Credit Derivatives

Credit Derivatives Copyright 2004 Deutsche Bank@

Credit DerivativesMike Pawley [email protected] Profitability Product Complexity High Medium-High

Global Markets Training

1


Credit Derivatives

What will you get out of this material ? Well, at the end of this session you should : Understand the mechanics of credit default swaps Understand the pricing of credit default swaps Know the key users of default swaps Be aware of the links between credit derivatives and synthetic CDOs Be able to explain constant maturity credit default swaps Be able to explain equity default swaps Be able to identify bond/default swap relative value opportunities

What you already need to know: Basic bond maths Simple repo applications.


2


Credit Derivatives This session covers the main uses of credit default swaps, the key product in the credit derivative market. In particular, we look at the applications of credit default swaps in customised structured products : credit linked notes, repackaged notes, and synthetic collateralised debt obligations. In the appendix we have briefly covered other types of credit derivative : total return swaps and credit spread options. We start with an introduction to the basic credit default swap product and its uses. Credit derivatives are bilateral contracts tied to the performance of an underlying reference credit(s) or loan(s). They allow users to manage their credit exposure via the isolation and transfer of credit risk. The figure below shows the structure of a credit default swap. It is a contract whereby one party (the protection seller) agrees to receive periodic payments (the CDS spread or fee) in return for making a contingent payment to the protection buyer, following a default on a specific security or loan. Credit Default Swap StructureContingent Payment

Protection Buyer

Protection SellerPeriodic Fee (bps), s.a. A/360

Reference Entity/Security (Bond/Loan)

The credit risk faced by the protection buyer as a result of holding the reference security has been stripped off, and effectively trades separately. So, what are the motives of the counterparties ? Perhaps the protection seller is a bank that wants credit exposure to a particular sector to which it does not have a have particularly strong presence. Alternatively, it could be an investor seeking access to a particular credit that is not normally available in the market. The protection buyer may simply want to hedge the credit risk on the bond for a period of time.Global Markets Training3


If a credit event occurs the protection seller makes a payment to the protection buyer. Payment is usually linked to the change in price of an agreed reference credit in order to cover the loss. The most common method of payment is physical settlement, in which the protection buyer exchanges the reference security for par. Settlement may also be in cash, and is usually the difference between par and the market value of the defaulted reference asset. The market value may be determined by a calculation agent or through a dealer poll. A credit default event would normally cover payment default, bankruptcy and rescheduling/restructuring of debt. It must be a material and objectively measurable default. In particular, note that ratings downgrades are not in themselves evidence of default. Users Banks tend to be major users of CDSs. They tend to use them to reduce credit line usage with major counterparties. Increasingly, clients give mandates for high valueadded investment banking business (derivatives, FX etc) to those banks that are willing and ready to offer deep credit lines. If credit lines are full, the CDS allows the bank to free up line usage to carry on lending to the valued client. The alternative would be to refuse to lend and subsequently to lose the chance to cross-sell other products. Banks also use CDSs to adjust portfolio concentration. Many banks tend not to be fully diversified - their portfolio may be geographically concentrated or concentrated on a particular sector. Credit default swaps therefore allow banks to manage assets and shift credit risk. One of the major advantages of credit default swaps is the ability to create maturity and credit exposures that are not available in the underlying cash market. For example, if an investor wants a four year maturity and duration to an issuer that only has two year and seven year securities trading in the secondary market, then a four year credit default swap can be created that suits the investors requirements. Alternatively, a credit linked note (see below) could be created for the investor. Its important to recognise that credit derivatives are not one-shot static products. They are dynamic products whose value changes as the risk of default on the underlying changes. Their mark-to-market performance is therefore very closely related to changes in credit spreads. They are therefore a very useful tool for hedging/speculating on changes in credit. Clearly, any negative sentiment on a corporate, or perhaps a negative outlook from the rating agencies will tend to increase the credit protection bought in the market and a subsequent increase in the default swap fee quoted for new business (see below). The mark-to-market of existing default swaps will therefore change (the MTM value of the CDS will go up in this case).


4


Default Swaps As Dynamic Products

Contingent Payment CDS Market Increased Fee required by CDS market for new default swaps on XYZ

Investors

Corporate XYZ (poor outlook)

For example, an investor could have sold protection on Ford in October 2002 at around 650bps and closed out the deal in January 2003 at 400bps, making 250bps (hindsight is a wonderful thing see figure below). This is effectively the same as buying Ford bonds (but without the duration risk). Alternatively, an investor could have bought protection on Ford around June 2002 at 200 bps and closed it out later in October 2002 at 650bps, a gain of 450bps. Buying protection is effectively the same as shorting the credit, but with much less hassle. Ford 5Y CDS Spread HistoryFord Motor Credit700

600

500

400 Ford Motor Credit 5Y CDS 300

200

100

0 04/01/2002 04/05/2003 04/11/2003 04/05/2004 04/09/2003 04/09/2001 04/07/2002 04/01/2003 04/03/2004 04/07/2004 04/11/2001 04/03/2002 04/05/2002 04/09/2002 04/11/2002 04/07/2003 04/01/2004 04/03/2003

The market can also be used for relative value trades. Relative value trades expressing intra- or inter- industry views are more efficiently expressed using default swaps. They can be used to isolate a credit view whilst limiting curve, rate and funding risk. They are easier to execute as there are less components/instruments in the trade.Global Markets Training5


Here, for example, is an intra-sector trade idea : Ford/GM basis. When the Ford/GM spreads diverge too much, sell Ford protection and buy GM protection in the expectation that spreads will come back together. Alternatively, when the spreads are too close to each other, buy Ford protection and sell GM protection. Ford and GM 5Y CDS Spread History.Ford Vs GM350

300

250

200 Ford 5Y GM 5Y 150

100

50

008 /0 9/ 22 200 /0 3 9/ 06 200 /1 3 0/ 20 200 /1 3 0/ 03 200 /1 3 1/ 17 200 /1 3 1/ 01 200 /1 3 2/ 15 200 /1 3 2/ 29 200 /1 3 2/ 12 200 /0 3 1/ 26 200 /0 4 1/ 09 200 /0 4 2/ 23 200 /0 4 2/ 08 200 /0 4 3/ 22 200 /0 4 3/ 05 200 /0 4 4/ 19 200 /0 4 4/ 03 200 /0 4 5/ 17 200 /0 4 5/ 31 200 /0 4 5/ 14 200 /0 4 6/ 28 200 /0 4 6/ 12 200 /0 4 7/ 26 200 /0 4 7/ 09 200 /0 4 8/ 23 200 /0 4 8/ 20 04

So, the market has rapidly moved from using CDS as a static hedging instrument to one that uses CDS to actively buy and sell credit risk.


6


CDS Pricing Credit default swap pricing tends to be driven by credit spreads in the cash markets and by the financing costs of hedging the deal. The key parameters are the asset swap rate and the repo rate. If buying protection, we have the following figure : Credit Default Swap Pricing : Buying Protection and Hedging, No DefaultAsset Swap Counterparty

Contingent Payment

Fixed

Float (Libor + Asset Swap Spread) Cash

Protection SellerPeriodic Fee (28 bps)

Deutsche Bank

Repo interest Libor - Spread

Repo Counterparty

Eurobond Fixed

American Express Eurobond

Deutsche Bank buys an American Express Eurobond and finances it in the repo market. The bank then buys credit default protection (via a CDS) and swaps the interest flows on the bond from fixed to floating. If there is no default, the two fixed rates cancel, the two libors cancel and the net position is : Net Position = Asset Swap Spread + Repo Spread Fee Cash Flows For Deutsche Bank (No Default) CDS : Interest Rate Swap : Repo : Default Swap : Net Position : Fixed Libor + Asset Swap Spread Fixed -Libor-Repo Spread -Fee Asset Swap Spread + Repo Spread - Fee

If we assume that the bank funds at libor (rather than Libor-spread) then the breakeven CDS fee is the asset swap spread. In reality, of course, banks do not fund at libor but somewhat below libor (ie the repo is libor minus a spread).Global Markets Training7


If there is a default, we have the following figure : Buying Protection and Hedging, Default OccursAsset Swap Counterparty

PV of Interest Rate SwapBond

Protection SellerPar

Deutsche BankPar

Repo Counterparty


The protection seller pays DB par, DB delivers the bond, pays par in the repo and the swap is closed at a mark-to-market loss or gain. The net position is thus the PV of the interest rate swap. Cash Flows For Deutsche Bank (Default Occurs) CDS Interest Rate Swap Repo Net Position Bond Delivered at Par PV of Interest Rate Swap -Par PV of Interest Rate Swap

If we assume that the bank funds at libor and the expected present value of the interest rate swap is zero, then the breakeven CDS fee is the asset swap spread. We will use the data in the table below to price up a CDS on American Express.


