Credit Derivatives Overview Iima

JULY 2006

CREDIT DERIVATIVES Vaidya Nathan

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

VAIDYA NATHAN

1

Credit derivatives in the context of financial markets growth

C R E D I T

D E R I VA T I VE S

O V E R VI E W

New applications expanding financial instruments use

re a c t s pa n u od to s ird r p w ging & th cts e u N er nd d o o m e sec pr n o ati r ne ge

As s ex et cl t tra ende asses dit d ion bey getti n al o ma nd g rke ts

VAIDYA NATHAN

2

Role of Credit Derivatives Motivations Motivations for for use use of of Credit Credit Derivatives Derivatives

C R E D I T

D E R I VA T I VE S

O V E R VI E W

Synthetically create loanbond; alternative to equity derivatives

Generate leverage or yield enhancement

Hedge, transfer and/or mitigate credit exposure

Manage regulatory capital ratios

Decompose and separate credit risks embedded in financial instruments

Proactively manage credit risk on a portfolio basis

VAIDYA NATHAN

3

Credit derivatives isolate and transfer credit risk Broad definition „ bilateral financial contract which allows specific aspects of credit risk to be

C R E D I T

Loan/bond

D E R I VA T I VE S

O V E R VI E W

isolated from the other risks of an instrument, and passed from one counterparty to another

Credit FX, Interest Rate

On-balance sheet

60 bps

6.60 % yield

Off-balance sheet VAIDYA NATHAN

4

Credit derivatives perform a market completion role Bond = Duration + Convexity + Credit

Convexity Risk Credit risk

C R E D I T

D E R I VA T I VE S

O V E R VI E W

including callability risk sometimes i.e. negative convexity

Duration Risk VAIDYA NATHAN

5

Efficiency gains arising from disaggregating risk

Auctioneer sells a number of risks, each to the highest bidder

C R E D I T

D E R I VA T I VE S

O V E R VI E W

JOB LOT

VAIDYA NATHAN

6

Spreads of Credit Default Swaps can be compared to bond yields Bond / Loan

Asset Swap

Credit Default Swap

Credit Risk Credit Risk Credit Risk Funding Risk Risk Free Rate

C R E D I T

D E R I VA T I VE S

O V E R VI E W

Funding Risk

VAIDYA NATHAN

7

The simplest instrument: single name credit default swaps

Reference Entity

Risk (Notional)

C R E D I T

D E R I VA T I VE S

O V E R VI E W

Fee/premium

B

A

Protection Buyer

Protection Seller

„ Buy CDS

Contingent Payment upon a credit event

„ Sell CDS

„ Buy Protection

„ Sell Protection

„ “Short Risk”

„ “Long Risk”

„ Pay periodic payments

„ Receive periodic payments

„ Receive contingent payment

„ Pay contingent payment

VAIDYA NATHAN

8

Indicative Summary Terms of CDS General General Terms Terms

Fixed Fixed Payments Payments

„ Effective Date: 23 Nov 2006

„ Fixed Rate Payer Notional: USD 25,000,000

„ Scheduled Termination Date: 23 Nov 2008

„ Fixed Rate Payer Payment Dates: The 23rd

„ Floating Rate Payer: X (the “Seller”) „ Fixed Rate Payer: Y (the “Buyer”) „ Business Day: London & New York

of February, May, August and November, commencing on February 23, 2007 „ Fixed Rate: X% per annum „ Fixed Rate Day Count Fraction: Actual/360

„ Business Day Convention: Modified Following

C R E D I T

D E R I VA T I VE S

O V E R VI E W

„ Reference Entity: ABC „ Reference Obligation(s) - The obligation(s)

identified as follows: „ Primary Obligor: ABC Corporation „ Maturity: 15 September 2011 „ Coupon: 6.5% „ CUSIP/ISIN: USXXX „ Original Issue Amount: USD 1,000,000,000

Floating Payment Payment Floating „ Floating Rate Payer Notional: USD 25,000,000 „ Conditions to Payment:

1. Credit Event Notice „ Notifying Party: Buyer or Seller 2. Notice of Publicly Available Information Applicable „ Public Source(s): Standard Public Sources „ Specified Number: Two 3. Notice of Physical Settlement VAIDYA NATHAN

9

Indicative Summary Terms of CDS Credit Credit Events Events

Settlement Settlement Terms Terms

„ Credit Events: The following Credit Event(s)

„ Settlement Method: Physical Settlement

shall apply to this Transaction: „ Bankruptcy „ Failure to Pay „ Restructuring „ Grace Period Extension: Not Applicable „ Payment Requirement: USD 1,000,000 or its

C R E D I T

D E R I VA T I VE S

O V E R VI E W

equivalent in the relevant Obligation Currency „ Default Requirement: USD 10,000,000 or its

equivalent in the relevant Obligation Currency „ Obligations: „ Obligation Category: Borrowed Money „ Obligation Characteristics: None

„ Physical Settlement Period: Section 8.5 of the

ISDA Credit Derivatives Definitions, subject to a maximum of 30 Business Days „ Portfolio: Exclude Accrued Interest „ Deliverable Obligation Category: Bond or Loan „ Deliverable Obligation Characteristics: „ Pari Passu Ranking „ Specified Currencies: Standard Specified

Currencies „ Assignable Loan „ Consent Required Loan „ Transferable „ Not Contingent „ Maximum Maturity: 30 years „ Not Bearer „ Restructuring Maturity Limitation Applicable

VAIDYA NATHAN

10

In reality assumes two names risk

Reference Entity

Risk X bps per annum

Counterparty

Bank

C R E D I T

D E R I VA T I VE S

O V E R VI E W

Contingent Payment

„ Buyer decreases exposure to Reference Credit(s), but assumes contingent (“two-

name”) exposure to Seller „ Seller receives a fee in return for making a Contingent Payment if there is a Credit

Event of the Reference Credit which in turn depends on the financial health of the bank

VAIDYA NATHAN

11

Replicating a Credit Default Swap

Corporate Asset

$ 100

T + Corporate Spread (SC) T + Swap Spread (SS)

Collateral

Swap Market

Investor

C R E D I T

D E R I VA T I VE S

O V E R VI E W

Libor $ 100 * ( 1 – haircut )

Repo rate (L – x)

Repo Market

Credit Default Swap Spread (approx.) = Corporate Spread (Sc) – Swap Spread (Ss) Assumption: Haircut is small ( ≈ 0) & repo rate spread is negligible ( ≈ 0)

VAIDYA NATHAN

12

Funding Cost Arbitrage Lender

AAA Rated Institution L – 20 bp L + 40 bp BBB Asset L + 25 bp

Lender

AA- Rated Institution

BBB Asset 25 bp AA- Rated Institution

C R E D I T

AAA Rated Institution Contingent payment in event of default

D E R I VA T I VE S

O V E R VI E W

L + 40 bp

L + 40 bp BBB Asset

Credit Assessment AAA to AA-

A+ to A-

BBB+ to BBB- BB+ to B-

Below B-

Unrated

Risk Weight

50%

100%

150%

100%

20%

100%

VAIDYA NATHAN

13

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

VAIDYA NATHAN

14

Degrees of leverage in various Credit Derivative structures

Baskets Capital protection + minimum coupon/interest Increasing Yield

Capital protection

First-To-Default

First-To-Default with Mark-To-Market

Senior

Mezzanine

First Loss (Junior)

C L N

&

L I N EA R

B A S K E T

Non-capital protected

Linear

Portfolio Tranching

VAIDYA NATHAN

15

Ab Initio - Credit-linked notes Tailored notes „ Structured for investor’s required currency, maturity, and coupon needs „ CLNs can be rated and / or listed if required „ Both physical and cash delivery are available to the investor „ Provide solutions for many investors restricted from entering into OTC transactions „ Provide investors with yield and minimum ratings requirements through leveraged

