<<

. 2
( 11)



>>

to enhance the yield on one™s capital, and, lastly, insurers perceive their
¬nancial strength as a potentially signi¬cant selling point in a market where
participants are looking for ways to mitigate their exposure to counterparty
credit risk.
Just as credit derivatives were a positive for banks during the wave of
corporate defaults in the early 2000s, they proved to be disappointing
investment vehicles for some in the insurance industry, especially in Europe.
Indeed, anecdotal evidence suggests that insurers may have decided to pull
back some from the credit derivatives market in the immediate aftermath of
that spike in corporate defaults, but the extent of such pull back is di¬cult
to quantify for the industry as a whole.
2.4 Common Market Practices 25

TABLE 2.1
Market Shares of Main Buyers and Sellers of Protection
(percent)


Protection Buyers Protection Sellers

Banks 52 39
Securities Houses 21 16
Insurers/Re-insurers 6 33
Hedge Funds 12 5
Corporates 4 2
Mutual Funds 2 3
Pension Funds 1 2
Others 2 0

Source: British Bankers Association (2002).


While one may conjecture that reduced credit risk appetite by insurers
may turn out to be an impediment to the further development of the credit
derivatives market, it appears that other entities, such as pension funds,
asset managers, and hedge funds, have been increasing their participation
in the sell side of the market. Non-¬nancial corporations have reportedly
also increasingly come to the market, but primarily to buy protection to
hedge their exposure in vendor ¬nancing deals.
Table 2.1 lists the main categories of market players and their estimated
relative participation in the buy and sell sides of the market in late 2001, as
estimated by the BBA. For instance, 52 percent of the sellers of protection
during that period were banks, whereas banks accounted for only 39 percent
of the protection sellers. In contrast, as noted, insurers and re-insurers were
signi¬cant net sellers of protection: They accounted for only 6 percent of
the protection buying positions, but corresponded to about one-third of
the protection selling positions.


2.4 Common Market Practices
Thus far, this chapter has made a few main points. First, the credit deriva-
tives market has experienced phenomenal growth in recent years. Second,
commercial and investment banks, insurers and re-insurers, hedge funds,
and a few other mainly ¬nancial institutions are the main players in the
credit derivatives market, buying and selling credit protection according to
their individual needs. Third, the market has continued to grow even in
the face of unexpectedly large defaults in the early 2000s. Let us take a few
moments now to highlight some of the common practices and procedures
that have underlain the evolution of the marketplace.
26 2. The Credit Derivatives Market

2.4.1 A First Look at Documentation Issues
As noted in Chapter 1, at its most basic level a credit derivative is a legally
binding contract between two counterparties whereby the credit risk of a
third party, the reference entity, is transferred from a protection buyer to
a protection seller. Consider now a situation where each credit derivatives
dealer has its own preferred set of stipulations for, say, a contract detailing
a credit default swap. Worse still, consider that each main end-user also
has strong preferences for what should and should not be covered in the
contract and about how di¬erent key terms should be de¬ned.
Now try to imagine how costly it would be to put a contract together that
would be mutually agreeable to both counterparties; imagine the di¬culties
in arriving at fair market values for the premiums associated with each type
of contract; imagine all the legal, pricing, and back-o¬ce headaches and
costs associated with keeping track of a myriad of contracts, each with its
own idiosyncrasies. Could a market like that experience the increasing size
and liquidity that have characterized the credit derivatives market since
the mid-1990s? Unlikely.
The steady convergence of documentation standards for basic credit
derivatives contracts has played a key role in facilitating the rapid growth of
the marketplace. The adoption of commonly accepted templates for con-
tracts and of marketwide de¬nitions of key terms of the contracts have
brought a measure of commoditization to the credit derivatives market,
helping the price discovery process and reducing legal risk. The common
contract speci¬cations used by over 90 percent of the market are con-
tained in the “Master Agreement,” “Credit Derivatives De¬nitions,” and
related supplements issued by ISDA, the International Swap and Deriva-
tives Association, a trade group formed by leading swap and derivatives
market participants.
Working in consultation with its members, ISDA issued its ¬rst set of
credit derivatives documents in 1997, and has since continued to improve
on them to address evolving market needs. In essence, the standard ISDA
con¬rmation templates are akin to forms that the counterparties ¬ll out and
sign. While some core terms and de¬nitions are unchanged across contracts,
the parties have the option to choose or “check” di¬erent clauses in the
standardized contract as they apply to their circumstances. For instance,
the standardized contract for a credit default swap gives the parties the
choice to either settle the contract “physically” or “in cash.”
We will turn to documentation issues in Chapter 24. For now, an impor-
tant factor to keep in mind about credit derivatives contracts is that the
market would probably not be even close to where it is now in terms of
participation and growth if it were not for the adoption of standardized
contracts and the consequent substantial reduction in transaction costs
and legal risk.
2.4 Common Market Practices 27

2.4.2 Collateralization and Netting
As the credit derivatives market has grown, so have market participants™
exposures to one another, and the potential for substantial losses related
to a default by a major credit derivatives counterparty has not gone unno-
ticed. As noted in Chapter 1, an important step that market participants
have taken to reduce counterparty credit risk is to require the posting
of collateral against the exposures resulting from credit derivatives posi-
tions. In reality, however, because posting collateral is expensive, amounts
pledged typically cover less than the total net exposure between counter-
parties. Indeed, a common practice in the interdealer market is for market
participants to call for additional collateral after their marked-to-market
exposure to a particular counterparty has risen beyond a previously agreed
upon threshold level.
Note that the previous paragraph mentioned the “net” exposure between
counterparties, and, indeed, this netting of counterparty credit risk expo-
sures is an important feature of market functioning. Consider a simple
example. AZZ Bank and XYZ Bank have a large number of credit default
swaps between the two of them. AZZ™s total exposure to XYZ amounts to
$100 billion, whereas XYZ™s exposure to AZZ is $90 billion. Netting means
that the exposures of the two banks to one another are o¬set before any
collateral is posted so that what matters in the end is the $10 billion net
exposure of AZZ to XYZ. This would be the only amount against which
any collateral would be calculated, and this would be the claim that AZZ
would have on XYZ in the event of a default by XYZ.
Taken together, collateralization and netting, along with standard-
ized documentation, have had the e¬ect of helping overcome some of
the “growing pains” of the credit derivatives market. While standardized
documentation has helped reduce legal risk and transaction costs, collater-
alization and netting have eased concerns about counterparty credit risk,
especially as potential risk exposures through credit derivatives have grown.
3
Main Uses of Credit Derivatives




As with any other derivative instrument, credit derivatives can be used
to either avoid or take on risk, in this case credit risk. Indeed, protection
buyers are credit risk avoiders, whereas protection sellers are credit risk
takers, and, obviously, the market would not exist without either of them.
As we saw in Chapter 2, banks are the main end-users of credit derivatives,
generally as net buyers of protection, and we shall start this chapter by
discussing bank-speci¬c applications of these instruments. We will then
look at the market from the perspective of insurance companies and other
typical sellers of credit protection.



3.1 Credit Risk Management by Banks
Taking credit risk is an inherent part of banking. Bankers have traditionally
earned a substantial share of their income as compensation for bearing such
a risk. In that regard, banks do seek credit risk and view it as a necessary
part of doing business. But, as with other aspects of life, it is always possible
to “have too much of a good thing””too much credit risk, that is, not too
much income.
Banks monitor the overall credit risk in their portfolios on an ongo-
ing basis and also watch for particular concentrations of credit vis-`-vis a
a given client or industry. The old adage “don™t put all of your eggs in
the same basket” applies with force here, especially in this day and age
when shareholders, regulators, and the credit-rating agencies have been
30 3. Main Uses of Credit Derivatives