8


American Express Pricing InformationUSD Libor Bid Protection Offer Protection 5% 5% Asset Swap Spread 17 bps 28 bps Default Swap Premium 28 bps 42 bps Breakeven Funding Rate 4.89% 4.86% Repo Rate 4.8% 4.75% Repo Spread - 20 bps - 25 bps

Source : DB Global Markets Research Credit Derivatives and Structured Credit August 30th 2000

In the figure below Deutsche Bank buys an American Express Eurobond and finances it in the repo market. The bank then buys credit default protection (via a CDS) and swaps the interest flows on the bond from fixed to floating. Deutsche Bank finances the purchase of the bond at a cost of 4.8% (Libor20), receives 5.17% (Libor+17) in the asset swap market (difference = 37 basis points) and buys credit protection for 28 basis points. There is therefore 9 bps of positive carry for the bank. Alternatively, we can use the expression above : Net Position = Asset Swap Spread + Repo Spread - Fee = 17 + 20 28 = 9 (positive carry) The breakeven financing cost is just the difference between the asset swap yield (5.17%) and the default swap premium (28 bps), which gives 4.89%. As the repo rate is below this, the bank earns positive carry (as mentioned above).


9


Credit Default Swap Pricing : DB Buys Protection and Hedges, No DefaultAsset Swap Counterparty

Contingent Payment

Fixed

Float (L+17= 5.17%) Cash

Protection SellerPeriodic Fee (28 bps)

Deutsche Bank

Repo interest L-20=4.80%

Repo Counterparty

Eurobond Fixed


If bond spreads tighten then the hedge will cost more for the dealer, and the dealer will push down the amount paid for protection. Similarly, if repo rates rise the hedge will cost more and the amount the dealer is willing to pay for protection will fall. Finally, if the asset swap spread tightens the hedge will also cost more and the fee quoted on default swaps will go down. All of these factors will drive down the credit default swap fee (see table below). Impact On Pricing Event Credit Spread Tightens : Repo Spread Falls : Asset Swap Tightens : Credit Spread Widens : Repo Rate Falls : Asset Swap Widens : Default Swap Fee Down Down Down Up Up Up

Alternatively, Deutsche Bank could sell credit protection. In the figure below, Deutsche Bank sells the Eurobond via the repo market, sells credit protection on the bond and receives fixed in the asset swap. [NB remember that in a repo the original seller of the bond receives the coupon back from the buyer of the bond].


10


Credit Default Swap Pricing : DB Sells Protection and Hedges, No DefaultAsset Swap Counterparty

Contingent Payment

Fixed

Float (L+Asset Swap Spread) Cash Proceeds

CDS Protection BuyerPeriodic Fee (42 bps)

Deutsche Bank

Repo Interest (L-Repo Spread)

Reverse Repo Counterparty

Fixed Eurobond Sale by DB

Eurobond

Counterparty

The two fixed rates cancel each other out, as do the libors, and the bank is left with the following position : Net Position = Fee Asset Swap Spread - Repo Spread Cash Flows For Deutsche Bank (No Default) Default Swap : Interest Rate Swap : Reverse Repo : Net Position : Fee Fixed Libor Asset Swap spread Libor Repo spread - Fixed Fee Asset Swap Spread - Repo Spread

If we assume that the bank funds at libor then the breakeven CDS fee is the asset swap spread. Now, lets look at this again using the American Express example. On the credit protection offer side Deutsche Bank sells American express (having reverse repod it in at a cost of 4.75% or L25), pays out libor plus 28 bps on the asset swap (5.28%) and receives 42 basis points for selling credit protection. Alternatively, we could calculate it as :


11


Net Position = Fee Asset Swap Spread - Repo Spread = 42 28 - 25 = (11) negative carry The breakeven financing rate is simply the difference between the asset swap level (5.28%) and the credit default swap fee (42 bps) which gives 4.86%. Note the repo rate is below the breakeven rate, so the dealer earns a negative carry. Credit Default Swap Pricing : DB Sells Credit Protection, No DefaultAsset Swap Counterparty

Contingent Payment

Fixed

Float (L+28 = 5.28%) Cash Proceeds

CDS Protection BuyerPeriodic Fee (42 bps)

Deutsche Bank

Repo Interest L-25=4.75%

Repo Counterparty

Fixed Eurobond Sale by DB

Eurobond

Counterparty

If bond spreads tighten the hedge will cost less for the dealer. The dealer will be willing to accept a lower default swap fee.


12


Credit Default Swap Pricing Using Swap Methodology It is also possible to use swap market techniques to price credit default swaps. A good analogy is with interest rate swap pricing. In practice, CDS spreads are known from the market and are used to derive default probabilities. Pricing credit derivatives thus follows the principles of no-arbitrage pricing and dynamic hedging. Therefore, the pricing benchmarks are prices of traded instruments and not historical data. Market CDS spreads reflect survival/default probabilities and the potential recovery value of the bond in default. Typically, the default curve and probabilities of default are bootstrapped from CDS market quotes for a number of maturities (using an assumption about recovery rates). Calculating Default Probabilities from CDSCDS Maturity Spreads 1 0.52% 2 0.97% 3 1.38% 4 1.70% 5 1.90% 6 1.96% 7 2.03% 8 2.07% 9 2.11% 10 2.16% Pvf Recovery (R) 0.985512959 40% 0.959923999 40% 0.92368145 40% 0.882572087 40% 0.839314814 40% 0.793244796 40% 0.753861716 40% 0.714914121 40% 0.676963455 40% 0.64128122 40% PV Fixed 0.508% 1.849% 3.820% 6.041% 8.107% 9.634% 11.177% 12.511% 13.782% 15.061% PV Probability Float of Default (P) 0.508% 0.8592% 1.849% 2.3487% 3.820% 3.6723% 6.041% 4.4993% 8.107% 4.6056% 9.634% 3.7764% 11.177% 4.1727% 12.511% 3.9685% 13.782% 4.1618% 15.061% 4.6109% 1-P 99.1408% 97.6513% 96.328% 95.501% 95.394% 96.224% 95.827% 96.032% 95.838% 95.389% PV Fixed 1-R PV Float 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% 60% 0.00% Cumulative P 0.8592% 3.1877% 6.7430% 10.9389% 15.0407% 18.2491% 21.6603% 24.7692% 27.9001% 31.2246% Survival Probability 99.1408% 96.8123% 93.2570% 89.0611% 84.9593% 81.7509% 78.3397% 75.2308% 72.0999% 68.7754%

The table above shows how default probabilities can be calculated from market CDS spreads (in this case, the Ford CDS curve). Think of CDS pricing like swap pricing there are two legs, fixed (the spread) and floating (the contingent payment). For a swap to be fairly priced at inception the present value (using PV factors from the USD swap curve) of the two legs should be equal : PV Floating (contingent payment) Leg = 60% x [(Prob of Default1 x PVF1) + (Survival Prob1 x Prob of Default2 x PVf2) + (Survival Prob2 x Prob of Default3 x PVf3) +.] PV Fixed Leg = CDS Spread x [ (Survival Prob1 x PVf1) + (Survival Prob2 x Pvf2) +] PV Fixed Leg = PV Float So, solve for probability of default given that the two legs must be equal Now, we know the CDS spread from the market, and assuming a particular recovery rate we can solve for the probability of default given that the two legs must be equal. In the excel spreadsheet we use the solver function to find the probability of default.


13


Go to spreadsheet Credit Derivatives Default Probs to see the calculations.


14


CDS Basis Trades It is potentially profitable for bond investors to pay attention to what is known in the credit derivative market as basis. This refers to the pickup that can be achieved by selling protection via a credit default swap rather than buying the cash bonds on an asset swapped basis. More formally, the basis is shown below : Basis = Protection Bid Bond Offer The choice is fairly simple : the investor should choose to take on credit risk in whichever way is most efficient (see below).