C L N

&

L I N EA R

B A S K E T

high grade structures

VAIDYA NATHAN

16

Structure of a Typical CLN

Interest Rate Swap

Protection Sale

Proceeds

Swap Counterparty

Collateral

SPV CDS Premium

Proceeds

CLN

C L N

&

L I N EA R

B A S K E T

Investors

VAIDYA NATHAN

17

Linear Baskets Linear basket swaps allow investors to gain exposure to multiple credits in one trade „ Risk buyer takes on exposure to each credit equal to the 1/N of the notional of the

basket, where N is the number of credits in the basket (assuming equal weighting) „ After the first credit event: „ swap on the defaulted credit terminates, „ notional of the trade is reduced by the notional of the defaulted credit, „ the investor bears exposure to the non-defaulted credits „ Yield on these structures is additive, since each credit is independent of the other „ Advantage of less documentation by taking exposure to many credits in one single

trade

C L N

&

L I N EA R

B A S K E T

(the same as yield on first-to-default basket with zero correlation)

VAIDYA NATHAN

18

Advantages of Credit Linked Notes

CP Risk Risk & & CP Credit Line Line Credit Usage Usage

No No Direct Direct Derivatives Derivatives Contract Contract Non-issuers Non-issuers reference reference

No Nosystem system requisites requisites

CLN Advantages Customized Customized Maturity Maturity

Relative Relative Value Value

Tailored Tailored Exposure Exposure

C L N

&

L I N EA R

B A S K E T

Canbe be Can listed listed

VAIDYA NATHAN

19

Disadvantages of CLNs

Medium Term View

Cheapest to Deliver Option

Funded Form & Lack of Leverage

C L N

&

L I N EA R

B A S K E T

Lack of Liquidity

VAIDYA NATHAN

20

Linear Baskets - an illustration

X bp per annum on notional 100

A

B

C

D

E

Investor Investor Contingent Payment

Protection Buyer

(Par — Recovery on Credit E)

Protection Seller

X bp per annum on notional 80

A

B

C

D

Investor Investor Contingent Payment

Protection Buyer

(Par — Recovery on defaulted credit)

Protection Seller

C L N

&

L I N EA R

B A S K E T

Example: Green Bottle Swap on 5 equally weighted names on a notional of 100 If Credit E defaults, Protection Seller pays Par and receives the Recovery Value on Credit E The rest of the swap remains, with the notional falling to 80

VAIDYA NATHAN

21

Linear Basket - Example

300 bps

A

100 bps

B

Equal weighted Linear Basket Spread

400 bps

C

150 bps

D

200 bps

E

=230 bps

C L N

&

L I N EA R

B A S K E T

= (300 + 100 +400 + 150 +200)/5

VAIDYA NATHAN

22

Benefits of CDX

Liquidity Liquidity

Diversification Diversification

Cost efficient and timely access to the Credit Markets via index swaps and credit-linked securities Daily reports on actual versus theoretical pricing

C L N

&

L I N EA R

B A S K E T

Transparency Transparency

Use Credit Default Swaps to maximize liquidity portfolios are composed of the most liquid credit default swap names

VAIDYA NATHAN

23

CDS indices: CDX and iTraxx Two major CDS indices trade actively at tight bid-ask spreads „ DJ CDX in North America has 125 names „ DJ iTraxx Europe has 125 names

Weekly fixings on three CDS indices: DJ iTraxx Europe, HiVol index and Crossover index Can also trade loss tranches on index „ Tranches are like synthetic CDO tranches „ Just as options are way to trade volatility, tranches are way to trade „ One-factor Gaussian copula is standard for quoting correlations

C L N

&

L I N EA R

B A S K E T

default correlations

VAIDYA NATHAN

24

CDS indices for Asia and Australia Nonlinearity Nonlinearity of of risky risky duration duration for for half half and and double double credit credit spreads spreads