increasingly focused on stricter controls on risk exposures and capital use
by banks.
Consider, as an illustration, the case of a bank that looks at the loan
amounts outstanding to various clients and decides that its exposure to
a given large corporation (XYZ Corp.) is more than that with which the
bank™s management and investors can feel comfortable. Short of not renew-
ing existing loans and curtailing its lending to XYZ Corp., what can the
bank do? It could, for instance, sell part of its XYZ loans to other lenders
directly in the secondary loan market, assuming that market is su¬ciently
developed. Alternatively, the bank could add some of those loans to a pool
of loans to be securitized and e¬ectively sell the loans to investors in the
asset-backed securities market.1 Either way, the bank would end up reduc-
ing its credit risk exposure to XYZ Corp., which is what it wanted to do
in the ¬rst place. The bank would be happy to see its exposure to XYZ
reduced. Would XYZ Corp. be happy?
Typically, selling a loan in the secondary market requires the bank to
notify and, sometimes, to obtain consent from the borrower. The same
principle generally applies to loan securitizations. Borrowers do not always
welcome such noti¬cations enthusiastically. Some see them as a vote of no
con¬dence by their bankers. It is as if they are being told by their bank:
“Listen, we like having your business and all the income it brings to us,
but we think you are a bit too risky for our taste so we will pass some
of the loans we made you along to other banks and investors . . .” A lot
of banking is about maintaining and nurturing client relationships so the
banker will be the ¬rst to receive a call when the client decides to embark
in a new venture or expand the range of ¬nancial services it purchases from
the banking sector. This is especially relevant nowadays, when banks and
their a¬liates are much more like ¬nancial supermarkets that o¬er a whole
gamut of services ranging from bond underwriting to equity placements.
Loan sales and transfers are not always consistent with this goal.
Let us go back to the XYZ Corp. example. What if the bank decided
that it did not want to jeopardize its relationship with XYZ Corp.? It could
turn to the credit derivatives market. Suppose the bank goes out and
buys default protection in a credit default swap that references XYZ Corp.
The bank has e¬ectively reduced its exposure to XYZ Corp., just as it
would if it had sold or transferred loans made to XYZ to someone else.
(Should XYZ Corp. go under, the bank would go to its credit deriva-
tives counterparty and receive the par value of XYZ™s defaulted assets.)
Unlike a loan sale or securitization, however, XYZ Corp.™s debt remains
on the bank™s books. More important for bank relationship purposes, XYZ
Corp. need not be noti¬ed about the credit default swap transaction. In a


1
We shall discuss securitizations further in Part II of this book.
3.2 Managing Bank Regulatory Capital 31

nutshell, the bank was able to reduce its exposure to XYZ Corp. anony-
mously because the reference entity is typically not a party in a credit
derivatives contract. In market parlance, by purchasing default protection
through a credit default swap, the bank was able to “synthesize” the e¬ects
of a loan transfer or securitization”it shed the credit risk associated with
those loans”but the loans themselves never left the bank™s books. In e¬ect,
credit default swaps, and credit derivatives more generally, help banks
manage their credit risk exposure while maintaining client relationships.
Synthetic loan transfers through credit derivatives have other advantages
over traditional sales and securitizations. They often involve lower legal
and other setup costs than do sales and securitizations. Moreover, buying
protection in the credit derivatives market can be a more tax e¬cient way
of reducing one™s risk exposure. In particular, banks can shed the credit
risk of a given pool of assets without having to face the tax and accounting
implications of an outright sale of the asset pool.
Lastly, we should point out that even though the above example centered
on buying protection through a single-name credit default swap, banks
can, and very often do, use other credit derivatives instruments, such as
synthetic CDOs, to transfer large pools of credit risk and manage their
credit risk exposure. We shall discuss these instruments in greater detail in
Part II of this book.


3.2 Managing Bank Regulatory Capital
Banks are required by law to hold capital in reserve in order to cover even-
tual default-related losses in their loan portfolios. The general framework
detailing guidelines on how much capital to hold vis-`-vis loans extended to
a
di¬erent types of borrowers were ¬rst spelled out at the international level
in the 1988 Basle Bank Capital Accord. The 1988 Accord was later followed
up by the so-called Basle II Accord, which is discussed in Chapter 25, but
the underpinnings of the original Accord have historically provided banks
with an additional motivation to use credit derivatives: the management
of their regulatory capital requirements. As a result, understanding banks™
participation in the credit derivatives market requires some discussion of
the 1988 Basle Accord.

3.2.1 A Brief Digression: The 1988 Basle Accord
The 1988 Accord assigned speci¬c “risk weights” to di¬erent types of bor-
rowers and prescribed how much of the banks™ exposure to such borrowers
should be held in reserve as a sort of “rainy-day fund.” Most borrowers
received a 100 percent risk weight under the Accord, which means that
banks were required to set aside 8 percent of their total exposure to such
32 3. Main Uses of Credit Derivatives

TABLE 3.1
Risk Weights Speci¬ed in the 1988 Basle Accord for Selected Obligors

Type of Exposure Risk Weight Capital Charge
(percent) (percent)


OECD Governments 0 0.0
OECD Banks 20 1.6
Corporates and Non-OECD Banks & Gov™ts 100 8.0




borrowers in reserve. For instance, if a bank extended $10 million to a
borrower that fell under the 100 percent risk weight rule, it should incur a
$800,000 “capital charge,” i.e., it should set aside that much as a capital
reserve related to that loan.2 What if the borrower had a di¬erent risk
weight under the Accord? Then the capital charge would change accord-
ingly. For instance, for a borrower with a 1988 Basle Accord risk weight of,
say, 20 percent, banks are allowed to hold only 20 percent of the capital
they would hold for a borrower with a 100 percent risk weight, resulting in
a capital charge of 1.6 percent of the bank™s exposure to that borrower.
More generally, the regulatory capital charge speci¬ed in the 1988 Accord
obeyed the following formula

regulatory capital charge = r — .08 — notional exposure (3.1)

where r is the risk weight assigned to the borrower”e.g., r = .20 for a
borrower with a 20 percent risk weight”and the notional exposure denotes
the extent to which the bank is exposed to that particular borrower.
In practice, the 1988 Accord speci¬ed only a small number of possible
values for the risk weight r. For instance, Table 3.1 shows the 1988 Basle
risk weights assigned to the most common types of entities referenced in
the credit derivatives market: (i) governments of member countries of the
Organization for Economic Cooperation and Development (OECD) were
assigned a 0 percent risk weight, meaning that banks needed to hold no
regulatory capital for loans extended to OECD countries; (ii) OECD banks
were given a 20 percent risk weight, resulting in the 1.6 capital charge

2
The 1988 Basle Accord di¬erentiated assets held in the banking book, typically
assets held by the banks as part of their normal lending activities, from those held in
the trading book, mainly assets held for short periods as part of the bank™s trading
activities. Our focus in this brief review is on the former.
3.2 Managing Bank Regulatory Capital 33

cited in the above example, and (iii) corporates and non-OECD banks and
governments were subject to a 100 percent risk weight.3
One obvious limitation of the 1988 Accord was that it lumped all (non-
OECD bank) corporate debt under one single borrower category, assigning
a 100 percent risk weight to all corporate borrowers regardless of their cred-
itworthiness. In particular, a loan extended to an AAA-rated corporation
would result in the same 8 percent regulatory capital charge as one made
to a corporation with a below-investment-grade credit rating. In addition,
the 0 and 20 percent risk weights assigned to all OECD governments and
OECD banks are also viewed as somewhat arbitrary by many banks as not
all OECD member countries and their banks are alike. For instance, both
Mexico and the United Kingdom, two countries with very di¬erent credit
histories, are OECD member countries and thus were subject to the same
risk weight of 0 percent for their sovereign debt and 20 percent for their
banks. In market parlance, the 1988 Basle Accord did not allow for su¬-
cient “granularity” in its assignment of risk weights to di¬erent categories
of borrowers.
The end result of this lack of granularity in the 1988 Accord was that the
notion of regulatory capital was often misaligned with that of “economic
capital,” or the capital that a prudent bank would want to hold in reserve
given its overall credit risk exposure. For instance, a bank may well want
to hold more than the prescribed 8 percent charge against a loan made to a
¬rm that is now ¬nancially distressed”nothing in the 1988 Accord would
have prevented it from doing so”but that same bank might feel that the
risks associated with a loan to a top-rated ¬rm are signi¬cantly less than
what would be suggested by the mandated 8 percent charge”but here the
bank would be legally prevented from reducing its capital charge. Holding
too much capital in reserve is expensive”the bank would have to forego
the income that the held capital could generate, for instance, if it were lent
to prospective borrowers”and banks have taken measures to reduce their
regulatory capital requirements while staying within the limits prescribed
by bank regulators.