Taking on Credit Risk : Choices

Sell Protection (Take on credit risk, bid side)

OR

Buy Bonds (Take on credit risk, offer side)

Works when basis is wide (positive) Basis = Protection Bid Bond Offer The table below shows some recent examples of the basis or protection pick-up. Protection Pickup For Various Credits Protection Daimler Chrysler Rolls Royce Allianz 69 44 21 Cash Bond Asset Swapped 35 12 0 Protection Pickup 34 32 21

Five year protection for Daimler Chrysler has a positive basis of 34 basis points. The default swap could also be done in funded form as a credit linked note, although this would reduce the spread by a few basis points. Switch Trade Alternatively, if an investor already owns the bond a switch trade could be done by selling the bond and replacing it by selling protection (see figure below).Global Markets Training15


Taking on Credit Risk : Switch

Sell Protection (Take on credit risk, Bid Side)

To Replace

Bond (Sell Bond already owned, Bid Side)

Works when basis is wide (positive) Basis = Protection Bid Bond Bid Selling a bond and replacing it by selling protection will result in a smaller pick-up due to the bid-ask spread. In the CIT case, the basis would be the bid side of the protection level (95 bps) and the bid side of the bond spread (82 bps), giving a pickup of 13 basis points. If the basis is negative ie asset swapped cash bonds trade wider than default swaps, then it makes sense to buy bonds and also buy protection. In other words, it is possible to both buy the bond and buy protection and lock in a positive spread (see below). Buying Bond And Hedging Credit Risk

Buy Protection (Hedge credit risk, offer side)

AND

Buy Bond (Take on credit risk, offer side)

Works when basis is negative Basis = Protection Offer Bond Offer Normally, the protection market trades wider than the asset swapped level of the underlying cash bond. As a result, buying protection as a hedge of a bond position usually results in negative carry after taking the bid/ask spreads into account ie the cost of protection exceeds the carry on the bond that is available to pay the protection fee.


16


The table below shows examples where it would be worthwhile buying bonds and buying protection. Negative Basis For Various Credits Protection Offer HVB Group 31 Michelin 34 ING 15 Cash Bond Asset Swapped 52 53 33 Basis -21 -19 -18

Market Dynamics : The Loan Market And Credit Default Swaps Feedback mechanisms between the loan markets and the credit derivative markets can often be seen. In the figure below a strong bid for protection emerges from banks soon after signing off on a large syndicated loan. The increased demand to buy protection on the part of the banks means that credit default swap spreads will widen. So, the basis will widen to accommodate the new demand for protection. Syndicated Loans And Impact On Default Swap FeeContingent Payment CDS Market Fee Large Syndicated Loan

Banks

Corporate

This means that potential investors in the corporates bonds may be able to buy exposure at a discount to the cash bond market by selling protection instead (as mentioned above). A good example of this is the AT&T $25billion syndicated loan facility in November 2000. This led to heightened demand amongst banks for one year protection, causing the default swap curve to flatten versus cash. This led to attractive levels for investors wanting to take on credit risk by selling protection either via a CDS or a credit linked note (see below).


17


AT&T Default Swaps Vs Cash Bonds 1 Year 5 Year Cash Bond L+35 L+95 CDS 65 110

It would be expected that the AT&T default swap curve would move back in line with the cash market as banks involved in the syndication are able to acquire the protection that they need. As the credit default swap market grows, feedback mechanisms of this kind between the loan and protection markets will become more prevalent and developed. Market Dynamics : Bond Issuance And Default Swaps It may also be the case that heavy new supply of bond issuance has an impact on the basis. Market Dynamics : Negative Sentiment

Bond Market

Investors Heavy New Supply, Spreads Widen, Basis May Tighten

A good example of the relationship between cash market spreads and default spreads occurred in April 2001 in the European telecoms market. Telecom spreads widened considerably on supply fears (expectation that telecoms companies would need to issue more paper to fund 3G licenses) and also lower expectations that telecoms would be able to sell off assets to fund license purchases. Default swap spreads for telecoms also widened, but by less than in the cash market, causing the basis to reduce. The protection pick-up (difference between selling protection and buying cash bonds on an asset-swapped basis) fell by between 5-21 basis points. The protection market still remained relatively cheap to the cash market, with Deutsche Telecom at 64 bp and France telecom and KPN both over 50bp. This is partly explained by the fact that many of these issues have a high name recognition and have an unnaturally strong retail bid pushing down spreads (particularly in short maturity paper). Even CLNs on telecoms offered pick-ups of 50 basis points or more in the telecom sector in March 2001. It is also possible for the protection market sometimes to take a different view to the cash market. The protection market might take a more prudent view and push up


18


protection levels faster than cash credit spreads rise, leading to a stronger basis. (ie protection trading above the cash asset swap level). Market Dynamics : Creating Credit Derivatives From Convertible Bonds Convertible bonds, in simple terms, are made up of a bond and an equity call option. It is possible to strip out the various elements embedded in the convertible to create credit derivatives. In the figure below the investor has created an arbitrage between the value of the embedded equity option in the CB and the value of the equity option in the equity derivatives market. The investor is benefiting from a higher implied vol in the equity derivative market than in the convertible market. CB Stripping

Swap (DB)

Contingent Payment

Fixed

Float

Option Premium

Default Swap (DB) BuyerPeriodic Fee (bps)

InvestorCB Equity Option

Equity Derivatives Desk (DB)

Convertible Bond

The investor sells an equity call option to Deutsche Bank. At the same time, the investor strips out the credit risk in the CB by buying default protection from Deutsche Bank in the form of a credit default swap. Finally, the investor has hedged the interest rate risk in the CB by swapping the fixed coupon to a floating coupon. All that the investor is left with is the arbitrage profit (once taking into account the floating return from the swap, the cost of funding the CB, the fee paid to Deutsche for protection and the premium earned on the sale of the equity derivative). The only way that the trade can work for the investor is if the equity option value in the equity derivative market is large enough for the investor to cover the cost of carrying the convertible, paying to buy credit protection, and still providing enough left over to make an arbitrage profit.


19


Consider two large convertible issues from November 2000, Corning (GLW) and Tyco International (TYC) (see table below). Cash, CDS and Stripped Converts (Offer Side) Issue GLW 0% 2015 TYC 0% 2020 Rating A2/A Baa1/AMaturity Cash Bond 5 Yr L+40 3 Yr L+50 CDS 120 115 Stripped Convert L+140 L+140

Notice that the cash bond levels of GLW and TYC are around the L+40 and Libor+50 area, but the same credit can be obtained through a stripped convertible with a spread of Libor+140. The implied vol on the equity call option embedded within the GLW issue traded in the 52%-54% range around issue (November 2000), whereas equity 6-month vol in the OTC options market was around 80% at the time. These large arbitrage gains mean that hedge funds are able to pay a premium for protection if necessary (and still make money), driving up default swap spreads. The level of liquidity in the credit derivatives market will also have an impact on CDS levels. The protection market typically trades in blocks of $5-$20million, whereas these were large convertible trades ($2.7 billion and $4.05 billion respectively) that added significant demand to the CDS market. Given that dealers will want to buy protection to hedge their long positions this will add upward pressure to the CDS market. In addition, the investor base for stripped convertibles and credit linked notes is still fairly limited, making it difficult for dealers to place large long positions in GLW and TYC protection. The stripped convertible market can have an impact on credit default swap spreads. It may be the case, for example, that a large issue by a corporate on generous terms leads to hedge fund arbitrage activity. As noted before, the hedge fund sells the equity option premium and hedges the credit risk by buying a credit default swap. This can cause credit default swap spreads to widen in the market off the back of heavy demand for protection. This, in turn, can lead to cheap opportunities for investors to gain credit exposure to a corporate through the credit default swap (either by selling protection as a substitute for buying the bonds or as a switch trade ie selling the bonds and then selling protection). This is because the basis has widened. Note that the wider default swap spreads in this example result from a technical situation and are not credit related. The wider spreads are related to the increased demand to purchase protection from convertible arbitrage funds, which are interested primarily in the equity option component. Of course, it is also possible to see the opposite situation, where investor demand for convertibles is relatively high. In these situations issuers and lead managers try to capitalise on strong demand for converts by pricing issues more aggressively. The spread between the implied vol in the equity market and the CB market will then tend toGlobal Markets Training20


tighten. This can mean that the equity arbitrage for hedge funds disappears and stripping becomes unattractive. So, hedge fund asset swap bids (ie hedge funds paying fixed) tighten, creating tighter asset swap levels for fixed income investors. Tighter asset swap levels will lead to cheaper credit default swaps. Moreover, hedge funds will buy less credit protection if they are not arbitraging the convert market and protection costs will see downward pressure. For arbitrage to occur again either equity volatility has to increase or recently issued convertibles have to cheapen. See the Convertible Bond Asset Swaps session for more detail.


21


Credit Default Swaps and Synthetic CDOs Credit default swap levels can be significantly affected by activity in the CDO synthetic arbitrage market. In these deals the notional portfolio is made up of single name default swaps rather than bonds (see figure below). Investors in the different tranches of the transaction sell credit protection and receive a corresponding fee as part of their return. It is often possible to create structures where the all-in cost of buying protection in the tranched portfolio form is less than the weighted average cost of the default swaps that make up the portfolio. This can be achieved even though the spreads on the tranches are substantially wider than on comparably rated corporate bonds. In putting the portfolio together, the sponsor buys protection in tranched form from investors and then sells single name protection to dealers. This creates significant supply of protection in the market, leaving dealers long protection and looking to sell protection. This can have a downward effect on protection levels and can in fact drive the basis negative. This, of course, will tend to make it attractive to buy protection (which is not normally the case). Investors can buy bonds, buy protection and lock in a positive spread. Synthetic CDO

Single Default SwapsCDS Counterparty 1

Default Protection Default Protection Fee Default Protection Fee Sponsor Default Protection Fee Default Protection

Tranched Portfolio of Default Swaps

Class A CDS

CDS Counterparty 2

Class B CDSFeeThreshold Amount (First Loss)

CDS Counterparty 3

Fee

Note, once again, that these are technical factors impacting on the CDS market. They do not reflect changes in credit outlook.