„ iTraxx CJ has 50 Japanese names, with sub-indices for capital goods, tech

and HiVol „ iTraxx Asia has 30 names from outside Japan with sub-indices for Korea,

Greater China and rest of Asia „ iTraxx Australia has 25 names „ CDX.EM has 14 emerging market sovereigns, including Korea, Malaysia and

C L N

&

L I N EA R

B A S K E T

the Philippines

VAIDYA NATHAN

25

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

VAIDYA NATHAN

26

First To Default (FTD) Baskets First-to-default basket swaps allow investors to leverage their exposure to a basket of credits „ After the first credit event, the first-to-default swap terminates and the investor no

longer bears exposure to the non-defaulted credits „ Yield enhancement in these structures basically depends on the correlation of the

names in the basket „ Risk buyer takes on exposure to each credit equal to the notional of the basket,

B A S K E T

P R O D U C T S

thus achieving leverage of the number of names in the basket

VAIDYA NATHAN

27

First To Default (FTD) Baskets — an illustration

X bp per annum

A

B

C

D

Protection Buyer

E Contingent Payment (Par - Recovery on Credit E)

Protection Seller

„ If Credit E defaults, Protection Seller pays Par and receives the Recovery Value on

Credit E „ The first-to-default swap is terminated and Protection Seller has no further

B A S K E T

P R O D U C T S

exposures „ The greater the correlation, the greater the probability of multiple defaults in the

basket

VAIDYA NATHAN

28

Mechanics of a First To Default Basket Structure „ Risk is sourced from the market „ First to Default Tranche is sold either in bond or swap form „ Bank retains “Senior” Tranche

Bank Credit Default Swaps on 5 Individual Credits

CDS on 1st to Default

OTC Investors

B A S K E T

P R O D U C T S

Market

VAIDYA NATHAN

29

From individual default probability to basket default probability Individual Individual default default probability probability

Basket default probability

Correlation Correlation of of assets assets

Taking a leveraged exposure to a basket is equivalent to trading the correlation between those names

150 bps

A

50 bps

B

200 bps

C

Intuition: Intuition: The The higher higher the the correlation, correlation, the the lower lower the the spread spread on on the the leveraged leveraged piece piece

100% correlation „ The basket behaves like 1 single credit

Protection Seller will expect to receive the widest of individual spreads

B A S K E T

P R O D U C T S

0% correlation

FTD Spread = 575 bps

Correlation = 1.0 75 bps

D

100 bps

E

„ Each name in the basket behaves

independently. Protection Seller should receive the sum of the individual spreads

Correlation = 0

FTD Spd = 200 bps

VAIDYA NATHAN

30

High and Low Correlation – the Tom & Jerry way

B A S K E T

P R O D U C T S

High High & & Low Low Correlation Correlation

VAIDYA NATHAN

31

Risk illustrated Correlation = 0

B

A D

C

E

Correlation = 1

B A S K E T

P R O D U C T S

B E C A 0 < Correlation < 1

A C

B E D VAIDYA NATHAN

32

Actual FTD trades – Example 1 Background Background

Trade Summary Summary Trade

„ Korean client buys FTD loan on six

„ Aggregate bid spread: 537 bps

investment grade credits „ 50% of names chosen were local credits „ 50% of names were foreign credits „ Inclusion of foreign credits helps

reduce correlation of the basket which inturn helps increase spread

„ FTD Basket coupon: 6.75% „ FTD spread over Libor: 335 bps „ FTD spread over Libor as a % of

aggregate spread: 62% „ FTD spread over Libor as a % of highest

bid spread: 163%

„ Non callable Credit

B A S K E T

P R O D U C T S

Kookmin Bank KEPCO POSCO Hutchison Whampoa Ford Standard Life Assurance

5-year (bids) 62 54 53 98 206 74 VAIDYA NATHAN

33

Actual FTD trades – Example 2 Background Background

Trade Summary Summary Trade

„ Client buys FTD note on five high yield

„ Aggregate bid spread: 950 bps

credits „ Names chosen were high rated (BB)

high yield credits „ Clustered spreads „ Spreads have low correlation to

maximize spread

„ FTD Basket coupon: 10.60% „ FTD spread over Libor: 720 bps „ FTD spread over Libor as a % of

aggregate spread: 75% „ FTD spread over Libor as a % of highest

bid spread: 335%

„ Non callable

B A S K E T

P R O D U C T S

Credit

5-year (bids)