3.2.2 Credit Derivatives and Regulatory Capital
Management
One general approach banks followed to better align regulatory and eco-
nomic capital has been to move loans to highly rated borrowers”for whom
the regulatory capital charge might be deemed excessive”o¬ their bal-
ance sheets, while retaining loans to lower-rated borrowers on the balance
sheet. One way to achieve such a goal is for the bank to sell or securitize

3
OECD membership is composed of primarily industrialized economies, although
some emerging market economies are now member countries.
34 3. Main Uses of Credit Derivatives

loans made to highly rated borrowers, the net e¬ect of which would be the
freeing up of capital that was previously tied to such loans. While banks
do engage in such sales, they are mindful, as we noted in the previous
section, of possible adverse e¬ects on their customer relationships. Here
again, banks have found that the anonymity and con¬dentiality provided
by the credit derivatives market make it a desirable venue for managing
their regulatory capital.
The credit derivatives market is so young that it was not even covered by
the 1988 Basle Accord. Nonetheless, national bank regulators attempted
to treat issues related to credit derivatives in a way that is consistent with
the spirit of that Accord. For instance, when banks sell default protection
through a credit default swap, most regulators treat that as being anal-
ogous to extending a loan to the reference entity speci¬ed in the swap.
For instance, if the reference entity is a non¬nancial corporation, banks
have to incur a capital charge equal to 8 percent of the notional amount of
the contract.
Consider now a bank that has extended a loan to a highly rated corpo-
ration. As we argued above, under the terms of the original Basle Accord,
that loan would be subject to the same capital charge assigned to a less
creditworthy borrower, even though it would typically embed much less
credit risk and, consequently, a lower yield to the bank. One way for the
bank to reduce the regulatory capital charge associated with this loan,
short of selling or transferring the loan o¬ its balance sheet, would be to
buy protection against default by that corporation from an OECD bank.
If the bank regulators were satis¬ed that the credit risk associated with
the loan had been e¬ectively transferred to the OECD bank, then the
regulatory capital charge of the protection-buying bank would fall from 8
percent to 1.6 percent, re¬‚ecting the fact that, from the perspective of the
protection-buying bank, the only remaining risk exposure associated with
the loan is the counterparty credit risk associated with the OECD bank.
(The OECD bank, of course, would have to hold the full 8 percent capital
reserve in conjunction with the protection sold under the contract.)
The use of credit derivatives by banks in this type of regulatory capi-
tal management under the 1988 Basle Accord played a signi¬cant role in
the evolution of the credit derivatives market. Banks used not just credit
default swaps, but also, and by some accounts mainly, portfolio products
such as synthetic CDOs to bring their regulatory capital requirements more
in line with what they perceive to be their economic capital needs. In this
context, the credit derivatives market has helped make banks™ use of capi-
tal more e¬cient, freeing up capital set aside in excess of true fundamental
risk and putting that capital to work elsewhere in the banking system.
There is still much debate about the implications of the Basle II Accord
for the future of the credit derivatives market. While some market observers
have noted that regulatory capital management will likely be less of a
3.3 Yield Enhancement, Portfolio Diversi¬cation 35

drive for banks™ participation in the credit derivatives market, others have
noted that, with capital charges that more closely correspond to debtors™
creditworthiness, banks will have a greater incentive to move lower-quality
credits o¬ their balance sheets, and that some of that activity will take
place in the credit derivatives market. In addition, credit derivatives are
treated explicitly in the context of the Basle II Accord, and there is some
disagreement among market observers on the question of how the credit
derivatives provisions of the new Accord will a¬ect banks™ incentives to
participate in the credit derivatives market.4


3.3 Yield Enhancement, Portfolio Diversi¬cation
There are two sides to every story, and if banks see bene¬ts in using the
credit derivatives market to lay o¬ some of the credit risk in their portfolios,
there must be others for whom that market has some appeal as a place to
take on credit risk. We have mentioned already, in Chapter 2, that insurers
and re-insurers tend to be sellers of protection in the credit derivatives
market. In particular, we argued that some insurance companies see credit
risk as being essentially uncorrelated with their underwritten risks and that
protection sellers in general see credit risk exposure as a way to enhance
the return on their portfolios and diversify their risks.
Having said all that, however, there are other ways for protection sellers
to obtain the desired exposure to credit risk. They could, for instance, and
they do, turn to the corporate bond and secondary bank loan markets to
essentially buy credit risk. Is there anything about the credit derivatives
market, other than banks™ desire to buy protection, that entices protection
sellers not to limit themselves to the cash (bonds and loans) markets?

3.3.1 Leveraging Credit Exposure, Unfunded Instruments
Certain credit derivatives, including credit default swaps, are “unfunded”
credit market instruments. Unlike buying a corporate bond or extending a
loan, which requires the investor to come up with the funds to pay for the
deal upfront, no money actually changes hands at the inception of many
credit derivatives contracts.5

4
Regulatory issues are treated in greater detail in Chapter 25.
5
As noted in Chapter 2, it is not uncommon for credit derivatives contracts to require
some degree of collateralization, but posting collateral is expensive. Still, it was also
argued in that chapter that the collateral pledged often covers less than the total net
exposure between the counterparties. Moreover, even when full collateralization of net
exposures is in place, net exposures are computed with respect to marked-to-market
values of the contracts involved, and, as discussed in Chapter 16, marked-to-market
values are typically substantially less than the underlying notional amounts.
36 3. Main Uses of Credit Derivatives

Take the example of a credit default swap. In its most common form,
the two parties in the contract agree on a value for the annualized pre-
mium that the protection buyer will pay to the protection seller such that
the contract has zero market value at its inception. As a result, provided
both the protection buyer and the reference entity remain solvent while
the contract is in place, the protection seller is guaranteed a stream of pay-
ments during the life of the contract without initially putting up any cash.
In contrast, were this same protection seller to buy a bond issued by the
reference entity, it would have to pay for the bond, either by using its scarce
capital or by raising the funds in the marketplace, before it could enjoy the
periodic payments made by the bond issuer. In other words, typical cash
instruments such as bonds and loans have to be funded on the investors™
balance sheet; typical credit default swaps, as well as many other forms
of credit derivatives, largely do not. This crucial di¬erence between cash
and derivatives instruments allows investors (protection sellers) e¬ectively
to leverage up their credit risk exposure.
Let us look at another example to see how the unfunded nature of some
credit derivatives makes them particularly appealing relative to traditional
cash instruments. Consider a leveraged investor with a relatively high cost
of funds. That investor would likely ¬nd it unattractive to invest in a bond
issued by a highly rated reference entity, the reason being that the yield
it would earn on the bond would tend to be lower than the investor™s own
cost of funds. The story would be di¬erent in the credit derivatives mar-
ket, however. The investor could enter into a credit default swap with a
highly rated dealer where it sells protection against default by the cor-
poration. The investor would earn the credit default swap premium paid
by the dealer, all while avoiding at least part of its funding disadvan-
tage in the credit markets and being subject to a relatively low level of
credit risk.


3.3.2 Synthesizing Long Positions
in Corporate Debt
Another potentially appealing application of credit derivatives to investors
is the ability to obtain credit risk exposures that would otherwise not be
available through traditional cash instruments. Suppose a given institu-
tional investor would like to take on some of the credit risk associated
with XYZ Corp., but all of XYZ™s debt is locked up in loans held on
banks™ books. The investor can essentially synthesize a long position in
XYZ™s debt by selling default protection in the credit derivatives market.
In principle, the income earned via the credit derivatives contract would
be closely related to what it would be earning had it lent to XYZ Corp.
directly.
3.4 Shorting Corporate Bonds 37

3.4 Shorting Corporate Bonds
In highly liquid ¬nancial markets, such as the market for US Treasury
securities and some equity markets, investors who have a negative view
regarding future market prices can hope to pro¬t from their opinions by
establishing short positions in those markets. For example, if one thinks
that US Treasury yields are headed higher, and that this sentiment is not
fully re¬‚ected in market prices, one could sell Treasuries short in the very
active repo market for Treasury securities in hopes of buying them back
later at a pro¬t when their prices will presumably be lower.6
While a short selling strategy can generally be implemented without
signi¬cant di¬culties in su¬ciently liquid markets, its applicability to cor-
porate debt markets can be problematic. Indeed, even in the United States,
which has the most advanced corporate debt market in the world, the
ability to short sell individual corporate bonds is, at best, very limited.
In particular, the market for these securities has not yet reached the level
of liquidity and transparency that facilitate the emergence of a viable repo
market. What is the aspiring short seller to do? One option is to go, you
guessed, to the credit derivatives market.
Consider an investor who has a negative view on XYZ Corp. and thus
expects that its credit quality will deteriorate in the near term. Moreover,
the investor thinks that market prices have not yet fully accounted for such
a scenario regarding XYZ™s fortunes. While that investor may be unable
to establish a short position on XYZ™s debt, it may be able to mimic at
least part of the economic e¬ects of such a position by buying protec-
tion against XYZ Corp. in the credit default swap market. Here is how it
would work. The investor buys protection against XYZ today. Should the
investor™s views on that reference entity prove to be right, the market value
of the original credit default swap will now be positive, i.e., a newly entered
credit default swap that references XYZ would require a higher premium
from protection buyers than the one written into the investor™s contract.
The investor can thus unwind its credit default swap position at a pro¬t,
essentially cashing in on its earlier view, now con¬rmed, that XYZ™s credit
quality was headed lower.7
The example just described o¬ers two important insights into the many
uses of credit derivatives. First, one need not have any risk exposure to a

