying companies stock is struggling the investment value
remains constant since the coupon on the bond is always being paid and the principal will be
returned upon maturity, barring a default. The conversion value is the value of the bond were it to
be converted into its equity equivalent, so if a convertible can be converted into 5 shares of stock
and the stock price is currently $10 then the conversion value is $50. When the conversion value
is higher then the investment value then this factor dominates in price determination and vice-
versa. However since conversion value can only be positive, since conversion is an option not an
obligation, a convertible bond will usually trade higher then both its investment and conversion
value. Finally, option value is simply the time value of the option to convert the bond into stock.
Convertible arbitrage is undertaken through establishing a short position in the stock that the
bond can be converted into. This practice, known as delta hedging, consists of dividing the price
of the convertible by the stock price conversion premium and then multiplying by the option delta.
Suppose a convert's price is $1000 and the company's current stock price is $50 with a 50%
conversion premium, so the value of the stock price conversion premium is $75. Option delta is
the movement in the price of the option for every 1% increase in the stock, suppose an option on
this stock has a delta of 0.65. The amount of shares to short, hedge ratio, is then:
($1000/$75)*0.65=8.6667. During small price movements in the stock this short position will act
as an effective hedge against a decrease in the price of the bond, since part of the price
component of a convertible is conversion value. This helps create a market neutral position with
returns solely dependent upon the coupon the bond is paying. During volatile markets this hedge
breaks down and a large decrease in the price of the stock will result in a profit because the short
position becomes more profitable then the losses from holding the bond. Conversely a large
increase in the price of the stock results in a profit because the rise in the price of the convertible
bond is greater then the loss from the short position. The central component of convertible
arbitrage is the current income derived from a position. That is, the income if the price of the
stock remains more or less constant. If the coupon on company X's bonds is 7%, their current
price $50, and in turn their current yield just under 14%, say 13.5%. Then the current income in
establishing a convertible arb. position would be 13.5% + the interest recieved on the proceeds
received from the short sale of company X's stock.
Credit Derivatives
There are three main types of credit derivatives, default swaps, total-rate-of-return swaps, and
credit-spread put options.
Default swaps simply transfer the credit risk of an asset from one party to another. The holder of
an asset, suppose a bond, would pay a periodic fee to the "risk buyer" and in return would recieve
some agreed upon payment in the case of some credit event, a default, bankruptcy, credit
downgrade, etc. The usefulness of these is straightforward as they help hedge against events
which would cause a dramatic decrease in the value of an asset.
Total-rate-of-return swaps have a "total-return purchaser" who pays a periodic fee to the "total-
return seller" in return for the total net payments of some underlying asset. In this case if the
underlying asset increases in value the "total-return purchaser" receives this higher return but if
the underlying asset depreciates in value then the "total-return purchaser" pays the value of this
decrease to the "total-return seller". In the case of a default or some other credit event of the
underlying asset the total-rate-of-return swap usually terminates. For an example of a partial, buy
side only, total-rate-of-return swap see funding arbitrage.
Credit-spread put options are somewhat similar to total-rate-of- return swaps in that they do not
rely upon any specific credit event occurring. A credit-spread put option consists of the buyer who
wishes to guard against an increase in the spread, where spread is the yield differential between
an asset held by the buyer and usually a interest-rate swap, and a seller who writes the put
option. The buyer pays a one-time fee to the seller and in return the seller agrees to pay some
agreed upon amount should the spread breach some preset level. The benefit here is that during
times of heightened uncertainty, especially concerning emerging markets, risky assets can
decrease in value with no apparent cause which renders default swaps useless since they only
provide protection in the case of some credit event. A credit-spread put option would be useful in
this situation because it would entitle the holder of the risky asset to a payment as soon the yield
spread passed some preset threshold. Therefore as this riskier asset now declines in value, due
not to a credit event but instead to some financial crisis or panic, the holder is compensated by
the seller of the credit-spread put option.
Credit derivatives can be used in a variety of ways, aside from simply hedging risk. An example
is their use in helping to construct synthetic positions which cannot be established outright in
standard markets. Suppose an investor wishes to buy a 4 yr. debt issue of a foreign company,
company Ba, but no such bonds have been issued by this corporation and he/she also does not
wish to undertake the currency risk which would be involved in such a transaction. Herein this
investor could purchase 4 yr. bonds of another corporation, preferably a more creditworthy one to
reduce risk, which are denominated in the domestic currency and simultaneously offer to "sell
insurance" against a potential default of company Ba. This investor would then receive the
coupon payments on the higher rated debt that has been purchased outright plus the periodical
payments from the credit swap. This creates a synthetic position where the return is now similar
to an outright purchase of a 4 year bond issue of company Ba denominated in domestic currency.

Volatility Swaps
As the name implies volatility (or variance) swaps allow investors to profit from or hedge the risks
of an increase or decrease in future volatility of an index of securities. A volatility swap on the
NIKKEI index would have a payoff at expiration of
     (σv-Dvol)*N
where:
σv= is the actual volatility of the NIKKEI index over the life of the contract.
Dvol= the volatility specified by the swap
N= the notional amount of the swap (in dollar, yen, etc.) per a unit of volatility.
This formula can also be modified to denote a variance swap where variance is simply the square
                         2
of volatility (sigma): (σ v-Dvar)*N
So the buyer of this swap receives N amount (dollar, yen, euros, etc.) for every unit increase in
volatility (variance) above the volatility (variance) denoted by the swap (D vol or Dvar). Dvol or Dvar
would usually be quoted as a percentage (ex. 20%) and N would usually be specified as amount
per 1% increase in volatility (ex. $1000/0.5% volatility change so $1000 for 0.5% increase in
volatility beyond 20%). Volatility swaps can serve many functions outside of simple speculative
positions on future increases (decreases) in volatility, examples being equity portfolios managers
purchasing volatility swaps as a hedge against severe downturns in markets (ex. NASDAQ
2000/2001) or hedge funds using volatility swaps to attain more market neutral positions.