Rating

Amerisource Corporation Chesapeake Energy Corporation

160 215

Ba3/BB Ba3/BB-

Mandalay Resort Group Flextronics International Ltd Georgia Pacific Corporation

210 167 208

Ba2/BB+ Ba2/BBBa2/BB+ VAIDYA NATHAN

34

Actual FTD trades – Example 3 Background Background

Trade Summary Summary Trade

„ Client buys FTD protection on five high

„ Aggregate offer spread: 795 bps

yield credits

„ FTD Basket premium: 5.0%

„ Reduced costs relative to hedging each

of the individual credits „ Sheds substantial portion of the risk „ Credits have high correlation to

„ FTD spread over Libor as a % of

aggregate spread: 63% „ FTD spread over Libor as a % of highest

bid spread: 222%

minimize cost

B A S K E T

P R O D U C T S

Credit Delphi Corporation Ford Motor Company General Motors Acceptance Corporation Lear Corporation Visteorn Corporation

5-year (bids) 113 213 157 87 225 VAIDYA NATHAN

35

Default Correlation Default correlations are key determinants of hedge ratios which determine basket premiums that dealers are willing to pay. The boundary conditions for the basket premium can be restated in terms of the default correlation as follows: „ Basket premiums should decline with an increase in correlation. A basket of

uncorrelated credits trading at similar spreads produces the largest relative increase in premium compared to the average single-name default swap premium „ Default correlations impact the likelihood of multiple defaults up to a given time

horizon. In practice, there is a lack of historical data that could be used to extract default correlations. Instead, market players use the asset correlation to calculate default correlation

B A S K E T

P R O D U C T S

„ Asset correlations can be extracted from the ability-to-pay process of a portfolio of

firms. Such a process is modeled for an individual firm as its market value of assets minus liabilities. Market inputs are equity and debt data. The asset correlation derived in this manner is deterministically related to the default correlation, i.e. one can be transformed into the other „ The most difficult factor to incorporate in the model for pricing basket products is

the estimation of the underlying correlations. Estimating the correlation between two default events cannot be achieved by standard statistical methods. Instead some proxy for default or credit behaviour must be used VAIDYA NATHAN

36

How to deal with correlation? Model the correlation between names as a result of the correlation of each name with a systemic “market” variable (CAPM) „ given a market environment, the names default independently „ the probability of default of any given name depends on the market environment „ As correlation approaches zero: FTD Basket Premium = Sum of Basket Component’s

CDS Spreads „ As correlation approaches one: FTD Basket Premium = Highest Individual CDS Spread

in Basket „ Investors wishing to maximize premium generated by selling FTD Basket protection

B A S K E T

P R O D U C T S

should consider selecting a basket that is relatively uncorrelated

VAIDYA NATHAN

37

Par spreads for FTD with different betas Par Par spreads spreads for for FTD FTD with with different different betas betas 550

Par Spreads with Time for various betas

500

450

400

350

Rate of decrease of spreads is high for high correlation Beta = 10%

300

Beta = 50%

Beta = 90%

Max Spread

250

B A S K E T

P R O D U C T S

200

Time ( yrs)

150 1.00

2.00

3.00

4.01

5.01

6.01

7.01

8.01

9.01

10.01

12.01

15.01

20.02

25.02

30.02

40.03

Basket of five names each trading at 100 bps. Rate of decrease of spreads is high for high correlation