6
In the repo market, short sellers sell borrowed securities now, hoping that by the
time that they have to repurchase the securities to return them to their original owners”
repo is the market term for repurchase”their prices will have fallen enough to produce
a pro¬t.
7
In Chapter 6 we discuss how a credit default swap can be unwound. For now it suf-
¬ces to note that by unwinding we mean the e¬ective termination of the contract where
the part for whom the contract has positive market value is compensated accordingly.
38 3. Main Uses of Credit Derivatives

given reference entity in order to have a reason to buy protection against
it. Second, in addition to providing protection against the possibility of
default by particular reference entities, credit default swaps can also be
used to express views about the prospective credit quality of such entities,
and thus one can ¬nancially bene¬t from them even when no default by
the reference entity takes place.



3.5 Other Uses of Credit Derivatives
Banks and would-be short sellers are not the only ones who stand to bene¬t
from buying protection in the credit derivatives market, just as leveraged
investors are not the only ones with something to gain from selling pro-
tection. Non¬nancial corporations, certain classes of investors, and even
some banks have found other uses for credit derivatives and increased their
market participation accordingly.

3.5.1 Hedging Vendor-¬nanced Deals
Once again, let us look at a hypothetical example to better understand this
alternative use of credit derivatives. Consider an equipment manufacturer
with a few large corporate clients. To facilitate its business, the manufac-
turer also provides its customers with at least some of the ¬nancing they
need to fund their orders. While that may be good for sales, such vendor-
¬nanced deals have one obvious drawback to the manufacturer. They leave
the manufacturer exposed to the risk that its customers may default on
their obligations. One way to hedge against such risk would be to buy
protection in credit derivatives contracts that reference the individual cus-
tomers. As a result, the manufacturer can concentrate on its core business,
producing and selling equipment, while hedging out its credit risk in the
credit derivatives market.
While this use of credit derivatives is not yet widespread, market partici-
pants see the potential for it to become more so in the future. For instance,
respondents to the 2002 British Bankers Association Credit Derivatives
Report expect corporations to increase their participation in the credit
derivatives market, although they have revised down their forecasts relative
to the 2000 report.

3.5.2 Hedging by Convertible Bond Investors
Convertible bonds are corporate bonds with an embedded call option on
the bond issuer™s stock. Equity-minded investors would be natural buy-
ers of such securities, except that some may not want to have to bother
with the associated credit risk. One strategy apparently followed by many
3.6 Credit Derivatives as Market Indicators 39

convertible bond investors, including certain hedge funds, is to buy the
bonds and simultaneously buy protection against the bond issuer in the
credit default swap market. The end result of these transactions is to syn-
thesize a pure call option on the issuer™s stock. While the bond purchase
involves buying both credit risk and the call option, the former is o¬set by
the credit default swap position.
Of course, the attractiveness of the approach just described depends on
how it compares to the cost of buying a call option on the bond issuer™s
stock directly from an options dealer. Apparently, many investors occasion-
ally do ¬nd that the call option embedded in the bonds is cheap relative to
the direct purchase approach. Indeed, when liquidity in the credit deriva-
tives market is thin, the cost of buying protection against a convertible
issuer can temporarily go higher even when the market™s assessment of the
issuer™s creditworthiness is unchanged.8

3.5.3 Selling Protection as an Alternative to Loan
Origination
In discussing the ways banks use credit derivatives we have thus far por-
trayed banks as protection buyers. While available information on banks™
participation in the credit derivatives market tends to con¬rm their role
as net buyers of protection, banks do sell protection over and beyond that
amount required by their market making activities.
From a bank™s perspective, selling protection in the credit derivatives
market can be thought of as an alternative to originating loans.9 More gen-
erally, a bank may view protection selling as portfolio diversi¬cation and
yield enhancement mechanisms, i.e., as a way to obtain exposure to partic-
ular credits that would otherwise not be easily obtainable in the loan and
bond markets.



3.6 Credit Derivatives as Market Indicators
We have thus far focused on the main uses of credit derivatives strictly
from the standpoint of the entities that participate in the credit derivatives
market. Not so obvious, but potentially very important, is the growing use
by participants and non-participants alike of prices of credit derivatives,

8
This is one example where temporary demand factors, discussed in Chapter 1, can
a¬ect the pricing of credit derivatives. We will encounter more examples later in the
book.
9
We mentioned an example of this phenomenon when we discussed regional European
banks in Chapter 2.
40 3. Main Uses of Credit Derivatives

especially credit default swaps, as indicators of market sentiment regarding
speci¬c reference entities (and credit risk in general).
Investors, credit analysts, and ¬nancial regulators already have at their
disposal several indicators regarding the creditworthiness of particular
¬rms. For instance, in the United States, such indicators include yield
spreads of corporate bonds over US Treasury debt, as well as a few equity-
market-based measures of credit risk developed by well-known analytical
¬rms. Increasingly, credit default swap premia have been added to the ranks
of major indicators of perceived credit risk. Indeed, some market observers
have even suggested that prices in the credit default swap market have a
tendency to incorporate information more quickly than prices in the corpo-
rate bond market given that, at times, it may be easier to enter into swap
positions than to buy or sell certain corporate bonds and loans. Whether
information truly is re¬‚ected ¬rst in the credit derivatives or cash markets
remains a point of empirical debate, but the fact that both investors and
regulators have started to pay closer attention to signals sent out by the
credit default swap market is di¬cult to deny.
An additional potential indirect bene¬t of the credit default swap mar-
ket is the possibility that it may encourage greater integration between the
corporate bond and bank loan markets, two segments of the credit mar-
kets that have remained largely segregated despite their natural common
ground. In particular, in part because most credit default swaps generally
are physically settled and allow the delivery of either bonds or loans in the
event of default, and in part because banks have been stepping up their use
of credit derivatives to manage their economic capital, one might expect
that banks will increasingly turn to prices observed in the credit derivatives
market as important inputs into their own lending decisions. Should this
phenomenon come to pass, it will have the potential to encourage greater
e¬ciency and discipline in the credit markets in general, with credit risk
being more transparently and consistently priced across the marketplace.
Part II

Main Types of Credit
Derivatives




41
4
Floating-Rate Notes




Floating-rate notes, FRNs or ¬‚oaters for short, are among the simplest
debt instruments. They are essentially bonds with a time-varying coupon.
In this chapter we go over the basics of FRNs and introduce some notation
and concepts that will be used throughout the remainder of the book.


4.1 Not a Credit Derivative...
Floating-rate notes are not credit derivatives, but they are featured promi-
nently in the discussion of so many of them”such as credit default swaps,
asset swaps, and spread options”that we decided to give them their own
chapter in this book.
The main reason for the close link between FRNs and credit derivatives
is that, as we shall see below, the pricing of a ¬‚oater is almost entirely
determined by market participants™ perceptions of the credit risk associated
with its issuer. As such, ¬‚oaters are potentially closer to credit default
swaps than to ¬xed-rate corporate notes, which, as the name suggests, are
bonds with a ¬xed coupon.