Options Strategies
There are many types of options strategies which are used to accomplish a wide variety of tasks
from hedging exposure to taking speculative positions, in somewhat more complex and rewarding
ways. This section will deal with 4 types of options strategies, synthetic puts, long straddles,
calendar spreads, and ratio call spreads. For simplicity each example, with the exception of
Calendar spreads, involves the purhcase/sale of one put/call contract.
Synthetic Put - simple strategy used to protect short sale of a stock by purchasing a call option.
Ex. ZZZ is trading at $20 and trader expects stock to decline in value, shorts 100 shares of ZZZ
at $20, net proceeds = $2000. Simultaneously trader purchases, suppose August, $22.5 call
option for $2, cost = $200. If ZZZ declines in value below $18 then the short position is profitable
(with ZZZ @ $18, short pos.=+$200 - cost of call option (-$200), with ZZZ @ $17.50 net profit
=$50, etc.), if ZZZ increases in price there is a loss from the short position up to $22 but after this
point any loss from the short is covered by profit from the call with a chance of a profit depending
upon volatility of stock ZZZ. The further out of the money the call is the weaker the hedge and
vice versa.
Long Straddle - strategy used to profit from a large move up or down in a stock. A straddle
consists of purchasing very near the money call and put options of the same security with the
same strikes & maturity. Ex. ZZZ stock is at $30, trader purchases August 30 call for $3 and
August 30 put for $2.50 (generally calls entail a higher premium then puts). The total cost of this
purchase = $550, therefore this position is profitable when ZZZ rises above, roughly, $35.50 or
declines below, roughly, $24.50. Anywhere in between these two ranges, except ofcourse exactly
at $30, and the loss is less then the cost of $550 since the profit from the call(put) will offset the
loss from the put(call).
Calendar(time)spread - this strategy can be structured to profit from a range bound stock or
highly volatile movements. For the first choice, where its expected a stock will not move much
over a period of time, a trader would sell a near term call option while purchasing a long term call
option. Ex. ZZZ is trading at $30, trader expects it to remain very close to $30 or below over the
month. Sells August 30 call for $3 and purchases October 30 call for $5.5, net cost = $250. If at
expiration of the August 30 call ZZZ is trading slightly below $30 then this call expires worthless
and the trader can close out the spread by selling the October 30 call, at say $5, for a small loss.
Net profit from the spread is ($300-$250)=$50 or 0.5*amount of calls sold and purchased. If
future expectations for ZZZ were that the stock would rise over a period of time then this spread
could be restructured as follows: ZZZ @ $30, sell August $32.5 call for $2 and purchase October
$32.5 call for $4.5. If at expiration of the August call ZZZ is still below $32.5 then the cost of the
position declines to ($200-$450) $250. Further increases in ZZZ over 2.5 points will cause the
long October call position to yield profits.
Ratio Call Spread - a generally neutral strategy where no great movement is expected in the
underlying stock. There are different ways to construct this spread the following example deals
with a net credit position: ZZZ is at $33, purchase August $31 call for, say, $3.5 and sell two
August $35 calls for, say, $2.5 each. Therefore net profit in establishing this spread is ($250*2-
$350) $250 and the ratio is 2:1 in favor of the short position. If all calls expire worthless the
spread still yields a profit of $250, if ZZZ rises to 35 at expiration the spread achieves maximum
profit and finally if ZZZ rises significantly then the break-even point is equal to maximum profit,
say 9, plus the higher strike price, $35, = $44.

Exotics - Barrier Options, Compound Options, and Asian Options
There are many forms of exotic options, three of the more popular types, barrier options, the most
common forms being a Knock-in option or a Knock-out option, compound options, and Asian
options are explained here.
Barrier Options (Knock-In and Knock-out Options) - A Knock-In option has a provision whereby
when the underlying assets value increases to or drops below a defined barrier a new option, with
preset characteristics, is created. A Knock-out Option has a provision whereby when the value of
the underlying asset reaches some point the option ceases to exist. Example of a Knock-in
option: August 1.85 USD/CHF call, the right to exchange 1 US$ for 1.85 Swiss Francs. Add a
knock-in provision of 1.79. If current USD/CHF spot is 1.84 then this option would not have a very
large premium due to the unlikeliness of USD/CHF reaching 1.79 before expiration at end of
August and in this case the barrier option is cheaper then the plain vanilla August 1.85 CHF call
option. The more likely it is that the knock-in barrier is reached the more expensive will be the
options contract. The opposite is true for a Knock-out option, as once the knock-out barrier is
reached this type of options contract becomes worthless. Again suppose USD/CHF spot is 1.84.
An August 1.85 USD/CHF call with knock-out provision at 1.80 will have a premium which is
determined by how likely it is that the knock-out barrier is reached. The closer the knock-out
barrier is to the current market level the cheaper is the knock-out barrier option. The above two
examples deal with barriers that are out of the money now suppose an option with a barrier that is
in the money. Again USD/CHF spot is 1.84, suppose August 1.89 USD/CHF call which, in plain
vanilla form, will have a high price due to the significant amount of intrinsic value. However
consider the same option but now with a knock-out barrier level of 1.82, the value of this option
will be significantly less then of the plain vanilla option due to the probability of spot 1.82 being
reached. If you thought that over the duration of this option (so over August) USD/CHF would not
reach 1.82 then the opportunity exists to purchase the August 1.89 USD/CHF call with knock-out
barrier at 1.82, exercise it to receive 1.89 CHF per US$ and lock in a notional profit of 0.05 (as
the current market spot rate for USD/CHF is 1.84). The higher the implied volatility on the option
the cheaper it will be when it has a knock-out provision and the more expensive it will be if it has
a knock-in provision.
Asian Options- an Asian option has its payoff determined by the fluctuations in the underlying
asset through the term of the option. For example, an average rate option has its payoff
determined by taking the average value of the underlying asset, subtracting it from the strike
price, and finally multiplying this by the notional amount instead of simply taking the value at
expiration. Lookback options give the buyer the right to, in the case of a lookback call, buy the
underlying asset at the lowest price reached during the term of the option while a lookback put
gives the right to sell at the highest price reached through the duration of the option.