VAIDYA NATHAN

38

Expected loss for FTD with different betas Expected Expected loss loss for for FTD FTD with with different different betas betas 100.00%

Expected Loss for FTD with time for various betas

90.00% 80.00%

Beta = 10%

Beta = 50%

Beta = 90%

70.00% 60.00% 50.00% 40.00% 30.00% 20.00%

B A S K E T

P R O D U C T S

10.00% Time (yrs)

0.00% 0.25

1.00

1.50

2.00

3.00

4.01

5.01

6.01

7.01

8.01

9.01

10.01

12.01

15.01

20.02

25.02

30.02

40.03

VAIDYA NATHAN

39

Dynamic Hedging Of The FTD Basket „ The hedging behaviour of a dealer provides some intuition behind the actual basket

premium „ A dealer that buys protection on a basket from an investor would normally hedge

this transaction by selling default protection on each individual name in the basket „ As the underlying default premiums shift, the deltas will change and the hedges

will need to be rebalanced dynamically „ The efficiency with which the hedge can be managed is a key factor that

determines the basket premium „ For small movements in the hedge ratio, the dealer may not be able to sell or buy

B A S K E T

P R O D U C T S

protection and may instead buy or sell bonds to hedge, thus taking on basis risk

VAIDYA NATHAN

40

Dynamic Hedging Of The FTD Basket Protection Buyer

Protection Seller X bp per annum

A

B

C

D

E

FTD FTDInvestor Investor Contingent Payment (Par - Recovery on Credit E) Notional = N

AANotional NotionalNNAA==Hedge HedgeRatio RatioAAxxNN BBNotional HedgeRatio RatioBBxxNN NotionalNNBB==Hedge

B A S K E T

P R O D U C T S

CCNotional HedgeRatio RatioCCxxNN NotionalNNCC==Hedge DDNotional HedgeRatio RatioDDxxNN NotionalNNDD==Hedge EENotional HedgeRatio RatioEExxNN NotionalNNEE==Hedge VAIDYA NATHAN

41

Dynamic Hedging of FTD Basket „ Following a credit event, the dealer will be forced to unwind

the hedges on the other credits (assuming non-zero deltas for these credits)

150 bps

A

50 bps

B

200 bps

C

75 bps

D

100 bps

E

„ The cost of unwinding the hedge would depend on the spread

movement for each of the non-defaulted credits. This, in turn, would depend on the correlation between the defaulted and the non-defaulted credits „ The greater this correlation, the greater the expected spread

B A S K E T

P R O D U C T S

widening for a non-defaulted single-name default swap. This would imply a greater cost of unwinding the hedge. The dealer would therefore maintain a lower delta i.e. sell a lower amount of protection, to minimise losses from the unwind. This would, in turn, provide a lower premium to pay for the basket protection „ On the other hand, a low correlation would imply a lower

expected spread change in a non-defaulted credit in the event of default and consequently a lower cost of unwinding that hedge. The hedger could therefore maintain a higher delta to manage the hedge i.e., sell a higher amount of protection. This provides a higher premium to pay for the basket protection

VAIDYA NATHAN

42

Investment rationale for a FTD Basket Swap „ Efficient leverage and limited downside : Investors are able to efficiently leverage

credit risk with a defined downside potential by executing a FTD Basket Swap, since the swap’s notional amount references a basket that is 5x to 10x its size „ Limited maximum loss: a $10mm FTD Basket Swap that references a basket of ten

names (aggregate credit exposure of $100mm) still limits the investor’s maximum loss to the first loss in the basket — or $10mm „ Basket component correlation premium : FTD baskets that are relatively

uncorrelated pay investors a premium that approaches the sum of the basket’s component default premiums rather the spread of the basket’s riskiest component

B A S K E T

P R O D U C T S

„ Efficient usage of regulatory capital: The notional amount of the swap, rather than

the notional amount of the reference basket, would be used in a risk-based capital requirement calculation for a US bank. This allows banks to take credit risk to a reference portfolio that is 5x-10x as large as the notional amount of the swap for which it must set aside capital