4.2 How Does It Work?
The variable coupon on a ¬‚oating-rate note is typically expressed as a ¬xed
spread over a benchmark short-term interest rate, most commonly three- or
44 4. Floating-Rate Notes

six-month LIBOR (London Interbank O¬ered Rate). LIBOR is the rate at
which highly rated commercial banks can borrow in the interbank market.
Therefore, one can think of LIBOR as re¬‚ecting roughly the credit quality
of borrowers with credit ratings varying between A and AA, and thus such
borrowers are able to issue FRNs with a spread that is either zero or close
to it. Intuitively, someone with a lower/higher credit quality than these
borrowers would presumably issue ¬‚oaters with a positive/negative spread
over LIBOR.
The mechanics of an FRN is quite simple and can be best understood
with an example. Consider a corporation that needs to raise $100 million
in the capital markets and that has decided to do so by issuing four-year
¬‚oaters that will pay coupons semiannually. Floaters are typically issued
at “par” or very close to par, meaning that their initial market price will
be equal or very close to their face value, $100 million in this case. Suppose
that in order for the corporation™s ¬‚oaters to be issued at par, they have
to be issued with a ¬xed spread of 80 basis points over six-month LIBOR.
This is called the par ¬‚oater spread or the spread that makes the ¬‚oater
be priced at its face value. (We will address pricing issues in further detail
later. For now, we can deduce that this issuer has a credit rating that is
likely inferior to A given that it had to o¬er a positive spread over LIBOR
in order to be able to sell its ¬‚oaters at par.)
When the ¬‚oater is issued, investors know what their ¬rst coupon
payment will be, although the actual coupon payment will be received
only six months forward. In particular, that coupon will be the sum of
the current value of six-month LIBOR plus the 80 basis point spread
required to sell the ¬‚oater at par. Thus, assuming that six-month LIBOR
is 6 percent, the ¬rst coupon will be equal to 6.8 percent, or 3.4 percent
on a semiannual basis. Future coupons are not known in advance as they
will be reset on each payment date according to the then prevailing six-
month LIBOR. For instance, suppose that when the ¬rst coupon becomes
due, six-month LIBOR happens to be 6.2 percent. That would result in
the ¬‚oater™s second coupon being reset to 7 percent on an annual basis.
The process continues like this until the end of the four-year period cov-
ered by the ¬‚oater. As with standard ¬xed-income bonds, the last payment
will also include the repayment of the full $100 million face value of the
¬‚oater.
Table 4.1 details the cash ¬‚ows of the four-year ¬‚oater under consider-
ation using hypothetical values for six-month LIBOR over the life of the
instrument. Again, at any point in time, investors only know the value of
the next coupon payment. The size of subsequent payments will be deter-
mined one at a time at each reset date of the ¬‚oater. All that investors
know about these future payments is that they will be based on an annu-
alized coupon 80 basis points higher than the six-month LIBOR prevailing
at the immediately preceding reset date.
4.4 Valuation Considerations 45

TABLE 4.1
Cash Flows of a Hypothetical Floatera
(Face value = $100 million; Spread = 80 basis points)


Reset Date 6-month LIBOR Coupon Cash Flow
(years from now) (percent) (percent, annual) ($ millions)

(1) (2) (3) (4)
’100.00
0 6.0 6.8
.5 6.2 7.0 3.40
1 6.2 7.0 3.50
1.5 6.1 6.9 3.50
2 6.0 6.8 3.45
2.5 5.8 6.6 3.40
3 5.8 6.6 3.30
3.5 5.7 6.5 3.30
4 5.4 ” 103.25
aA negative cash ¬‚ow denotes a payment from the investor to the note issuer.




4.3 Common Uses
Banks are relatively frequent issuers of FRNs, especially in Europe.
US corporations occasionally do issue ¬‚oaters as well, but it is fair to say
that the market for FRNs is signi¬cantly smaller than that for ¬xed-rate
corporate bonds. Some sovereign debt is also issued in the form of FRNs.
From the credit investor™s standpoint, FRNs have one key advantage
over their ¬xed-rate cousins. As noted, the value of a portfolio of ¬‚oaters
depends almost exclusively on the perceived credit quality of their issuers
represented in the portfolio. Thus, similar to someone who has sold protec-
tion in a credit default swap, the FRN investor has primarily bought some
exposure to credit risk. In contrast, an investor who is long a corporate
bond with a ¬xed coupon is exposed primarily to both credit and interest
rate risk, the latter arising from the fact that prices of ¬xed-coupon bonds
move in the opposite direction of market interest rates. (As we shall see
below, ¬‚oaters have very little interest rate risk.)


4.4 Valuation Considerations
Let us take a closer look at the factors that in¬‚uence the pricing of a
¬‚oating-rate note. As the above example suggested, the most important
factor is, by far, the credit quality of the issuer. Our intuition should
46 4. Floating-Rate Notes

tell us that the riskier the issuer, the higher the spread over LIBOR will it
have to o¬er in the marketplace in order to attract willing investors. At the
same time, an AAA-rated issuer would generally be able to sell its notes
at a negative spread relative to LIBOR as its creditworthiness would be
superior to that of even the highly rated banks that borrow and lend in
the interbank market.
While understanding the relationship between FRN spreads and credit
risk is straightforward, the fact that ¬‚oaters are relatively insensitive to
other types of risk, particularly interest risk, may not be so obvious. To see
this, let us consider ¬rst a ¬xed-rate note, perhaps the corporate debt
security with which people are the most familiar. For simplicity, assume
that the note was bought for its par value. Suppose a couple of months
go by and market interest rates rise unexpectedly. What happens to the
market value of the note? The purchaser of the note ¬nds itself in the
unenviable position of holding a security that now pays a coupon based on
a below-market interest rate. The market value of the bond naturally falls.
Of course the opposite would have been true if market interest rates had
fallen. The main point here is that the ¬xed-rate note investor does not
know whether market interest rates will rise or fall. That is the nature of
interest rate risk!1
Would the holder of a ¬‚oating-rate note, also bought for par and at the
same time as the ¬xed-rate note, fare any better than the ¬xed-rate note
investor under the same circumstances? Suppose that the sudden rise in
market interest rates happened just before a reset date for the ¬‚oater. That
means that the ¬‚oater™s coupon will soon be increased to re¬‚ect the recent
rise in market rates. More important, the rise in the coupon will be such
that, provided the credit quality of the issuer has remained the same, the
¬‚oater will continue to be valued at par. Not much interest rate risk here!
When market rates increase, the coupon increases; when rates decline, the
coupon declines . . . and the ¬‚oater continues to be valued at par. (We will
show this in a numerical example below, after we introduce some simple
valuation principles more formally.)
The careful reader probably noticed that we assumed that the sudden
rise in market interest rates happened just before one of the ¬‚oater™s reset
dates. What if the increase in market rates had happened right after a
reset date? Given that the ¬‚oater™s coupon changes only at reset dates,
the holder of the ¬‚oater would temporarily be receiving a coupon based on

1
There is one sense in which a ¬xed-rate note investor may not care about interest
rate risk. Suppose the investor paid par for the note, plans to hold on to it until its
maturity, and is not particularly concerned about what will happen to its value during
the intervening time. Assuming no default by the note issuer, the investor will receive
par back when the note matures regardless of where market interest rates are at that
point.
4.4 Valuation Considerations 47

a below-market rate. The key word here is “temporarily” as the ¬‚oater™s
coupon would eventually adjust to prevailing market rates and thus the
resulting e¬ect on the market price of the ¬‚oater is generally very small.
The risk of movements in market rates in between reset dates is what
market participants call “reset risk.” We will examine reset risk more closely
below. For now it su¬ces to say that reset risk is not a major factor in
the pricing of ¬‚oaters as the time between reset dates is typically short,
generally six months at the most.
Table 4.2 uses a numerical example to highlight key aspects regarding
the pricing of ¬‚oating-rate notes and to draw some comparisons with ¬xed-
rate notes. Let us continue to use the hypothetical four-year ¬‚oater detailed
in Table 4.1; indeed columns (1) through (3) and (5) are taken directly
from Table 4.1. What is new in Table 4.2 is the fact that we now also
consider a hypothetical ¬xed-rate note issued by the same corporation.
By assumption, the ¬xed-rate note pays a ¬xed-coupon of 6.8 percent per
annum”the same initial coupon of the ¬‚oater”and has the same maturity,
face value, and payment dates as the ¬‚oater. We now also make the explicit
simplifying assumption that the corporation faces a ¬‚at term structure of
interest rates, i.e., it can borrow at a ¬xed rate of 6.8 percent regardless of
the maturity of its debt. The table shows, in the ¬rst row of columns (5)
and (6) that the investor paid the par value of both securities at time 0.
How can we show that these are “fair” market prices? Let us start with the
¬xed-rate note, which in some respects is easier to price than the ¬‚oater.
A basic valuation principle for any ¬nancial security is that its market
value today should re¬‚ect the (appropriately de¬ned) expected present (dis-
counted) value of its future cash ¬‚ows. In the absence of default, we know
exactly what the cash ¬‚ows of the ¬xed-rate security will be. The only thing
we need to do to derive the market value of the ¬xed-rate note is discount
these future cash ¬‚ows to express them in terms of current dollars and add
them up. Let D(0, t) denote the discount factor that corresponds to today™s
(time-0) value of a dollar to be received at time t. Those familiar with sim-
ple bond math will recognize D(0, t) as today™s price of a zero-coupon bond
that will mature at time t”see Appendix A for a brief refresher on basic
concepts from bond math. Assuming that the debt instruments of the issuer
have no recovery value in the event of default, we can write today™s value
of the four-year ¬xed-rate note, denoted below as V F X (0, 8), as