Compound Options - Compound options are simply options on options. So say a compound put
option giving the buyer the right to purchase a put option on stock XYZ at a specified price or a
compound call on a put giving the holder the right to purchase a put option on stock XYZ at a
specified price. Compound options, like volatility swaps, can be used to speculate on movements
in volatility among many other uses.

Risk Reversals - 25 Delta Risk Reversals and related measures of market sentiment.
Risk reversals refer to the manner in which out-of-the-money options, usually FX options, are
quoted by dealers. Instead of quoting prices these options are quoted in terms of their volatility.
The greater the demand for an options contract the greater will be its volatility and its price. An
options contract that is currently in the money, one which gives you the right to conduct some
transaction at a cheaper rate then current market prices, will usually have a premium assigned to
it both in terms of volatility and in terms of price while for an out of the money options contract it is
often the opposite. In this case the volatilities quoted by a dealer for, say out-of-the-money yen
calls versus out-of-the-money yen puts, will indicate which scenario the market is currently
placing a higher probability on.
25 Delta Risk Reversals: A 25 delta risk reversal consists of either buying a call with a delta of 25
and selling a put with a delta of 25 or vice-versa. Thus, this is a directional trade with the volatility
quote consisting of the volatility on the 25 delta call option and the 25 delta put option. Whichever
contract the volatility favours that is the direction that the market currently expects the underlying
currency to move. So if the 25 delta risk reversal favoured yen puts over yen calls then current
market sentiment is more heavily in favour of the yen making a large movement down rather then
up.
Interest-rate Swaps|
An interest-rate swap is essentially an agreement between parties to exchange interest
obligations on a principal amount over a specified period of time. IRS' can take many forms, the
following example demonstrates how two companies can engage in an IRS to lower borrowing
costs. Suppose Company X can borrow floating rate at LIBOR + 1% and 10yr fixed rate at 11%
while company A can borrow floating rate at LIBOR + 0.50% and 10yr fixed rate at 9%. Here
company X, which has the lower credit rating, wishes to borrow fixed rate but at a lower cost then
currently available while A wishes to borrow floating rate but, as with company X, at a lower cost
(say simply at LIBOR, instead of LIBOR +0.5%). Note that even though A can borrow floating
and fixed cheaper then X, A has a comparative advantage in borrowing fixed rate. An IRS would
be arranged as follows, company A borrows 10yr fixed rate at 9% while company X would borrow
floating rate at LIBOR + 1%. Company A would now take on X's LIBOR payment while company
X takes on A's 9% fixed rate payments (while continuing to pay the 1% from its original loan).
Hence Company X now pays 10% (9% from IRS + 1% from original loan) for a 10yr. fixed rate
loan, savings of 1%, while company A simply pays LIBOR, savings of 0.50%.

Don't bank on arbitrage advantage, NRIs told
Vinson Kurian
Viewed from the adverse implications resulting from the continued influx of hot money into the
country, the RBI prescriptions were only to be expected.
Thiruvananthapuram , Aug. 22
INTEREST rate caps and other restrictions on non-resident deposits present the most conclusive
evidence of an emerging scenario where arbitrage opportunities would increasingly become
scarce for NRI customers to bet their savings on.
This is what Mr K.P. Padmakumar, Chairman, Federal Bank, makes of the Reserve Bank
prescriptions that seem to have set a `policy-level' cat among the homing NRI pigeons used to
feeding merrily on the interest rate differentials thrown up by gyrations in the foreign exchange
market.
Speaking to Business Line here after addressing an NRI customers' meet, Mr Padmakumar said
this was an issue arising out of the forex surpluses that the country was accumulating in recent
times. Understandably, the Central Bank was worried about the arbitrage opportunities that were
becoming too plentiful for comfort.
"In a declining interest rate scenario all around, keeping an island of high interest rates structure
and aiming to attract all those monies with no apparent advantage for the country in terms of
deploying them at a reasonable rate of return would not have made economic sense. It was
therefore natural that the central bank sought to kill this arbitrage advantage," he said.
The NRIs have to get used to the fact that even as they seek to keep their money in India as any
patriotic Indian would do, they should not be seen as trying to make a kill - as demonstrated in
their perceived lookout for what might appear as patently unreasonable interest rate structures.
"This, I think, is something that they would need to get adjusted to as we go along, at least in the
short to medium term. You can't be too sure about the long term," Mr Padmakumar said.
For another, India's competitiveness has improved worldwide, despite an appreciating rupee.
These are facts that the RBI is only well aware of. Unlike in the past, notably the 1990s, the kind
of desperate need for building our reserves is not being felt anymore. "So to a large extent, I
would say that the arbitrage opportunities available till now have been bringing in hot money, not
any real savings, taking undue advantage of the interest rates differential. Viewed from the
adverse implications resulting from the continued influx of hot money into the country, I think the
RBI prescriptions were only to be expected."
In Mr Padmakumar's view, the Resurgent India Bonds that mature in October would present all
private banks in the State, as also those outside, with a denouement of unique dimensions. Many
leading banks, including multinationals, are keen on attracting as much RIB monies as is
available for their indulgence.
"Bankers in Kerala have traditionally priced NRI funds higher, if only to cultivate this privileged
constituency as an assured source of funds. And some of us also launched specific products, in
tune with their demands aired from time to time. At the end of the day, if the customer is not
getting the expected return compared to what is internationally possible, then it will start hurting
the NRI portfolios of banks. My expectation is that restrictions on NRI deposits would be only a
short-term measure and a reversal would take place sooner than later. Personally, I feel NRIs
need to be paid on a par with what is being paid to domestic depositors.
"Especially so, in view of the fact that their remittances would yield only fewer rupees tomorrow.