22


Deutsche Bank Funding Spread The DB Funding Spread metric aims to provide a better measure of relative and absolute richness/cheapness of bonds versus CDS than either asset swap spreads or Z-Spreads versus CDS. The funding spread calculates a theoretical bond price, taking into account recovery rates and survival probabilities computed from the CDS curve. The theoretical bond Price is calculated like this : PV Bond Cash Flows x Survival Probability + PV Recovery Cash Flows x Default Probability The basis is then the difference between the theoretical bond price and the market price. The funding spread is then the parallel shift to the swap curve required to make the theoretical and market bond prices equal. For as positive basis (bond rich to CDS, credit spread lower than CDS spread) the swap curve needs to be shifted downwards, meaning a negative funding spread. For a negative basis (bond cheap to CDS, credit spread higher than CDS spread) the swap curve needs to be shifted upwards, meaning a positive funding spread. Trading the Basis Negative Basis Trades If the basis is negative the bond is cheap to CDS, the credit spread is higher than the CDS spread, and the funding spread is positive. The trade is then : Buy bond, buy protection, if you expect one of following : 1. Bond richens (bond price rises, credit spread falls), long bond position profitable. 2. CDS spread widens, protection resold at profitable level. 3. Both events happen, profits made. 4. Bond richens more than CDS (bond price rises, bond credit spreads fall, and by more than fall in CDS spreads), gain on long bond more than offsets CDS loss. 5. CDS spread widens more than bond cheapens (bond price falls, bond credit spreads rise, but less than rise in CDS spreads), gain on CDS more than offsets loss on bond. All of these events push the funding spread lower, ie bond richens relative to CDS. Interest rate risk is typically hedged with an asset swap, the bond financed with a repo, so you can replace Credit Spread with asset swap spread above, also corporate bonds on special are good for this trade (positive carry).


23


Trading the Basis Positive Basis Trades If the basis is positive, the bond is rich to CDS, the credit spread lower than the CDS spread and the funding spread is negative. The trade is then: Short bond, sell protection, expecting one of the following : 1. Bond cheapens (bond price falls, credit spread rises), short bond position profitable. 2. CDS spread tightens, protection bought back at profitable level 3. Both events happen, profits made. 4. Bond cheapens more than CDS (bond price falls, credit spread rises, and credit spread rises by more than CDS spreads), gain on short bond more than offsets CDS loss. 5. CDS spread tightens more than bond richens (bond price rises, credit spread falls, but less than fall in CDS spreads), gain on CDS more than offsets loss on bond. All of these events push the funding spread more positive ie the bond cheapens relative to CDS. The interest rate risk is typically hedged with an asset swap, the bond financed with a repo, so you can replace credit spread with asset swap spread above, also corporate bonds on special are bad for this trade (negative carry), although investors who own bonds can sell them, invest the proceeds at Libor and sell protection, enhancing the yield on their portfolios until the positive basis disappears. The return from basis trading comes in three parts : Carry P&L on unwind, assuming no default Payoff on default

For a face value hedge the CDS and bond notionals are set equal. For an expected recovery hedge the CDS notional N is set such that the payoff in default is zero ie N (1-R) P + R = 0. Lets concentrate on a negative basis trade. The carry is : Carry = Proceeds of bond x (ASW+Repo Spread) Notional on CDS x CDS spread


24


Carry on Negative Basis Trade

ASWProceeds of bond x Libor + ASW

Coupon

Protection SellerCDS Notional x Spread

InvestorProceeds of Bond x Libor Repo Spread

Repo

Coupon

Bond

For buyers of protection higher CDS notional leads to lower carry, all other things equal, therefore the expected recovery hedge has highest carry if the bond is below par, the face value hedge has the greatest carry for bonds above par. For the P&L on Unwind, assuming no default, and assuming the basis hits zero at the unwind date (bond becomes richer and CDS curve remains constant) : for a buyer of protection higher CDS notional leads to lower P&L (same result as for carry), all other things equal, therefore expected recovery hedge has highest P&L if bond below par, face value hedge has greatest P&L for bonds above par). P&L on Negative Basis Trade

ASWProceeds of bond x Libor + ASW

Coupon

Protection SellerCDS Notional x Spread (Nets Off)

InvestorProceeds of Bond x Libor Repo Spread

Repo

Coupon

Bond

For the payoff on default : Payoff on default = N(1-R) P + R N = CDS Notional, 1 = Bond Face Value, R = Recovery Rate, P = Initial Bond Proceeds25



Payoff on Default for Negative Basis Trade

ASW

ASW Unwind

Protection Seller

N(1 R)

P + Repo Rate

Investor

Repo

R

Defaulted Bond

As most bonds are currently above par and defaults not generally expected, face value hedges are likely to be more prevalent, see the figure below. Hedge Rankings for Negative Basis TradesScenario Full Price Below Par, Default Not Expected Full Price Below Par, Default Expected Full Price Above Par, Default Not Expected Full Price Above Par, Default Expected Best Hedge Expected Recovery Hedge Face Value Hedge Face Value Hedge Expected Recovery Hedge Justification Highest carry and P&L on unwind Highest payoff when default occurs Highest carry and P&L on unwind Lowest loss when default occurs

Source : Global Markets, Quantitative Credit Strategy, 25 August 2004 Market Value Hedges are are also possible, where CDS notional is the same as bond proceeds, but these are always outperformed by either face value or expected recovery hedges).


26


Hedge Rankings for Positive Basis TradesScenario Full Price Below Par, Default Not Expected Full Price Below Par, Default Expected Full Price Above Par, Default Not Expected Full Price Above Par, Default Expected Best Hedge Face Value Hedge Expected Recovery Hedge Expected Recovery Hedge Face Value Hedge Justification Highest carry and P&L on unwind Highest payoff when default occurs Highest carry and P&L on unwind Lowest loss when default occurs

Source : Global Markets, Quantitative Credit Strategy, 25 August 2004


27


CDS Forwards and Options It is also possible to create forward CDS curves from spot CDS curves, analogous to the way in which forward interest rates are created from yield curve data. We can then compare the forward curve for a credit with the generic forward curve for the same credit rating to see if there is a difference and if there might be trading opportunities. The forward CDS Spread can be calculated as follows : (CDS Spread2 x Risky PV012),- (CDS Spread1 x RiskyPV011) Risky PV012 Risky PV011 Where : Risky PVf : Product of Riskless PVf and Survival Probability Risky PV01 : Cumulative PVf The table below shows the forwards calculated from the formula above. Forward CDS SpreadsMaturity 0 1 2 3 4 5 6 7 8 9 10 USD Swap Rates 1.47% 2.06% 2.66% 3.13% 3.50% 3.84% 4.01% 4.16% 4.29% 4.39% PVf 1 0.985513 0.959924 0.923681 0.882572 0.839315 0.793245 0.753862 0.714914 0.676963 0.641281 Survival Risky Probability PVf 99.14% 96.81% 93.26% 89.06% 84.96% 81.75% 78.34% 75.23% 72.10% 68.78% 0.9770 0.9293 0.8614 0.7860 0.7131 0.6485 0.5906 0.5378 0.4881 0.4410 Risky CDS Forward 1Y Forward 2Y Forward 3Y Forward 4Y Forward 5Y Forward 6Y Forward 7Y Forward 8Y Forward 9Y PV01 Curve Maturities Maturities Maturities Maturities Maturities Maturities Maturities Maturities Maturities 0.9770 1.9064 2.7678 3.5538 4.2669 4.9154 5.5059 6.0438 6.5319 6.9729 52 97 138 170 190 196 203 207 211 216

144 229 283 290 235 261 248 261 290

185 254 286 264 248 255 254 275

215 265 271 263 248 257 265

231 259 269 260 251 264

232 259 265 260 257

236 258 265 264

237 258 267

239 261

243

Go to spreadsheet Credit Derivatives Forward CDS to see the calculations.