8 ¯
C
FX
V (0, 8) = D(0, t) + D(0, 8) P (4.1)
2
t=1


¯
where C is the ¬xed annual coupon rate paid by the note at time t”as
¯
shown in Table 4.2, C = 6.8 percent or 0.068”P is the face value of the
note”$100 million”and D(0, t) is the discount factor, as seen from time 0,
48 4. Floating-Rate Notes

TABLE 4.2
Valuation of Fixed- and Floating-rate Notesa
(Face value = $100 million; Spread = 80 basis points)


Reset Date 6-month LIBOR Coupons Cash Flows
(yrs hence) (percent) (percent, annual) ($ millions)
Floater Fixed Floater Fixed

(1) (2) (3) (4) (5) (6)
’100.00 ’100.00
0 6.0 6.8 6.8
.5 6.2 7.0 6.8 3.40 3.40
1 6.2 7.0 6.8 3.50 3.40
1.5 6.1 6.9 6.8 3.50 3.40
2 6.0 6.8 6.8 3.45 3.40
2.5 5.8 6.6 6.8 3.40 3.40
3 5.8 6.6 6.8 3.30 3.40
3.5 5.7 6.5 6.8 3.30 3.40
4 5.4 ” 6.8 103.25 103.40
aA negative cash ¬‚ow denotes a payment from the investor to the note issuer.




relevant for a cash ¬‚ow received at time t.2 Thus, for instance, given the
¯
assumption of semiannual payments, D(0, 2) C P is today™s value of the
2
coupon payment that will be received one year from the present date.
Which zero-coupon bond prices should we use to discount the future
payments promised by the note issuer? If there were not default risk asso-
ciated with the issuer, one might just use zero-coupon bond prices derived,
say, from the US Treasury yield curve, which is typically assumed to be
a good representation of the term structure of default-free interest rates.
In the presence of credit risk, however, the prudent investor might want
to discount these future payments at a higher rate. How much higher?
The answer is in the issuer™s own yield curve, which we assumed to be
¬‚at at 6.8 percent in this example. Given this yield curve, one can derive
the prices of zero-coupon bonds associated with the issuer. For now we
will focus on discretely compounded rates. As shown in Appendix A, for
j = 1, 2, . . . , 8, one can write D(0, j) as

1
D(0, j) = (4.2)
(1 + R(0, j)/2)j

2
The zero-recovery assumption means that the note becomes worthless upon default.
We make this assumption here for the sake of simplicity only. Di¬erent recovery
assumptions are discussed in Part III.
4.4 Valuation Considerations 49

where R(0, j) is the semiannually compounded zero-coupon bond yield,
derived from the issuer™s yield curve, that corresponds to the maturity
date j. Given the ¬‚at yield curve assumption, R(0, j) = .068 for all maturi-
ties j. Thus, we obtain D(0, 1) = 0.9671, D(0, 2) = 0.9353, etc. By carrying
on with these calculations and substituting the results into equation (4.1)
we ¬nd that V F X (0, 8) = $100 million, which is what is shown in Table 4.2.
We can use similar principles to value the ¬‚oater, except that in this
case the associated future cash ¬‚ows are unknown even in the absence of
default. For instance, we do not know today what six-month LIBOR will
be in one year™s time. How can we verify that the ¬‚oater is indeed worth
par given the assumptions underlying Table 4.2?
It can be shown that the following relationship holds for a par ¬‚oater:
N
F — (0, t ’ 1, t) + S
ˆ
P= D(0, t) + D(0, N ) P (4.3)
2
t=1

where F — (0, t ’ 1, t) is the discretely compounded forward LIBOR rate, as
seen at time 0, that corresponds to a future loan that will start at a future
date t ’ 1 and end at t, and S is the issuer™s par ¬‚oater spread. Thus, the
ˆ
forward rate that applies to this particular issuer would be F — (0, t’1, t)+ S.
ˆ
By de¬nition of the forward rate, see Appendix A, we have

D(0, t ’ 1)
ˆ1
F — (0, t ’ 1, t) + S ≡ ’1 (4.4)
δ D(0, t)

where δ is the accrual factor that corresponds to the period between t ’ 1
and t. For instance, if F — + S is expressed on an annual basis, δ = .5
ˆ
corresponds to the case where the period [t ’ 1, t] is equal to one-half of
a year.
If we substitute the above expression for the forward rate into equation
(4.3) we ¬nd that the term in square brackets simpli¬es to 1 so that (4.3)
holds, verifying what we wanted to show.
Going back to the example in Table 4.2, we can use equation (4.4) to com-
pute the forward rates associated with the note issuer and check whether
the assumed spread of 80 basis points over LIBOR is indeed consistent with
the ¬‚oater being sold par at time 0. It is!
Note that we can use an expression analogous to (4.3) to price any ¬‚oater,
not just a par ¬‚oater. In particular, we can write:
N
F — (0, t ’ 1, t) + S
FL
V (0, N ) = D(0, t) + D(0, N ) P (4.5)
2
t=1

where S is the ¬‚oater™s spread, which is not necessarily the same as the
ˆ
par spread S. Thus, if we know the discount curve for a particular issuer,
50 4. Floating-Rate Notes

we can ¬nd the fair market value of a ¬‚oater paying a generic spread S
over LIBOR.
We are now in a position to take another look at the ¬‚oater™s sensitivity
to changes in market interest rates (reset risk) and to changes in the credit
quality of the issuer. Let us consider ¬rst a surprise across-the-board 50
basis point increase in market rates”a parallel shift in the yield curve.
We consider two scenarios. In the ¬rst scenario, six-month LIBOR increases
50 basis points six months after it was issued, just before the ¬‚oater™s ¬rst
reset date, but immediately after the ¬rst coupon payment is received.
In the second, the rise in rates happens immediately after the ¬‚oater was
issued.
As noted, the ¬rst case illustrates a situation of minimum, virtually
zero, reset risk. Rates rise just before the reset date so the higher level is
immediately re¬‚ected in the ¬‚oater™s coupon. The end result, which can
be veri¬ed with help from equation (4.5) is that the market value of the
¬‚oater is not a¬ected by the rise in market interest rates.
The second scenario illustrates a case of maximum reset risk. Market
interest rise unexpectedly but investors have to wait for a full accrual period
(six months) before the ¬‚oater™s coupon will be adjusted. What happens
to the market value of the ¬‚oater? It falls 0.24 percent, from its par value
of $100 million to $99.76 million. Once the reset date does arrive, however,
the ¬‚oater™s coupon is adjusted to the prevailing LIBOR, and the ¬‚oater™s
price reverts to par.
How does the interest rate sensitivity of the hypothetical ¬‚oater in the
example in Table 4.2 compare to that of the ¬xed-rate note? Using equation
(4.1), we ¬nd that the price of the ¬xed-rate security falls about 1.5 percent
and 1.7 percent in the ¬rst and second scenarios, respectively. Another way
to compare the interest rate sensitivities of the ¬‚oating- and ¬xed-rate notes
is to look at their durations, which one can approximate with the following
calculation:

change in price 1
Duration = ’ (4.6)
change in interest rate initial price

This calculation puts the duration of the ¬‚oater between 0 and just under
6 months and that of the ¬xed-rate note at between 3 and 3 1 years.
2
An interest rate sensitivity measure that is closely related to duration
is PV01 sensitivity, de¬ned as the change in the price of an instrument
in response to a parallel one basis point shift in both the issuer™s and
benchmark yield curves. The PV01 sensitivities of the hypothetical ¬‚oating-
rate note in Table 4.2 were $0 and ’$4,835 in the two scenarios examined,
given the $100 million initial price of the ¬‚oater. For the ¬xed-rate note,
the PV01 sensitivities were substantially more pronounced, at ’$30,681
and ’$34,506 for the two cases analyzed and same initial price.
4.4 Valuation Considerations 51