The rupee is perhaps destined to appreciate further. There are people who are predicting Rs 45
by March 2004. And Rs 42 or thereabouts, going forward. In that context, the NRI doesn't gain;
he actually loses on value in rupee terms if he is fixing the deal now. By offering him a rate which
is not competitive vis-à-vis international rates, I think bankers would only stand to lose the
confidence of a constituency known best for its habit of putting most, if not all, of its savings in
banks. Because, after the stock market fiasco and the real estate crash, by and large, banks
deposits have been the most preferred destination for parking their hard-earned money," Mr
Padmakumar said.
"Effectively, with the kind of a rate structure that the RBI is prescribing, we may lose the NRI
advantage. It is this fear that has driven us to hold this type of a meeting here. We are going to
hold such meetings at other centres also. This is one way of assuring NRIs that we are standing
with them in this hour of uncertainty, and also to allay their fears that banks might lose interest in
continuing to maintain the respective NRI portfolios. That kind of a feeling should not be allowed
to bother NRI minds. Yes, of late, NRIs have been openly airing such fears," he added.

The Latest 'ZITCOM' and My New Tax Shelter Bank
Author(s): C. Eugene Steuerle
Other Availability: PDF | Printer-Friendly Version
Published: May 05, 2003
Citation URL: http://www.urban.org/url.cfm?ID=1000482


"Economic Perspective" column reprinted with permission.
Copyright 2003 TAX ANALYSTS
The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of
public consideration. The views expressed are those of the authors and should not be attributed
to the Urban Institute, its trustees, or its funders.

It's too bad that the lifetime savings account (LSA) proposal of the administration seems to have
no legs. I had already seen what a great boon this was going to be for society (or at least my
personal corner of it) and had begun preparations for a special Steuerle cyberspace bank to help
out the lowly taxpayer. Now my bank would be unique because it would allow people to "save"
and gain the tax breaks without coming up with any money at all. Because of my deep social
concern for the individual with no money to save — not unlike my tax shelter brethren who really
only want to help out the lowly corporation with no money to invest — I would set up an account
with net "Zero deposIT," a ZIT. Someone opening a ZIT would merely have to pay me a little bit of
a fee, and then at the end of the year they would get back much more than their fee in the form of
a tax reduction due to additional net interest deductions on their tax return. My main concern was
to attract enough deposits before Merrill Lynch, Bank of America, and other financial
intermediaries caught up and started competing with my bank. ZIT.com would be to the new tax
shelters as Amazon.com was to book sales.
First, a bit of a clarification. What is involved here is tax arbitrage, a type of tax sheltering that
involves not simply taking advantage of low or zero tax rates on preferred assets, but leveraging
those tax breaks by multiplying up financial assets and liabilities. The most common form of tax
arbitrage involves borrowing to buy preferred assets, but tax deductions can also be generated
through a short sale of one asset and simultaneous purchase of another. It is through the
additional leveraging that one can generate extra tax deductions without doing any saving at all.
Certainly tax arbitrage is not new. Much of the growth in private debt in the economy is due to
these opportunities. Individuals borrow against their houses and business and then effectively
invest the proceeds in other assets that generate little in the way of taxable income. They do this
now when they borrow (or pay down their mortgages more slowly) and buy individual retirement
accounts (IRAs), make deposits in 401(k) plans, buy stock on which they defer capital gains from
tax, and put money into Coverdell education savings accounts, section 529 plans, or medical
saving accounts. It works in reverse as well. People make the deposits in these preferred assets,
then find themselves a bit short on money, and so borrow a bit more, take out a second
mortgage, or pay off their debt a bit slower. Individuals don't even have to know they are engaged
in this arbitrage as long as they simultaneously write off debt payments from a mortgage or
business loan and avoid tax on the receipts from some assets.
Ever wonder how Congress could enact so many saving incentives over the years and yet
generate little or no net personal saving? Now you know. Congress doesn't really provide saving
incentives. Instead it provides incentives for making gross purchases of certain types of assets
even if there is no saving at all.
The opportunities are so many and the limits so high in cases like section 529 plans that the initial
Treasury effort to establish LSAs seems to have been directed more at simplification than
expansion of tax sheltering opportunities. In particular, Treasury wanted to combine together all
the separate savings vehicles for selective purposes outside of retirement. Thus, it would remove
separate requirements in various accounts that the savings be spent on students under age 18 or
in college or on medical expenses. The effort got out of hand, however, as the White House
wanted higher limits for the new saving vehicle, no lifetime limit was established, and it refused to
restrict any existing tax preference as a means of folding them together or limiting revenue cost.
Why would my bank help expand arbitrage opportunities beyond the extraordinary extent to which
they exist already? Under existing law, it is often hard to match up liabilities with assets in a
riskless manner. Often the asset has a different maturity or risk structure than the liability. As an
example, it is possible to borrow to invest in an education account, but educational needs are
often way down the road and the money not easy to withdraw until then. One can also borrow to
invest in stock whose yield is expected in the form of unrealized capital gains, but the returns
received on the stock investment are unlikely to match exactly the timing and risk of the payments
made on the borrowing. As a final example, borrowing to put money into retirement accounts
creates some problems of liquidity: If the money is needed at some point, withdrawal from the
retirement account may lead to a penalty.
Not that attempts are wanting to create essentially riskless tax arbitrage opportunities. Many of
the new corporate shelters come close. A common type of device fought by the IRS for years was
a straddle in futures, where a trader might sell, say, wheat futures due in one year and buy wheat
futures due in one year and one month. Then, as wheat prices changed, the trader would
recognize the loss on one part of the straddle but not the gain on the other side. Here the law now
requires most traders to recognize all gains and losses at the end of the year — "mark them to
market" is the technical term — so that losses cannot be recognized in the current year and gains
deferred.
With lifetime savings accounts, new opportunities would abound for even the most risk-averse
individual to buy and sell short the same asset, recognize the loss (the interest payment), and
never pay tax on the gain (the interest receipt). If I borrow $10,000 and put $10,000 in the bank,
both at a 5 percent interest rate and both with a common maturity date (or simply some common
interest rate compounded daily), then my net assets are zero, my net interest is zero, and my risk
exposure is essentially zero.