The question the trader has to ask is this : Do I believe the evolution of the forward curve ? A pure forward trade would be to buy protection at one maturity and sell protection at another using equal notionals. For example, buy 3Y protection, sell 5Y protection synthetically creates selling two year protection starting in three years. For years 1-3 receive difference between 5Y and 3Y CDS spreads (carry). For years 4-5 pay 5Y CDS, but expect to unwind before then. Outright forwards are possible in some markets, but there are often liquidity problems. Also, there can be wide bid-offers, around 5bps for 5Y CDS, and 10bps or more for other maturities, meaning that significant curve changes would be necessary for the trade to make money.Global Markets Training28


It is possible to trade the shape of the curve, as opposed to a specific forward rate. If the curve is too steep then buy three year protection and sell five year protection. The notionals of each leg should be weighted such that the trade is immunised against parallel shifts in the credit curve. For example, the three year leg should be weighted by the relative risky PV01s of the three and five year legs, 4.2576/2.7668 = 1.54 (these data come from the forward table explained in more detail below). This gives PV01 neutral two year short protection stating in three years. Note that the forward CDS spread depends on the level and slope of the CDS curve. A steeper and/or higher spot CDS spread will result in a higher forward CDS spread. Credit Spread Options A Payer Swaption is the right to buy protection on an index or credit. It can be thought of as a put on the forward credit price or a call on the forward CDS spread. Other features are : European exercise, premium is paid/received upfront, maturities tend to be 3-6 months, strikes are often at-the-money although others are possible, physical delivery upon exercise. Upon exercise parties enter into a standard default swap on index or single name. Index options allow for exercise if credit events occur before expiry, whilst single name options knock out if a credit event occurs prior to expiry. Payer Swaption

payout

Forward CDS Spread

A Receiver option is the right to sell protection on an index or credit. It can be thought of as a call on the forward credit price or a put on the forward CDS spread. Receiver Swaption

payout

Forward CDS Spread


29


The market for trading volatility on credit spreads has been developing rapidly, with markets now existing in trading options on credit indices, and a focus on options to express volatility views as well as directional credit bets. So, heres a simple model to price CDS swaptions. It is basically the Black model for pricing options on forward prices. Here is the VBA code for pricing it :Function Swaption(ForwardSwap, Strike, Volatility, PVfSum, ValueDate, ExpiryDate, PayOrReceive) Dim Time As Single Dim D1 As Single Dim D2 As Single Dim OptVal As Single Time = (ExpiryDate - ValueDate) / 365 D1 = (Application.Ln(ForwardSwap / Strike) + (Volatility ^ 2 * Time / 2)) / (Volatility * Sqr(Time)) D2 = D1 - Volatility * Sqr(Time) OptVal = PayOrReceive * ForwardSwap * Application.NormSDist(PayOrReceive * D1) - PayOrReceive * Strike * Application.NormSDist(PayOrReceive * D2) Swaption = OptVal * PVfSum End Function

Here we price up 1 year into 5 year payers and receivers swaption (ie the right in one years time to enter into a five year CDS) on Ford. These knockout if there is a credit event before expiry of the option. So, payers and receivers swaptions would be 2.67% (note they are the same price because the strike is at-the-money).Forward CDS Spread Strike Volatility PV01 Value Date Option Expiry Date Option to Buy Protection : Option to Sell Protection : Payer Swaption Receiver Swaption 231.67 232 75% 3.9443214 28/09/2004 28/09/2005 267 267

Go to spreadsheet Credit Derivatives CDS Swaptions to see the calculations.


30


In theory, a knockout payer swaption should be delta hedged with a short protection CDS to the final maturity of the underlying CDS and a long protection CDS to the default swaption expiry date. This produces a synthetic forward CDS that knocks out at default before the expiry date. However, CDS with short maturities tend to still have low liquidity, so swaptions with expiry of one year or less (the majority) are hedged only with CDS to final maturity (as mentioned earlier, there are also liquidity issues with outright forwards). The market for trading volatility on credit spreads has been developing rapidly, with active markets in trading of options on credit indices and a focus on options to express volatility views (see the straddles in the figure below) as well as directional credit bets. The liquidity of the credit volatility market has benefited from several factors : increased liquidity in the underlying product, i.e. single name CDS and index products; better understanding of credit spread volatility and modelling approach; the increased demand for credit options as part of an integrated credit trading strategy.

Straddle on Index, Long Vol

Straddle on Index, Short Vol

Buy Receiver

Buy Payer

Sell Receiver

Sell Payer

Constant Maturity Credit Default Swaps (CMCDS) The premium in the first year is X% of the 5 year spot CDS spread, subsequent premiums are reset at a pre-determined frequency (quarterly, semi-annually), to X% of the then spot 5 year spread. X% is known as the participation rate and stays constant. Selling CMCDS protection provides spread income plus the upside to spread widening/curve steepening (whilst still taking default risk ie long credit) the best analogy is fixed bonds versus FRNs, CDS versus CMCDS. CMCDS should be used when spreads are tight, there are flat curves and vol is low ie Now ! CMCDS generally provide higher spread income than short-dated risk and lower MTM volatility compared to a plain vanilla CDS. Note that the maturity of the contract and the maturity of the CMCDS reference rate need not be the same.Global Markets Training31


CMCDS Spread Setting ProcessConstant 5 Year CDS maturity set every periodbps

18 months forward 12 months forward 6 months forward Spot CDS Curve

1

5

Maturity (years)

Consider the example of Ford Motor Company (F):Investor View

Is it likely that F will default within the next 5 years? Is it likely that F CDS spreads tighten further relative to historical levels? Is it likely that F CDS spreads will widen within the next year or so?

NO NO YES

The investor can sell CMCDS protection on F at 74% of the 5yr spread = 230bps * 0.74 = 171bps. By selling CMCDS protection the investor benefits from potential future spread widening through spread reset.


32


Ford CDS Curve Vs ForwardsCDS Curve Vs Forwards300

250

200

150

CDS Curve Constant Maturity 5Y

bps

100

50

0 1 2 3 4 5 6 Years 7 8 9 10 11

Pricing a CMCDS means finding the fair participation rate. The payoff on the loss leg of a CMCDS is identical to the payoff on the loss leg of a plain vanilla CDS, so the spread legs on both instruments must be equal too : CMCDS Spread Leg x Participation rate = CDS Fixed leg Lets assume the CMCDS spread leg can be priced using the forwards, then the participation rate becomes : Participation Rate = CDS Spread Average of Forward CDS Spreads

If the CDS curve is flat, the forwards equal spot and the participation rate will equal 1. If the CDS curve is upward sloping, the forwards will be higher than spot and the participation rate will be lower than 1. If the CDS curve is downward sloping, the forwards will be lower than spot and the participation rate will be higher than 1.CDS Curve 52 97 138 170 190 196 203 207 211 216 Forward 5Y CDS Spread 1 x 5 CMCDS Participation 2 x 5 CMCDS Participation 3 x 5 CMCDS Participation 4 x 5 CMCDS Participation 5 x 5 CMCDS Participation

Maturity 0 1 2 3 4 5 6 7 8 9 10

0.27 0.46 0.61 0.72 0.79 190 232 1st Year 1st Year 1st Year 1st Year 1st Year 259 265 Spread Earned Spread Earned Spread Earned Spread Earned Spread Earned 52 87 116 137 150 260


33


0.7% 0.6% 0.5% 0.4% 0.3% 0.2% 0.1% 0

Go to spreadsheet Credit Derivatives CMCDS to see the calculations.

The participation in the CMCDS depends on the shape of the curve. Steeper curves typically have lower participation levels whilst inverted curves typically imply a participation greater than 100%. For example, the figures below show the CDS curves for Lafarge and Corus and their participation levels. Lafarge Credit Curve (5y Lafarge CMCDS: 60%)

2

4

6

8

10

12

14

Corus Credit Curve (5y Corus CMCDS: 102%)8.1% 7.1% 6.1% 5.1% 4.1% 3.1% 2.1% 1.1% 0.1% 0 2 4 6 8 10 12 14

For views solely on credit spreads widening and/or steepening the trade would be to sell CMCDS protection, buy CDS protection, which gets rid of default risk. The CMCDS technology can be applied to the entire continuum of credit derivative products, single name credits, capped CMCDS, first-to-default baskets, credit indices, credit portfolios.