While the ¬‚oater has very little sensitivity to changes in market interest
rates, a sudden deterioration in the credit quality of the issuer would have
a noticeable e¬ect on the prices of both the ¬‚oating- and ¬xed-rate notes.
In particular, continuing to use Table 4.2 as an illustration, suppose the
issuer™s yield curve shifts up 50 basis points in a parallel fashion while
other market interest rates remain unchanged. Looking at the ¬‚oater ¬rst,
both the forward LIBOR curve and the 80 basis spread written into the
¬‚oater are unchanged in equation (4.5). Nonetheless, all future payments
to be received from the issuer will now be discounted at a higher rate to
re¬‚ect the issuer™s lower credit quality. As a result, assuming the widening
in spreads happens immediately after the notes were issued, the price of
the ¬‚oater declines 1.7 percent. In this particular example the price of the
¬xed-rate note also declines 1.7 percent as a result of the deterioration in
the issuer™s credit quality.3
To sum up, the above examples highlight two key points about ¬‚oating-
rate notes. First, they have very little sensitivity to changes in market
interest rates. Second, changes in their prices re¬‚ect mainly changes in the
creditworthiness of issuers. As such, ¬‚oating-rate notes are closely related
to many types of credit derivatives, an issue we shall discuss in greater
detail in the next two chapters.




3
A related concept is that of spread duration, which measures the sensitivity of a
bond™s price to a change in its yield spread over a given benchmark rate”typically the
yield on a comparable government bond. In the above example, both the ¬‚oater and the
¬xed-rate bond have a spread duration of about 3 1 years.
2
5
Asset Swaps




Asset swaps are a common form of derivative contract written on ¬xed-rate
debt instruments. The end result of an asset swap is to separate the credit
and interest rate risks embedded in the ¬xed-rate instrument. E¬ectively,
one of the parties in an asset swap transfers the interest rate risk in a
¬xed-rate note or loan to the other party, retaining only the credit risk
component. As such, asset swaps are mainly used to create positions that
closely mimic the cash ¬‚ows and risk exposure of ¬‚oating-rate notes.



5.1 A Borderline Credit Derivative...
There is some disagreement among credit derivatives market participants
on whether an asset default swap is a credit derivative. Some apparently
focus on the fact that the asset swap can be thought of as not much more
than a synthetic ¬‚oater, and a ¬‚oater is de¬nitely not a credit derivative.
Others seem to emphasize the fact that asset swaps can be thought of as
a way to unbundle the risks embedded in a ¬xed-rate security, isolating
its credit risk component, much like what other credit derivatives do. For
these and other reasons, the di¬erence in opinions regarding asset swaps
persists. Indeed, while Risk Magazine™s 2003 Credit Derivatives Survey
(Patel, 2003) decided to exclude asset swaps from its range of surveyed
products, the 2002 British Bankers Association Credit Derivatives Report
included assets swaps in its credit derivatives statistics. The BBA acknowl-
edged the ongoing debate among market participants, but reported that
54 5. Asset Swaps

a majority of key participants considers asset swaps as part of their credit
derivatives activities.
Regardless of where the asset swap debate settles, however, asset swaps
are important in their own right if one™s goal is to develop a better under-
standing of credit derivatives. For instance, similar to ¬‚oating-rate notes,
asset swaps are closely related to credit default swaps, which are probably
the best-known type of credit derivative.


5.2 How Does It Work?
Those familiar with interest rate swaps will ¬nd several similarities between
the mechanics of an asset swap and that of an interest rate swap. As with
an interest rate swap, an asset swap is an agreement between two parties
to exchange ¬xed and variable interest rate payments over a predetermined
period of time, where the interest rate payments are based on a notional
amount speci¬ed in the contract. Unlike the vanilla interest rate swap,
however, where the variable rate is LIBOR ¬‚at and the ¬xed rate is deter-
mined by market forces when the contract is agreed upon, the ¬xed rate
in an asset swap is typically set equal to the coupon rate of an underlying
¬xed-rate corporate bond or loan, with the spread over LIBOR adjusting
to market conditions at the time of inception of the asset swap.
But this is not a book about interest rate swaps, nor is detailed knowledge
about such instruments a prerequisite for reading this chapter.1 Figure 5.1
lays out the basic features of an asset swap. Consider an investor who wants
to be exposed to the credit risk of XYZ Corp. (the reference entity), but
who wants to minimize its exposure to interest rate risk. Assume further
that XYZ Corp.™s debt is issued primarily in the form of ¬xed-rate bonds,
which, as we saw in Chapter 4, embed both credit and interest rate risk.
The investor can enter into an asset swap with a dealer where the investor
will be a ¬‚oating-rate receiver. (In market parlance, the investor is called
the asset swap buyer and the dealer is the asset swap seller.)
The typical terms of the agreement are as follows.
• The investor (the asset swap buyer) agrees to buy from the dealer (the
asset swap seller) a ¬xed-rated bond issued by the reference entity,
paying par for the bond regardless of its market price. This is shown
in the upper panel of Figure 5.1.
• The investor agrees to make periodic ¬xed payments to the dealer
that are equal to the coupon payments made by the reference bond.
The investor essentially passes through the coupon payments made by

1
See Hull (2003)[41] for an overview of interest rate swaps.
5.2 How Does It Work? 55




FIGURE 5.1. Diagram of an Asset Swap



the bond to the dealer. In return the dealer agrees to make variable
interest rate payments to the investor, where the variable payments
are based on a ¬xed spread over LIBOR, and the notional principal
is the same as the par value of the reference bond. This is shown in
the lower panel of Figure 5.1. The coupon payments made by bond
and passed along by the investor are denoted as C, and the variable
rate payments made by the dealer are denoted as L + A, where L
stands for the relevant LIBOR”typically corresponding to the three-
or six-month maturity”and A is the so-called asset swap spread.
• A is typically set so that the initial market value of the asset swap is
zero. This implies, for instance, that if the reference bond is trading
at a premium over its face value, dealers must be compensated for the
fact that they sell the bond to the asset swap buyer for less than its
56 5. Asset Swaps

market value. That compensation comes in the details of the interest
rate swap embedded in the asset swap. In the case of a bond that is
trading at more than its par value, A must be such that the dealer™s
position in the swap has a su¬ciently positive market value at the time
of inception of the swap so it will compensate the dealer for selling the
bond to the investor “at a loss.”2

Assuming there is no default by the reference entity during the life of
the asset swap, what happens at its maturity date?
• The investor receives the par value of the reference bond.

• The dealer and the investor are freed from their obligations under the
swap after they make their last exchange of payments. As with an
interest rate swap, there is no exchange of notional principals at the
end of the swap.
What if the reference entity defaults while the asset swap is still in force?
• The asset swap lives on, that is, all rights and obligations of the dealer
and the investor are una¬ected by the reference entity™s default.
• The investor, however, loses the original source of funding for the ¬xed
rate payments that it is obliged to make to the dealer. The investor
also loses its claim on the full par value of the bond, receiving only
the bond™s recovery value upon the reference entity™s default.

If the reference entity defaults, the investor and the dealer may decide
to terminate the interest rate swap by cash settling it. They will look at
the market value of the interest rate swap at that point in time, and the
party for whom the interest rate swap has positive market value will be
paid accordingly by the other in order to terminate the swap. For instance,
suppose both parties agree that the interest rate swap has positive market
value of Y to the dealer. In order to the terminate the swap, the investor
will pay Y to the dealer.