The main difference between the old world and the new LSA world is that the tax preference — in
this case, the nontaxation of the returns from the asset — is now available for short-term assets,
not just for items like retirement and education. There is no penalty for withdrawing at almost any
time. Now, for the first time, short-term liabilities could be matched up with short-term assets,
yielding an exponential growth in tax arbitrage opportunities.
This ability to save taxes using assets and liabilities that almost match in short-term risk and
maturity is what would make my bank so successful. With ZITs, I would offer homeowners as
high a mortgage as possible, while keeping principal payments to zero (perhaps even doing a
little bit of reverse mortgaging) and essentially refinancing any interest owed. For small business
owners, I would insist on a small business loan to finance their tax avoidance activity, and then
have them purchase separate LSAs as members of households.
Alas, it is true. Congress would eventually try to write some laws or Treasury would try to write
some regulations preventing my direct marketing of these ZITs. They would try to come up with
requirements similar to those that restrict direct borrowing to buy tax-exempt bonds. Of course,
such borrowing takes place all the time, as long as it cannot be traced directly. Also, state and
local bonds are limited in supply, hence people pay a premium for them, thereby limiting demand.
With my ZITs, there would be no limit on supply, hence the interest rate wouldn't have to fall on
the assets relative to what is paid on the debt.
When these new laws or regulations were promulgated, my bank would have to operate on a
more subtle basis of providing advice to clients to borrow to the hilt and then buy whatever LSAs
were available, but, of course, no longer tying the two transactions together. Hey, what do I know
when the individual's mortgage goes up by $5,000 a year and the LSA account goes up by
$5,000?
The Steuerle bank brings to light one of the greatest problems with saving incentives in current
law — the political unwillingness to treat borrowing as negative saving when subsidies are
provided to positive saving. The administration's proposal, in turn, reveals that one cannot go
much further in pushing the income tax toward a consumption tax without tackling this overriding
issue. In the meantime, we will continue to have a law that encourages excess debt and provides
saving subsidies that are often independent of any saving at all. You could say we've got a real
ZITCOM situation already.
Investment Management
Two International Place
Boston, MA 02110
January 14, 2002
Via Electronic Mail and Federal Express
Mr. Jonathan G. Katz
Secretary
U.S. Securities and Exchange Commission
450 Fifth Street, NW
Washington, D.C. 20549-0609
Re: File No.: S7-20-01- Comments on Actively Managed Exchange-Traded Funds
Dear Mr. Katz:
State Street Bank and Trust Company ("State Street") welcomes this opportunity to comment on
the U.S. Securities and Exchange Commission's (the "Commission") concept release regarding
actively managed exchange-traded funds ("ETF")(the "Concept Release"). State Street has been
a leader in the ETF marketplace since it partnered with the American Stock Exchange to launch
the SPDR Trust, Series 1 in 1993. State Street believes that actively managed ETFs will be a
valuable investment vehicle for both institutional and individual investors and encourages the
SEC to grant the exemptive relief needed to facilitate the development of actively managed ETFs.
The Concept Release identified a number of areas for which the Commission has requested
comment. For purposes of our comments, we have followed the outline of the Concept Release
and have identified in order those items for which our comments are intended to apply.
I. Executive Summary
A. Structure, Management and Operation of Actively Managed ETFs
We believe there are many potential product structures for actively managed ETFs. A key
consideration will be the amount of portfolio disclosure in the actively managed ETF design. The
management of actively managed ETFs, similar to traditional mutual funds, could vary greatly,
depending on the design of the product and nature of the underlying investment objective of the
actively managed ETF. The operations of an actively managed ETF will need to be developed
and enhanced to support the product structure and investment management goal of the actively
managed ETF.
B. Investor Use and Benefit
State Street believes that actively managed ETFs will be beneficial to investors in a number of
situations. Actively managed ETFs could potentially serve as short-term or long-term trading
vehicles, allow investors to gain exposure to an asset category in a manner similar to index-
based ETFs and play a significant role in an investor's hedging strategies. State Street believes
that like index-based ETFs, actively managed ETFs will appeal to both individual and institutional
investors.
Similar to their index-based counterparts, State Street believes that the most important uses and
benefits of actively managed ETFs will be their flexibility, tradability and tax efficiency. We also
believe that actively managed ETFs provide investors with the ability to "trade the market" in one
transaction. With traditional mutual funds, investors are forced to subscribe or redeem shares at
unknown prices. In contrast, actively managed ETFs will trade on a stock exchange throughout
the trading day. This will provide investors with more flexibility, and allow them to determine when
they want to transact.
C. Exemptive Relief
As discussed below, State Street believes that the exemptive relief granted to existing index-
based ETFs would be appropriate for actively managed ETFs. State Street anticipates that
actively managed ETFs will operate in much the same way as index-based ETFs and require the
same exemptive and no-action relief.
Further, we believe that the Commission should consider adopting rules to allow for the creation
of ETFs without the need for exemptive relief. Although situations that require additional relief
may arise, State Street believes that rulemaking would allow for more efficient creation of ETFs.
II. Index-Based ETFs vs. Actively Managed ETFs
In the Concept Release, the Commission asks if there is an appropriate way to distinguish
between index-based and actively managed ETFs, and further if there are any reasons to
distinguish between different types of actively managed ETFs. State Street believes that the
differences between existing index-based ETFs and those actively managed ETFs that plan to
maintain a similar level of transparency will not be significant. The more significant difference that
could arise with actively managed ETFs is the level of portfolio transparency that is provided to
the public. As discussed in greater detail below, those ETFs that propose to operate with less
transparency than existing index-based ETFs would be required to add additional disclosure
regarding the risks of investing in such a fund, including the possibility that such a fund may trade
at discounts and premiums that are higher than those for fully transparent ETFs. The Commission
may also wish to consider whether no-action relief from Section 16 of the Securities Exchange
Act of 1934, as amended (the "Exchange Act") is appropriate for ETFs that propose to operate
with less transparency than existing index-based ETFs.