34


Credit Indices The Dow Jones Indices have become the de facto indicators for the synthetic credit default swap market. The DJ indices represent the merger of two leading indices in June 2004 (iBoxx and DJ TRAC-X). Here are (some) of the indices available : CDX.NA.IG 125 US names iTraxx Europe 125 names iTraxx Asia (ex Japan) 30 names (plus sub-indices for Korea, Greater China and other countries) The Dow Jones iTraxx Indices have become the de facto indictor for the synthetic credit default swap market. Represents the merger of two leading indices in June 2004 (iBoxx and DJ TRAC-X). iTraxx Japan 50 names iTraxx Australia 25 names The figure below shows the main users of the indices.Whos Using the Indices? Asset Managers Quick credit exposure Liquidity Management tool

Hedge Funds Relative value trades Efficient tool for directional trading

Bank Prop Desks Relative value trades Efficient tool for directional trading

Bank Portfolio Managers Suitable for portfolio balancing Credit diversification tool

Correlation Trading Desks Suitable for portfolio hedging Easy ramp-up

Corporate Treasury Easy access to diversified US risk

Insurance Companies Proxy hedge against senior CDO credit portfolio

For investors that do not have the credit resources to fully research a large number of names one approach would be to sell protection on one of the broad indices and concentrate credit analysis on sectors or names that might offer the prospect of superior returns, such as the high yield sector. The DJ indices allow investors to trade sector exposure in large size. This could include strategies where single name risk is hedged out to create specific baskets. For example, an investor could sell protection on the HVOL index and buy protection on names that are deemed to be above a particular level of risk (or those names marked down as underweight by DB research. In addition, investors could sell extra protection on names which are expected to outperform (eg those which DB research has placed on an overweight recommendation) in order to generate extra spread income. Some traders take views on the indices according to the difference between the fair value spread and the actual spread on the index (known as the index basis). The fair value spread is that implied by the spread on the underlying names (calculate the NPV for each CDS in the index, add them up and reverse out one spread that equates to the aggregate NPV DB research also makes an adjustment for the no restructuring clause of an index by taking 5% off the spread, and adjusting for maturity mismatch of around 1bp per quarter).Global Markets Training35


The basis between the index and the underlying names is taken as an indicator of relative richness/cheapness of the index. If the basis (index spread fair value spread) is positive the index is cheap to the underlying names. A negative basis means that the index is rich relative to the underlying names. The basis itself is rarely traded because of the difficulty of trading 125 names given the liquidity and transactions costs. Generally a directional position is simply taken in one of the indices. First-to-Default Basket Default Swaps Basket default swaps can be used to buy and sell protection on baskets of securities or portfolios of assets, where default protection is traded on each of the constituent securities or loans in the portfolio. In the figure below Deutsche Bank makes periodic payments for default protection on a basket of four securities and receives a contingent payment when any reference credit in the basket experiences a credit event, at which point the swap terminates. First-to-Default Basket Default Swap Structure

Security 1

Contingent payment Seller Deutsche Bank (Buyer) Fee

Security 2

Security 3

Security 4

The seller gains higher payments than on a single security of the same credit quality but the sellers risk is limited to the market loss that follows a credit event on a single security in the basket. First-to-Default baskets are considered in more detail in the CDO materials as they are basically correlation products. The figure below shows how first, second and third to default baskets are priced in relation to varying correlation between the names in the baskets. Correlation Versus Pricing for Default BasketsNB implicit ceiling on price of baskets : dealers unwilling to pay more than about 85% of the sum of spreads)

100% Basket Spread as % of Sum of Spreads

First-toDefault

Second-to-Default

0% 0% Global Markets Training Correlation

Third-toDefault

36

100%


A good way of thinking about FTD baskets and correlation is to compare them with the various tranches on a CDO. The equity is similar to first-to-default, mezzanine like second to default, and senior like third to default (in simple, relative terms). You can immediately see that the figure above is very similar to the figure in the right hand corner below. With low correlation the assets are more or less independent and there is a low probability of large losses, low probability of zero losses spread on senior tranche (third, fourth, fifth to default etc) is small, spread on equity tranche (first-to-default) is large. For medium correlation the assets are more likely to default together and the distribution of the tail is longer more risk in senior tranche, larger senior spread (third,fourth, fifth to default etc) than before; probability of zero losses has increased, equity less risky and equity spread (first-to-default) smaller than before. Finally, for high correlation the portfolio behaves like one asset with zero diversification, the names either all default or all survive. The probability of a large number of losses is significant, the spread required to hold the senior tranches (third, fourth, fifth to default etc) is relatively large; probability of zero losses is very high so equity spread (first-todefault) is small.Probability Probability Probability 20% 18% Probability Probability 16% 14% 12% 10% 8% 6% 4% 2% 0% 0% 4% 8% 12% 16% 20% 24% 28% 32% 36% 40% 44% 48% Portfolio Loss (%) 5.00% 0.00% 0% 4% 8% 12% 16% 20% 24% 28% 32% 36% 40% 44% 48% Portfolio Loss (%) 15.00% 10.00%

Low correlation, high equity returns, low senior returns

35.00% 30.00% 25.00% 20.00%

Medium correlation lower equity returns, higher senior returns

High correlation, lowest equity returns, highest senior returns80.00% 70.00% Probability 60.00% Spreads (%) 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% 0% 4% 8% 12% 16% 20% 24% 28% 32% 36% 40% 44% 48% Portfolio Loss (%)

Correlation and Tranche Spreads25.00%

20.00%

15.00%

Equity Mezzanine Senior

10.00%

5.00%

0.00% 10% 20% 30% 40% 50% 60% 70% 80% Default Correlation


37


Implied correlations for standardised baskets as of September 2004 are given in the table below for comparative purposes.Basket Consumer Industrial Financial Energy TMT HVOL Correlation (1/09/04) 28% 24% 51% 38% 34% 23%

Standardised FTD baskets are a new development to the market (summer 2004). Sixteen dealers have undertaken to make markets in the standardised US baskets listed above. Each basket has five corporate names in it, and average bid-offers at DB are 25bp (a lot tighter than before the standardisation came about). For credit views investors can buy/sell straight FTD protection (there will still be a correlation element here, but it can be said that the credit view dominates). For correlation views investors can buy/sell FTD protection and delta hedge (this eliminates default risk and concentrates on the correlation element). Here are some relationships/terminology to remember : Buy FTD and delta hedge : short correlation (expect correlation to fall) Sell FTD protection and delta hedge : long correlation (expect correlation to rise) For correlation trades the key drivers are idiosyncratic single name credit spread movements (lower correlation) and economy/industry wide risks (higher correlation). Heres a good example of correlation trading with FTD baskets. In July 2003, Sears sold its credit card unit and Sears spreads subsequently tightened on the market relief. At the same time, Altria, also in the Consumer basket of five names, suddenly faced litigation threats, leading to a big widening of Altria spreads. Spread correlation for the basket fell (see the figure below). These are idiosyncratic risks that reduce correlation. Consumer FTD basket Spread CorrelationSpread Correlation 40%

Consumer FTD Basket 30% 0% Jul03 10% 20%

Nov03

Apr04


38


Equity Default Swaps (EDS) EDS offer the opportunity to trade/invest in a credit-like instrument without the need to deal with the complexities of the ISDA credit derivatives conventions (this can be especially attractive for retail investors and equity derivatives investors) as EDS are documented under ISDA 2002 Equity Derivative definitions rather than ISDA Credit Derivative definitions. An EDS trigger even is based on the performance of the underlying share price, typically a 70% fall in the share price. An EDS has fixed recovery in the case of a trigger event, usually 50% of notional. For fixed-income investors, EDS enable access to an entirely new segment of the continuum of corporate risk, in a familiar CDS format. For equity investors, EDS provides the ability to tap into a part of the volatility surface that previously did not trade with any material liquidity In particular, EDS offer a wide range of opportunities to extract value from the equity versus credit market. In the figure below you can see that a buyer of EDS protection pays a running premium on the notional to protect against the occurrence of a trigger event. The seller of EDS protection pays 0% if there is no trigger event or [Par - Recovery] if there is a trigger event. A trigger event occurs if the stock price of the underlying reference entity ever closes below the barrier level set at the inception of the contract. Equity Default Swap MechanicsContingent Payment on Trigger Event

Buyer of EDS protection

Seller of EDS protection

Running Premium

The tables below show a comparison between a CDS on siemens and an EDS on siemens. Clearly, EDS spreads are a lot higher, reflecting the higher probability of the share price dropping by 70% compared to the probability of default.