5.3 Common Uses
Investors who want exposure to credit risk, without having to bother much
about interest rate risk, but who cannot ¬nd all the ¬‚oaters they want, can
go to the asset swap market to “buy synthetic ¬‚oaters.” Likewise, investors


2
We will look at this more explicitly in Section 5.4.
5.3 Common Uses 57

who fund themselves through ¬‚oating-rate instruments but who hold ¬xed-
rate assets might want to transfer the interest rate risk of their assets to
someone else by becoming buyers of asset swaps.
Banks are signi¬cant users of asset swaps as they tend to fall in the
category of investors mentioned at the end of the last paragraph. Banks tra-
ditionally fund themselves by issuing ¬‚oating-rate liabilities, e.g. deposits,
but may hold some ¬xed-rate assets on their balance sheets, such as ¬xed-
rate bonds and some ¬xed-rate loans. As we saw in Chapter 4, the interest
sensitivity of ¬‚oating-rate instruments is vastly smaller than that of ¬xed-
rate instruments. As a result, banks and other investors with potentially
sizable mismatches between the durations of their liabilities and assets may
bene¬t from buying asset swaps. For instance, a ¬‚oater-funded investor who
invests in ¬xed-rate bonds could be substantially adversely a¬ected by a
sudden rise in market interest rates: Although the value of its liabilities
would be relatively una¬ected by the rise in rates, that of its assets could
potentially plunge.
Some market participants, such as hedge funds, use the asset swap
market to exploit perceived arbitrage opportunities between the cash and
derivatives market. As discussed above, buying an asset swap is akin to
buying a synthetic ¬‚oater issued by the entity referenced in the swap.
It should then be the case that, in the absence of market frictions such as
segmentation, poor liquidity, and other transactions costs, par asset swap
spreads will tend to remain close to par ¬‚oater spreads.3 When the two
spreads are deemed to be su¬ciently out of line that associated transac-
tions costs of taking positions in both markets are not binding, arbitrageurs
will spring into action. For instance, suppose the asset swap spread associ-
ated with contracts referencing XYZ Corp. are perceived to be too narrow
relative to the spreads paid by ¬‚oaters issued by XYZ. An arbitrageur could
exploit this apparent misalignment in prices in the asset swap and ¬‚oater
markets by selling asset swaps written on XYZ ¬xed-rate debt and buying
XYZ ¬‚oaters.4
Asset swaps can be used to leverage one™s exposure to credit risk.
For instance, consider an investor who wants to buy a corporate bond
that is trading at a substantial premium over its face value. The investor
can either buy the bond in the open market for its full market value or buy
it through a par asset swap where the initial cash outlay would be only the
bond™s face value. Of course, as with other leveraging strategies, buying
the asset swap in this case has the implication that the investor could lose
more than the initial cash outlay in the event of default by the bond issuer:
If you bought the bond through a par asset swap, and thus paid less than

3
We will take a closer look at the relationship between asset swap spreads and ¬‚oater
spreads towards the end of this chapter.
4
This arbitrage strategy is discussed further in Section 5.4.2.
58 5. Asset Swaps

the market price of the bond because you also entered into an asset swap
that had negative market value to you. Thus, if and when the bond defaults,
you may ¬nd yourself in a situation where not only you incur a default-
related loss in the bond, as would an investor who bought the bond directly
in the open market, but you could also ¬nd yourself with an interest rate
swap that is still against you. In contrast, if you bought the bond directly
in the open market, your maximum loss will be your initial cash outlay.


5.4 Valuation Considerations
To understand what goes into the pricing of an asset swap (AP), it is best to
think of it as a “package” involving two products: (i) a ¬xed-rate bond (B),
which is bought by the investor from the dealer for par, and (ii) an interest
rate swap (IRS) entered between the investor and the dealer. Thus, from
the perspective of the asset swap buyer (the investor), the market value of
the asset swap can be written as

V AP (0, N ) = [V B (0, N ) ’ P ] + V IRS (0, N ) (5.1)

where P denotes the face value of the bond, and the notation V Y (.) is used
to represent the market value of Y , whatever Y may be. For the interest
rate swap, we express its market value to the asset swap buyer; the market
value to the asset swap seller would be the negative of this quantity.
In words, equation (5.1) says that the value of the asset swap to its
buyer at its inception is essentially given by the sum of its two components.
The buyer paid par for the bond, but its actual market value, V B (0, N ),
could well be di¬erent from par, and thus the buyer could incur either
a pro¬t or a loss if the bond were to be immediately resold in the open
market. This potential pro¬t or loss is shown by the term in brackets on
the right side of the equation. The second component in the valuation of the
asset swap is the market value of the embedded interest rate swap between
the buyer and the dealer, denoted above as V IRS (0, N ). That swap too
may have either positive or negative market value to the buyer.
We see then that the market prices of the reference bond and the embed-
ded interest rate swap determine the market value of the asset swap.
Now, the market value of the bond is given to both the dealer and the
investor so the main issue of negotiation between the two of them would be
the interest rate swap component, more precisely, the spread A over LIBOR
that will be part of the variable-rate payment made to the investor”both
LIBOR and the bond™s coupon, which determines the ¬xed leg of the swap,
are also outside of the control of either the dealer or the investor.
Market convention is to set the asset swap spread so that the market
value of the entire package, the asset swap, is zero at the inception of
5.4 Valuation Considerations 59

the contract. This means that, if the market price of the bond is above its
face value”e¬ectively meaning that the investor bought the bond from the
dealer for less than its market price”the dealer must be getting something
in return. Indeed, for the privilege of buying the bond at below its market
price in this example, the investor agrees to enter into an interest rate swap
that has positive market value to the dealer. Thus, for a par asset swap, the
asset swap spread will be set such that the market value of the interest rate
swap will exactly o¬set the di¬erence between the par and market values
of the bond.


5.4.1 Valuing the Two Pieces of an Asset Swap
Valuing the bond piece of the asset swap is relatively straightforward.
We have seen how to do it in Chapter 4”equation (4.1). For convenience,
let us repeat the relevant equation here:

N
¯
B
V (0, N ) = D(0, i)δi C + D(0, N ) P (5.2)
i=1



where V B (0, N ) is the time-0 price of a ¬xed-rate bond that matures at time
¯
N ; C is the ¬xed annual coupon paid by the bond at time i; P is its face
value; and D(0, i) is the time-0 discount factor, derived from the issuer™s
yield curve, that represents the present value of $1 payable by the issuer
at a future time i.5 δi is the accrual factor that corresponds to the period
between i ’ 1 and i”for instance, if the bond pays coupons semiannually,
δi = .5. In practice, of course, the price of the ¬xed-rate bond may be
directly observable in the market, but we write down the above expression
here as it will come in handy in the valuation of the embedded interest
rate swap.
To ¬nd the market value of the embedded interest rate swap, we should
remember the fundamental principle that the market price of any security
essentially is equal to the appropriately discounted value of the future cash
¬‚ows associated with the security (see Chapter 4). Now consider the fact
that an interest rate swap can be thought of as an exchange of a ¬xed- for
a ¬‚oating-rate liability and we are essentially home! Recall that the asset
swap buyer agrees to pay ¬xed in the interest rate swap, which is akin to
selling a ¬xed-rate bond to the dealer. In particular the market value of
the so-called ¬xed leg of the swap, denoted below as V XL (0, N ) can be


5
To keep things simple at this point, it is assumed that the bond has no recovery
value.
60 5. Asset Swaps

written as
N
D— (0, i)δi C
¯
XL
V (0, N ) = P (5.3)
i=1

where, as is customary for an asset swap, we assumed that the ¬xed leg has
the same coupon, notional principal, and payment dates as the underlying
bond. Before moving ahead, let us take notice of two points. First, we
omitted the principal payment at the end of the swap, given that there
is no exchange of notional amounts in these contracts. Second, and more
important, we used a di¬erent set of factors, D— (0, i), to discount the
future payments of the swap.6 These should re¬‚ect the credit quality of
the swap counterparties, either on a stand-alone basis or enhanced via
collateral agreements and other mechanisms. For simplicity, let us assume
that D— (0, i) is derived from the LIBOR curve, which tends to correspond
to the average credit quality of the main participants in the asset and
interest rate swap markets.
Turning to the other leg of the interest rate swap, the dealer has agreed
to make variable-rate payments to the asset swap buyer. This is analogous
to selling a ¬‚oating-rate note to the asset swap buyer. From Chapter 4,
equation (4.5), we know how to value a ¬‚oater. Let V LL (0, N ) represent
the market value of the ¬‚oating-leg of the interest rate swap. Then, we can
write
N
(D— (0, i)δi (F — (0, i ’ 1, i) + A)) P
LL
V (0, N ) = (5.4)
i=1

where A is the spread to be paid over LIBOR, which in the current context
shall be called the asset swap spread, and F — (0, i ’ 1, i) is forward LIBOR,
as seen from time 0, for a deposit to be made at time i ’ 1 with maturity

<<

. 2
( 11)



>>