III. Operational Issues Relating to Actively Managed ETFs
A. Importance of Arbitrage and Significance of Premium and Discounts
We believe that the ability to effectively arbitrage an ETF product is integral to the ETF trading at
or close to fair value. An ETF structure that does not have an effective arbitrage mechanism
could result in wider bid/ask spreads which, in turn, could lead to larger premiums and discounts
to the underlying NAV of the ETF. While the importance of an effective arbitrage mechanism is
clear, there are potential ways to achieve an effective arbitrage mechanism with less than full
transparency, and, potentially, with no portfolio transparency. This may be accomplished with
proper disclosure of an actively managed ETF's investment strategy and portfolio characteristics.
Therefore, placing a requirement that a certain percentage of the portfolio of a particular fund be
transparent, without taking into consideration other characteristics of its design, may be placing
undue restrictions on the ETF.
In any and all cases, clear and detailed disclosure with respect to the ETF's portfolio
transparency, arbitrage mechanism and the potential impact of the ETF trading at a
premium/discount should be required disclosure.
B. Transparency of an ETF's Portfolio
To some degree actively managed ETFs exist today. Index ETFs whose portfolio is managed
through a sampling technique, rather than the full replication of the index ("Sampled ETFs"), and
which are currently registered with the SEC and trade on the AMEX could be considered actively
managed ETFs. The portfolio holdings of a Sampled ETF generally do not fully replicate its
underlying benchmark index. Instead, a Sampled ETF invests in a selection of the securities from
the underlying index which it is tracking. With this in mind, one can argue that sampling in such a
manner allows the asset manager to actively manage the ETF's holdings. For this reason,
actively managed ETFs can be effectively categorized into two broad categories, "Transparent
Actively Managed ETFs," which exist today in the form of sampled ETFs, and "Non-Transparent
Actively Managed ETFs."
While some actively managed ETF investment managers may elect to fully disclose the contents
of their portfolio holdings, other investment managers may not. The active ETF investment
manager who does not desire full disclosure of his or her portfolio should have the option to enjoy
the same portfolio disclosure requirements of existing actively managed open-end and/or closed-
end funds.
Regardless of the extent of ETF portfolio disclosure, in order to avoid unfair advantage and
artificially creating a separate class of ETF shares, the same portfolio disclosure must be made to
all parties (i.e. market makers, brokers, retail and institutional investors, etc). Similar to an actively
managed open-end fund, an actively managed ETF should have, at a minimum, the same
requirements of portfolio disclosure. Keeping in mind, for the case of a transparent actively
managed ETF, disclosure may be greater, more frequent, or may be fulfilled by providing a
sample of the portfolio.
Greater portfolio disclosure of an ETF definitely aids in creating greater arbitrage opportunities,
narrower bid/ask spread, and lower premiums and discounts. With that said, a non-transparent
actively managed ETF will be no worse off than closed-end funds trading today. In fact, the
premium/discount of a non-transparent ETF should be narrower due to the ETF's open-ended
qualities. Because an open-end ETF allows daily redemptions and creation at NAV, the spread
(and premium/discount) should be even narrower than that experienced by similar closed-end
funds.
State Street does not believe that an investment manager of an actively managed ETF should be
required to disclose intra-day changes in the portfolio. State Street does believe, however, that in
the case of a material event, actively managed ETFs should be required to disclose changes
similar to those required by other exchange listed entities.
Further, State Street does not believe that it would be necessary at this point to require the
specified Creation or Redemption Basket to change during the day to reflect changes in the
actively managed ETF's portfolio.
C. Liquidity of Securities in an ETF's Portfolio
State Street also believes that the market should be allowed to determine which funds will be
successful and which will not be successful. We assert that funds whose investment objectives
and strategies hinder the arbitrage process will trade at higher discounts and premiums and, as a
result, may prove to be less successful than those with strategies which promote effective
arbitrage.
In the Concept Release, the Commission questions whether an actively managed ETF should be
                                                      1
permitted to invest in certain types of securities. State Street believes that actively managed
ETFs should be treated in the same manner that actively managed open-end investment
management companies are treated. We believe that portfolio managers need the flexibility to
take advantage of a variety of investment vehicles in order to effectively meet their investment
objectives.
Actively managed ETFs should be required to adhere to the Commission's policy requiring an
open-end investment company to restrict investment in illiquid securities to 15% of the fund's total
         2
assets. We believe the investors in actively managed ETFs that invest in illiquid securities would
be subjected to the same level of risk as those open-end mutual funds that invest in illiquid
securities.
State Street further supports allowing actively managed ETFs to invest in securities that cannot
be included in a Creation or Redemption Basket. Index-based ETFs currently accommodate
investors with restricted securities in their standard Creation or Redemption Baskets by allowing
custom baskets in which cash is substituted for the restricted securities. Index-based ETFs also
currently have the flexibility to allow Creation or Redemption Baskets to vary from portfolio
holdings and each other in certain situations. In addition to accommodating investors with
restricted securities in their baskets, portfolio managers of actively managed ETFs may need to
invest in certain other securities which may not be available for inclusion in Creation or
Redemption Baskets. As a result, State Street anticipates that cash or other securities will be
substituted for the unavailable securities in the same manner that index-based ETFs allow
custom baskets.
We also believe that actively managed ETFs should also be permitted to invest in securities that
are not registered under Section 12 of the Exchange Act. Currently, all "Authorized Participants,"
those parties that may purchase or redeem Creation Units, are "qualified institutional buyers." As
such, each Authorized Participant may transact in securities that are not registered under Section
12. In the event that such securities are illiquid or otherwise unavailable for inclusion in a Creation
Basket, cash or other securities may be substituted for the unavailable securities in the same
manner as discussed above. State Street further asserts that actively managed ETFs should be
permitted to invest in "unsold allotments" provided the fund does not violate Sections 12 and 17
of the Investment Company Act of 1940, as amended.
Although preferable, in-kind purchases and redemptions of Creation Units are not absolutely
necessary. While cash purchases and redemptions of Creation Units could impact the funds by
decreasing the tax efficiency, such use of cash for Creation Units may provide more efficient
portfolio management in cases where illiquid securities are included in the fund's portfolio.