39


Siemens Credit Default SwapReference Entity Reference Notional Scheduled Maturity Spread Trigger : : : : : Siemens AG EUR [10,000,000] 5 years 26 bp (1) Bankruptcy (2) Default (3) Restructuring Redemption : 100% (ie EUR10m) or Recovery Value

Siemens Equity Default SwapReference Entity Reference Notional Scheduled Maturity Spread Trigger Redemption : : : : : : Siemens AG EUR [10,000,000] 5 years 102 bp 70% drop in the Share Price 100% or 50% (ie EUR10m or EUR5m)

EDS are available in note format and in first-to-default or nth-to-default contracts. For some investors, simple relative-value trades are becoming popular : Sell EDS protection and Buy CDS protection For a broad universe of names, this is a massively positive carry position where the investor is proxy-hedging the EDS exposure with CDS protection. If the investor believes that credit spreads and equity prices are negatively correlated, then if equity prices go down the investor will lose money on the EDS but make some of it back through the CDS (on a MTM basis). The CDS protection can be purchased on a larger notional than the EDS protection sold, due to the large EDS-CDS basis, and still allow for a positive carry. For the Siemens EDS/CDS positions above, the positive carry would be 102bp - 26bp = 76bp. Net Payment should the equity price collapse by more than 70% would be EUR 5,000,000 - (Value of CDS) Lets assume a recovery assumption for the CDS of 30% (Trader receives EUR 7m if Siemens AG triggers a credit event).Global Markets Training40


The maximum gain: Should Siemens AG default, then the Trader would make : (EUR5 m) + EUR 7m = EUR 2m The maximum loss: Should the CDS not be worth anything the trader would lose : (EUR 5m) + EUR 0m = (EUR 5m) Credit Linked Notes This section explains how credit linked notes (CLNs) are structured and their attraction to investors. CLNs are typically issued by corporates or banks, often through a medium term note or structured note programme. A CLN is a structured note that combines a debt product and an embedded credit derivative (typically a credit default swap). In the figure below the investor in the CLN receives interest that is enhanced by the default premium received for selling credit protection. If a credit event occurs on the reference asset, the note typically matures immediately and the investor takes a loss of principal based on the loss on the reference asset. Credit Linked Note (Embedded Credit Default Swap)Contingent Payment Proceeds

Default Swap (DB) BuyerPeriodic Fee (bps)

Issuer

CLN InvestorInterest, Default Premium

Interest Rate Swap, Currency Swap etc (DB)

The issuer is simply trying to reduce its cost of funds. It does this by buying the default protection embedded in the note and then selling that protection to Deutsche Bank. The difference between the value of the credit default swap embedded in the note and the credit default swap sold to Deutsche creates the reduced cost of funding for the issuer. The notes are typically issued by highly rated corporates or banks, with an embedded default swap on another (typically lower rated) reference entity. A variety of investors simply are not allowed to enter into credit derivatives or other offbalance sheet products. CLNs are on-balance sheet instruments and therefore tend not to be caught by these restrictions. These investors can therefore gain access to the credit derivative markets via CLNs (an anomaly, certainly, but the sort of anomaly that drives many markets). CLNs, like plain credit default swaps, can also be customised such that they create maturities and credit features that may not otherwise be available in other markets.Global Markets Training41


CLNs may be rated by one of the agencies. The rating is based on the embedded default swap and the credit of the CLN issuer. Some investors are wary of buying CLNs directly because of the counterparty credit risk involved. They therefore prefer to buy credit linked notes structured by banks via bankruptcy-remote special purpose vehicles (SPVs see repackaging below). Insurance companies and retail investors tend to buy Credit Linked Notes in the form of Principal Protected CLNs. With these, there is a guaranteed full return of principal. If a credit event occurs in relation to the reference security, the CLN remains outstanding until its maturity date but stops paying interest. At maturity the CLN investor receives the principal amount. The investor is therefore partly protected from default under the reference security, but still has the risk of principal loss if the issuer of the CLN defaults. CLNs offer structural characteristics similar to cash instruments : Notes can be structured with DB as the host asset (DB CD or MTN) via a shelf program or with third-party collateral (AAA ABS) via an SPV or Trust Return of principal is contingent upon a Credit Event of the Reference Entity or the default of the host asset Notes can pay a fixed or floating coupon, quarterly or semiannually Notes are DTC/Euroclear eligible and generally 144A Notes can be rated for an additional cost


42


How to create a tailored structured note : Does client want exposure to a single name or a customized index or basket? What maturity fits the clients risk profile? What are the clients spread target/benchmark and investment guidelines? Does client want exposure to a particular industry or set of industries? Does client want exposure to a particular geographical region? Does client have a rating requirement for credits in which it invests? Does client need note rated by S&P and/or Moodys? Does client want a fixed or floating note? LIBOR or other base? USD or foreign currency exposure? How else can we customize the note to meet the needs of client e.g. Can add interest rate or FX overlay?


43


Appendix 1 : ISDA 1999 and 2003 Credit Derivative Definitions The introduction of the 1999 ISDA Credit Derivative Definitions led to a reduction in legal risk, greater standardisation of terms, faster execution and confirmation and enhanced market liquidity and transparency for credit derivative market participants. Further modifications have been made to the definitions in a number of Supplements, following specific market events (notably the Conseco restructuring and the Railtrack default). The 2003 Definitions have adopted the various Supplements and also introduced new terms and conditions that are a further step in the evolution of the credit derivative market. Notice of Physical Settlement Protection buyers must notify the protection seller of the bond/loan it will deliver to settle a triggered default swap contract. The provision also sets out explicit steps to be taken if the buyer changes the deliverable, or is unable to deliver as scheduled. The overall effect is to clarify that a protection contract is not subject to cancellation if timely delivery cannot be made. Business Day Convention If a business day convention is not specified in the confirmation, the fallback will be following instead of modified following. Under following if a date falls on a nonbusiness day, an adjustment is made to the first following business day. Under modified following an adjustment is made to the first following business day unless that day falls in the next calendar month, in which case it is adjusted to the first preceding business day. Quarterly Roll Date Default swaps will generally have standardised matuirty dates falling on 20th March, June, September and December, and the maturity date will roll over quarterly. Effective Date The effective date for a default swap is now the day following the trade date (T+1) instead of the previous T+3 convention. Restructuring Market participants can choose Full Restructuring, where a reference obligation is restructured so as to reduce the interest or principal due, delay interest payments or extend the maturity date, become subordinated to another obligation. The multiple holder obligation is standard, unless the counterparties agree to waive this provision. Under the multiple holder requirement, an obligation must be held by at least four unaffiliated holders and holders representing at least two-thirds of the principal balance must consent to trigger a credit event.


44


The intention here is to exclude restructurings in connection with bilateral loans. Note that a bilateral loan is still capable of being delivered, even if its restructuring cannot trigger a restructuring credit event. Supplementary definitions and provisions to the 1999 Definitions have been made in the Restructuring Supplement to the 1999 ISDA Credit Derivatives Definitions, subsequently adopted in the 2003 Definitions after full restructuring proved problematic. These were constructed in the light of market dissatisfaction with the situation in the Conseco case. In the Conseco situation, buyers of protection could deliver long-dated bonds that were considerably cheaper than short-dated bonds. They therefore gained significantly from the cheapest to deliver option, even though they were not materially worse off after the restructuring. In the US, dealers started to charge a premium (10% to 15%) for CDS with restructuring compared to CDS without restructuring. In Europe dealers continued to include restructuring as a standard event. In the Restructuring Supplement the guidelines for deliverable obligations following a restructuring are as follows : Restructuring Maturity Limitation (Modified Restructuring or Mod-R) If physical settlement and Restructuring Maturity Limitation Applicable are specified in the confirmation, and restructuring is the only credit event specified in a credit event notice delivered by the buyer, then the maturity of the deliverable obligation is limited to the earlier of the restructuring date plus thirty months, or the maturity of the restructured obligations, subject to a general provision that the maturity limitation cannot be earlier than the termination date of the default swap or later than thirty months following the termination date. The maximum maturity date is thus at a minimum the termination date, and may extend to the earlier of the restructuring date plus thirty months or the termination date plus thirty months. Under the 2003 Definitions, mod-r is selected by choosing Restructuring Maturity Limitation and Fully Transferable Obligation : Applicable The changes are intended to prevent delivery of long-dated obligations that might be trading at prices well below short-dated obligations. The idea is that the maturity of the deliverable obligation should be consistent with the term of the protection contract. Sellers can therefore reduce their exposure to the cheapest to deliver option that can arise when the reference obligations of an entity have many different maturities and hence many different values. Furthermore, the selected delivered obligation must be freely transferable, meaning that it is not subject to restrictions on who can hold it or to approvals from an original lender before it can be transferred. NB the restructuring definitions do not take into account statutory or regulatory restrictions such as banking regulations that prohibit certain types of entities from holding loans and/or bonds.


45


Mod-R proved to be too restrictive for many European banks, leading to the introduction of so-called Modified Modified Restructuring, or Mod Mod-r. This eases mod-R in two ways. The maximum maturity for deliverable restructured obligations is extended to the earlier of the restructuring date plus sixty months or the termination date plus sixty months. Non-restructured obligations are also deliverable but are subject to the thirty month rule of Mod-R. In both cases the minimum maturity is the termination date of the default swap contract. Deliverable obligations can be conditionally transferable to banks and other entities that make or invest in loans, with the stipulation that borrower consent for novation, assignment, or transfer cannot be unreasonably withheld or delayed. Under the 2003 Definitions, mod mod-r is selected by choosing Modified Restructuring Maturity Limitation and Conditionally Transferable Obligation: Applicable. It is of course possible to choose no restructuring (restructuring is not chosen as a credit event) and credit events are then generally limited to failure to pay and bankruptcy. The US, Australia and New Zealand are expected to select mod-R with the multiple holder obligation standard, Europe mod mod-R with the multiple holder standard, Japan and Asia full restructuring without the multiple holder standard, Emerging Markets (exasia) full restructuring including the multiple holder standard. Credit Event Notice Upon Restructuring parties may elect to partially exercise credit protection following a restructuring credit event. In effect, more than one credit event notice can be delivered. eg a protection seller may have sold protection to two or more entities, and then hedged the exposure by buying protection from a single seller. If some buyers do not trigger upon restructuring, the seller would not have deliverable obligations to deliver under its hedge without this provision. However, if the credit event is not a restructuring event then the exercise amount must be equal to the floating rate payer calculation amount (and not a portion thereof). Pari Passu Ranking any reference obligation in the context of a restructuring will have its seniority determined as of the trade date or the date on which the obligation was issued, without regard to any changes in seniority. The objective here is to ensure that should a reference obligation be subordinated in a restructuring the protection buyer can only deliver the reference obligation and not some other subordinated security or loan. Supplement Relating to Successor and Credit events to the 1999 ISDA Credit Derivatives Definitions ISDA has used this supplement to develop cutoff levels to determine how CDS contracts should be allocated in a successor situation (merger, demerger, consolidation, amalgamation, transfer). The Supplement has been adopted in the 2003 Definitions.


46


The original 1999 Definitions expected that successor events would result in an entity that assumes all or substantially all of the obligations of the original reference entity. It did not anticipate a situation where a corporate might break itself up, as happened with National Power (a UK utility company that split into Innogy Holding Plc and International Power Plc). The initial test will be 75%. If one entity succeeds to 75% or more of the relevant obligations of the reference entity that are outstanding immediately prior to the succession event, then that entity will be the sole successor. If no entity meets the 75% test, then a more than 25% test will be used. If only one of the surviving entities succeeds to more than 25% of the relevant obligations, this entity will be the sole successor. If one or more entities succeeds to more than 25% of the relevant obligations then the transaction will be divided into separate transactions for each of these entities and the notional amount evenly divided among them. The division will be done evenly between the contracts to avoid odd-sized contracts and disputes as to exact proportions. If no successor meets the 25% test then the G6 committee recommends turning to the original reference entity. If the original reference entity remains in existence, regardless of the amount of the relevant obligations it has succeeded to, then there will be no successor and the transaction will not change. If the original reference entity no longer exists and no entity succeeds to 25% of the relevant obligations then the entity that succeeds to the greatest percentage of bonds and loans of the original reference entity will become the sole successor. If, however, two entities succeed to equal portions of the bonds and loans then the one that succeeds to the greatest percentage of obligations will be the sole successor. It is still possible for an investor to reference a specific obligation as deliverable in the CDS, making it a single stock trade. The 2003 Definitions also now include spin-off as a succession event. The Successor Supplement also clarified the currency of payment definitions. In the original 1999 Definitions, any currency switch would trigger a restructuring event. Now, any change in the currency or composition of any payment of interest or principal to any currency that is not a Permitted Currency will be defined as a restructuring. Permitted Currency is defined as a currency from a G7 member country or that of a country that has a long-term debt rating equal to or more favourable than AAA from S&P, Aaa from Moodys, and AAA from Fitch IBCA. These currencies are sufficiently liquid to be interchangeable without causing a credit event. The Obligation Default has largely fallen into disuse by the market (it captures default situations other than failure to pay, where an obligation is in technical default, eg perhaps due to a violation of covenants).


47


Obligation Acceleration refers to situations where an obligation has been accelerated for technical reasons other than bankruptcy or non-payment (eg violation of covenants). Typically, loans or obligations are rarely accelerated unless default is close. The US market currently tends to exclude both Obligation Default and Obligation Acceleration, whereas the European market tends to include Obligation Acceleration. ISDA is taking this stance because Obligation Default and Obligation Acceleration can be triggered even though the reference entity has the ability and willingness to meet the original terms of the obligation. ISDA is therefore trying to exclude from the definitions events that are some distance away from outright default. The Successor definitions are widely used globally. Supplement to the 1999 ISDA Credit Derivatives Definitions Relating to Convertible, Exchangeable or Accreting Obligations. Issues arose in the credit derivative market following the placing of Railtrack into Administration in October 2001 (Railtrack is the infrastructure owner of Great Britains railway system). Protection buyers delivered the 3.5% 2009 convertible as it was the cheapest to deliver (trading at around a 15 point discount to other Railtrack obligations). The CB was trading at a discount because the government had announced that it was cutting off equity holders, resulting in the CB trading at its bond floor, whilst the government continued to service straight Railtrack debt. With a 3.5% coupon, the CB was at a large discount to straight Railtrack bonds once the equity conversion option had been taken away. Protection sellers pointed out that the 1999 definitions disallowed contingent securities as deliverable obligations and therefore questioned the ability of buyers of protection to deliver CBs. In the Supplement on convertibles, adopted in the 2003 Definitions, ISDA indicated that convertibles, along with exchangeable, or accreting bonds, could be delivered into CDS even if the Not Contingent condition had been specified. ie convertibles do not count as contingent securities and are deliverable obligations. The key point is that the convertible/exchangeable feature must be solely at the investors discretion, and that it not have been exercised. Emerging Markets Repudiation/Moratorium the 1999 definitions have been expanded to clarify that a repudiation/moratorium must be declared by an authorised officer of the reference entity or appropriate government authority and it must be accompanied by a failure to pay or restructuring.


48


Appendix 2 : Total Return Swaps and Credit Spread Options Total Return Swaps Total Return Swaps (TRS) allow investors to exchange the total economic returns of an asset for fixed or floating interest payments or vice versa. They provide a synthetic method to access off balance sheet assets or manage portfolio risk more efficiently. Total return swaps give clients exposure to particular sectors without incurring the costs of ownership and administration of the underlying assets. They optimise the efficiency of capital by achieving exposure to market and credit risk with reduced transaction, administrative and taxation costs. In many markets TRS can lock in term financing costs more efficiently than repos. It also allows the investor to finance an asset for which there is no traditional repo market. Total Return Loan Swap Structure

Libor + Spread

Funding SourceLibor-spread

Deutsche Bank

Coupon + Appreciation

Investor

Total Return

Depreciation

Bond/Loan

The TRS gives protection against the risk of portfolio assets without removing them from the balance sheet. Of course, the return payer need not hold the asset on its balance sheet to enter into a TRS - the TRS then allows the payer to short the asset. In a TRS there is no exchange of principal or exchange of ownership. TRS can be settled for cash or by physical delivery. If it is to be cash settled the market value is established by polling a group of dealers. Credit Spread Options Options on credit spreads allow investors to isolate credit risk from market risk and to express a view about an assets credit risk profile in the future. They can be used to49



earn premium income to profit from spread tightening or widening and to buy securities on a forward basis at favourable prices. Credit spread options are normally written on bonds. The spread or yield differential represents the risk premium the market demands for holding the issuers bonds. Spread calls make money from decreasing spreads (bond values rise if yields fall). Spread puts make money from widening spreads (bond values fall if yields rise). Maturities are usually between six months and two years and are settled in cash or through delivery of the bond. Essential Reading : Credit Derivatives and Structured Credit, Deutsche Bank Global Markets Research, August 30 2001. 2003 ISDA Credit Definitions, Deutsche Bank Global Markets Research, June 13th 2003 Trading Forward Spreads : Where is the Value ? DB Global Markets Research, July 2004 CMCDS : The Path to Floating Credit Spread Products DB Global Markets Research, March 2004 Trading the CDS-Bond Basis DB Global Markets Research, August 2004 Equity Default Swaps : A Primer DB Global Markets Research, February 2004


50


Thats the end of the session (at last !) You should now : Understand the mechanics of credit default swaps Understand the pricing of credit default swaps Know the key users of default swaps Be aware of the links between credit derivatives and synthetic CDOs Be able to explain constant maturity credit default swaps Be able to explain equity default swaps Be able to identify bond/default swap relative value opportunities Confused ? email Mike Pawley at [email protected] publication is for internal use only by Deutsche Bank Global Markets employees. The material (including formulae and spreadsheets) is provided for education purposes only and should under no circumstances be used for client pricing. Examples, case studies, exercises and solutions may use simplifying assumptions that do not apply in practice, and may differ from Deutsche Bank proprietary models actually used. The publication is provided to you solely for information purposes and is not intended as an offer or solicitation for the purchase or sale of any financial instrument or product. The information contained herein has been obtained from sources believed to be reliable, but is not necessarily complete and its accuracy cannot be guaranteed.


51