D. Other Operational Issues
In general, the operations of an actively managed ETF would work similar to existing index-based
ETFs. Some potential challenges could be technology enhancement necessary to more
frequently update portfolio information if that is a requirement of an active ETF. If significant
deviations in market price and NAV of an active ETF did occur, we do not believe this would
compromise the operations of the ETF.
IV. Uses, Benefits and Risks of ETFs
A. Uses and Benefits of Index-Based ETFs
The most important use and benefit of index-based ETFs is that they enable investors to gain
almost instantaneous access (short or long) to the performance of various market indexes in a
low cost and tax efficient manner. State Street believes that expense ratios have become
increasingly important to both retail and institutional investors because they can have a significant
impact on a fund's return and the investor's potential for wealth accumulation. Additionally,
because redemptions and subscriptions occur in-kind and the cost of these transactions are
borne by the subscribing or redeeming Participant, the costs of acquiring and or disposing of
securities are less than a typical mutual fund. There are additional cost savings due to
shareholder recordkeeping being maintained by Depository Trust Company.
ETFs provide investors with the ability to "trade the market" in one transaction. With traditional
mutual funds, investors are forced to subscribe or redeem shares at unknown prices. In contrast,
ETFs trade on a stock exchange throughout the trading day. This provides investors with more
flexibility and allows them to determine when they want to transact.
Index funds generally offer greater tax benefits than their active fund counterparts because they
typically generate fewer capital gains due to lower portfolio turnover. Indexed mutual funds,
however, may be forced to realize capital gains as a result of shareholder redemptions. On the
other hand, because ETFs trade on an exchange, most shareholder activity is the result of
matching buyers and sellers, versus trading activity in the underlying portfolio. This results in
greater tax efficiency because the portfolio is not impacted by the shareholder activity that occurs
on the exchange.
B. ETFs Problems, Confusion and Undesirable Consequences
We believe that the introduction of ETFs have had a positive impact on the Capital Markets, in
general, as they have made available to all investors (institutional, retail, buy and hold and active
traders) an efficient mechanism for gaining instant access, either long-term for buy and hold
investors or short-term for active traders, to broad segments of the Capital Markets. This
mechanism has helped to "level the playing field" between retail and institutional investors as
both have access to the benefits of ETFs described above.
While investor education and proper disclosure in ETF prospectuses and fund literature have
contributed significantly to ensuring that investors understand the differences between traditional
mutual funds and ETFs, any future mandates to support actively managed ETFs should continue
this education and disclosure effort.
C. ETFs and Market Volatility
We believe that ETFs impact market volatility in a similar way that derivatives impact market
volatility. Arguments are made that because ETFs and derivatives facilitate the trading of an
underlying basket of securities, they may increase market volatility in periods of market
uncertainty as investors may use these vehicles as a means for communicating their views on the
current market. We believe this use of ETFs, as well as the use of derivatives, at times of market
volatility will help to facilitate price discovery in the marketplace.
D. Desirability and Potential Benefits of Actively Managed ETFs
Similar to their index-based counterparts, State Street believes that the most important uses and
benefits of actively managed ETFs will be their flexibility, tradability and tax efficiency. Because
we envision that redemptions and subscriptions will occur primarily in-kind, there will be fewer
internal costs associated with operating an ETF versus a typical actively managed open-end
mutual fund. Essentially, the commissions and other related expenses that mutual funds pay to
buy and sell securities triggered by shareholder activity will also be greatly reduced in an actively
managed ETF.
Similar to index-based ETFs, actively managed ETFs also provide investors with the ability to
"trade the market" in one transaction. With traditional mutual funds, investors are forced to
subscribe or redeem shares at unknown prices. In contrast, actively managed ETFs will trade on
a stock exchange throughout the trading day. This will provide investors with more flexibility, and
allow them to determine when they want to transact.
Again, because ETFs trade on an exchange, most shareholder activity is the result of matching
buyers and sellers, versus trading activity in the underlying portfolio. This results in greater tax
efficiency because the portfolio is not impacted by the shareholder activity that occurs on the
exchange.
E. Principal Uses of Actively Managed ETFs
State Street believes that actively managed ETFs will be beneficial to investors in a number of
situations. Actively managed ETFs could potentially serve as short-term or long-term trading
vehicles, allow investors to gain exposure to an asset category in a manner similar to index-
based ETFs and play a significant role in an investor's hedging strategies. State Street believes
that, like index-based ETFs, actively managed ETFs will appeal to both individual and institutional
investors.
State Street also believes that the potential for a closed-end mutual fund to be converted into an
actively managed ETF exists. However, the closed-end fund's shareholders and sponsor would
have to agree that such a conversion would be beneficial to the fund and the shareholders.
Additionally, the investment limitations applicable to an open-end fund, which would also apply to
an actively managed ETF, may not be appropriate for a closed-end fund. While the potential
exists for any fund to be an ETF, the success of the ETF will be dependent on the merits of the
investment strategy
V. Exemptive Relief from the Investment Company Act for Actively Managed ETFs
A. Relief for ETFs to Redeem Shares in Large Aggregations Only
State Street believes that actively managed ETFs will not present any new issues with respect to
the exemptions which allow for current ETFs to redeem their shares only in creation units. While
State Street recognizes that the potential for more significant deviations between market price
and NAV exists with actively managed ETFs, the fact that a fund's cash component may vary
more significantly should not affect the relief granted from the definition of "redeemable security."
Further, as mentioned above, State Street believes that the current disclosure requirements are
sufficient to safeguard against investor confusion.
B. Relief for ETF Shares to Trade at Negotiated Prices
State Street believes that actively managed ETFs will not present any new issues with respect to
the exemptions which allow ETF shares to trade at negotiated prices. State Street believes that
with proper disclosure to all parties, actively managed ETFs do not create any new potential for
discrimination or preferential treatment among investors purchasing and selling shares in the
secondary market and those purchasing and redeeming creation units. State Street would
suggest that, like existing ETFs, actively managed ETFs include disclosure which states that
individual shares of actively managed funds will be sold at market price while creation unit
aggregations of ETF shares will be processed at NAV.
C. Relief for In-Kind Transactions between an ETF and Certain Affiliates
State Street believes that actively managed ETFs will not present any new issues with respect to
the exemptions which allow for in-kind transactions between an ETF and certain affiliates. We
expect that all shareholders, regardless of affiliation, will be notified simultaneously of material
events.
D. Relief for Certain ETFs to Redeem Shares in More than Seven Days
State Street believes that actively managed ETFs will not present any new issues with respect to
the exemptions which allow certain shares to redeem in more than seven days. Actively managed
ETFs with the flexibility to invest in certain foreign securities will still need to request this relief.
VI. Potential New Regulatory Issues
A. Potential Discrimination Among Shareholders
In response to the Commission's question regarding whether investors who create and redeem in
Creation Units would be in a different position than retail investors who buy and sell ETF shares
only at market price, we assert that actively managed ETFs would present no greater risk of such
difference than existing ETFs.
B. Potential Conflicts of Interest for an ETF's Investment Adviser
In response to Commission concerns regarding conflicts of interest for an ETF's investment
adviser, we believe that actively managed ETFs present no greater risks than open-end
investment companies. We further assert that an investment adviser to an actively managed ETF
would not be in a position to create supply or demand for securities if the Creation and
Redemption Baskets are the same. And even if the Creation and Redemption Baskets were
different, there is no guarantee that daily trading activity would occur.
State Street thanks the Commission for this opportunity to provide comments on the issue of
actively managed ETFs. If you have any questions concerning these comments, please contact
the undersigned at 617-664-4489.
Sincerely,
/s/Agustin J. Fleites
Agustin J. Fleites
Principal

Footnotes
1
    State Street does not believe that limiting actively managed ETFs to certain investment
    objectives or policies designed to ensure sufficient liquidity is necessary.
2
    See Guide 4 to Form N-1A.

Limiting the scope for arbitrage
By C. R. L. Narasimhan
On July 17 the Reserve Bank of India set a ceiling on the interest rates banks can offer on certain
non-resident Indian deposit schemes. The category of deposits so affected is the NRE rupee
deposits, which from now on have been linked to LIBOR, which are well known reference rates in
the inter-bank market. The RBI has said that banks should not offer for their NRE depositors rates
that are more than 2.50 per cent of the ruling LIBOR. It is perhaps necessary to clarify what this
means to understand its implications.
London inter-bank offered rates (LIBOR) are reference rates and meant to convey the rate at
which one first class bank will lend to another in the inter-bank market. The LIBOR rates do not
represent the rates at which banks lend to most of their customers. Barring the top notch, other
bank borrowers will naturally have to pay more .The exact borrowing rates are therefore
described as 100 or 200 basis points above the LIBOR which means LIBOR plus one or two per
cent as the case may be. The other important point about Libor (and other similar reference rates)
is that they are currency specific and relate to specific periods such as three months, six months
and so on. They are therefore floating rates that can change at the end of a short period.
Domestic interest rates have tended to be higher compared to those of other major currencies.
The dollar interest rates have now touched historically low levels. A non-resident Indian, who
borrows abroad, can earn a decent spread by putting the money in an NRE deposit and that too
without any risk arising out of adverse interest rate movements. For instance when banks were
offering 5.25 per cent for a three-year deposit and the six-month LIBOR was just one per cent, it
was possible for the non-resident depositor to earn a tidy spread. By borrowing abroad at LIBOR
plus 2.50 per cent his all in costs will not be more than 3.5 per cent giving him a safe return of
1.75 per cent.
Of course there was always a catch to the arrangement. Libor rates are "floating'' meaning that
they are relevant only for short periods. Should interest rates start going up for any reason, the
non-resident borrower will be caught in a bind, as he has to pay more to the foreign bank. The
deposit rates in India are for a fixed period. In many ways the rates offered by Indian banks on
NRE deposits and the rates at which the overseas Indian raises money abroad are not
compatible as they are usually for different periods. Hence the interest advantage that the
depositor gets can be misleading. The other major constraint has to do with exchange rates. The
depositor borrows in dollars whereas the deposit is in Indian rupees. Obviously exchange
fluctuation can be hazardous, a point overlooked at a time the rupee has been appreciating. Also,
when the spread derived was of the magnitude of 2 to 3 per cent (the difference between the
depositors' cost of funds and the yield on NRE depositors) those risks — arising out of a
mismatch and in the exchange rates — were ignored.
As the markets become more integrated it is possible for an overseas investor to hedge those
risks but at the moment few individuals investors have the wherewithal. Besides hedging has a
price that may neutralise the interest advantage.
Hence what is popularly called arbitrage has taken roots through the NRE schemes. It is possible
of course that genuine savings are sent to India but the distinction between the two often gets
blurred in the NRE statistics
At a time when the country's forex reserves are at an unprecedented high level, the RBI has
thought it fit to discourage certain types of NRI inflows. Clearly those that are not in the nature of
genuine savings are going to be discouraged. The new policy is to be welcomed. Over time it will
contribute to making the reserves more stable.
However, there is an inconsistency. It is not that everyone was unaware of what was happening
on the NRE front, In fact arbitrage created deposits was accepted and encouraged through
several steps as extending the scheme to overseas corporate bodies (which can raise larger
funds than individuals). The biggest sop to the NRIs was of course the Resurgent India Bonds
(RIB) scheme of July-August 1998 (and its successor) the IMDs (India Millennium Deposits).
The RIBs are coming up for redemption soon. The RIBs paid stupendously high interest rates, a
fixed 7.75 per cent in dollars and guaranteed repatriation. Neither the interest rate risk nor the
exchange risk has been quantified in a transparent manner.
The RBI is right in capping NRE deposit rates .The RIBs though five years old carried all the
deficiencies that the RBI is now trying to rectify. It is time the entire RIB episode is made more
transparent.

								
To top