VAIDYA NATHAN

43

Investment rationale for a FTD Basket Swap „ Investment grade ratings available: Rating agencies have been willing to provide

investment grade ratings to FTD Basket Swaps that meet rating agency requirements regarding credit risk and diversification „ Fund managers with limits on scope of assets to invest: This has allowed investors

that are limited to specific ratings-based investment guidelines to earn premiums well in excess of those typically found in the single-name investment grade bonds or credit default swaps „ Executable via swap or funded note/deposit: Investors may elect sell FTD Basket

protection via a swap or funded credit linked note (CLN) „ The settlement mechanics of a FTD Basket Swap are identical to a single name

B A S K E T

P R O D U C T S

credit default swap (physical settlement of defaulted securities)

VAIDYA NATHAN

44

First 2 To Default (F2TD) Baskets „ First-2-to-default basket swaps allow investors to leverage their exposure to a

basket of credits to a lesser extent „ The protection buyer is protected against the first two defaults „ Total premium for a F2TD basket should be lesser than for a first to default „ After the first credit event, contract settles partially and notional of the trade is

reduced by half

B A S K E T

P R O D U C T S

„ After the second credit event, contract settles fully and the trade terminates

VAIDYA NATHAN

45

First 2 To Default (F2TD) Baskets

X bp per annum

A

B

C

D

E Contingent Payment (Par - Recovery on Credit E)

Protection Buyer

Protection Seller

„ If Credit E defaults, Protection Seller pays Par and receives the Recovery Value on Credit E

X bp per annum

A

B

C

D

B A S K E T

P R O D U C T S

Protection Buyer

Contingent Payment (Par - Recovery on Credit D)

Protection Seller

„ If Credit D defaults, Protection Seller pays Par and receives the Recovery Value on Credit D

Trade terminates

VAIDYA NATHAN

46

First-to-default vs First-2-to-default Basket Spreads 200 180

180 170

BPS

160 140

158 145

140

135

125

120

112 100 92 80 4Y

5Y

B A S K E T

P R O D U C T S

F2TD

FTD1

8Y FTD2

FTD1 : Pacific Dunlop, Pasminco, MIM, Mayne Nickless, Qantas FTD2 : CWOptus, Boral, United Energy, Woodside, Western Mining F2TD : FTD1 + FTD2

VAIDYA NATHAN

47

Other Basket default products „ Second to default : The protection buyer is protected only against the second default,

and pays a premium until the occurrence of this event. The premium will be less than for a FTD, since two defaults are always less likely than one (assuming that the credits are not perfectly correlated) „ Second, third, fourth or fifth to default : The premium for an nth to default decreases

as n increases, since the probability of n defaults becomes more unlikely. The effect of correlation is to increase the premium, since this increases the probability of multiple defaults. Relatively speaking, this effect is more dramatic for larger n „ Sum of first, second, third, fourth and fifth to default : The total premium for these

five contracts is always greater than the sum of the individual default swap premiums

B A S K E T

P R O D U C T S

„ Quanto basket default swap : Similar to other quanto products. Assume that the FTD

basket is cash-settled. Upon a credit event, observe the reference obligation (presumably denominated in USD), then protection buyer receives 1-R x Quanto FX i.e. an FX rate agreed „ Actual Quanto Trade: a Taiwanese investor buys protection and pays a running TWD

fee. The basket consists of 5 names with reference obligations denominated in USD. There is a credit event, recovery on the deliverable is 40. The Taiwanese investor gets (100-40) x whatever FX rate that was agreed at the beginning of the trade. For trades in general, the fee and the payout don't have to be the same currency VAIDYA NATHAN

48

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

VAIDYA NATHAN

49

A synthetic tranche is a specific allocation of risk of a synthetic CDO A synthetic CDO pools the risk of various synthetic assets — corporates, sovereigns, financials — in the form of a portfolio of CDS Credit risk is re-allocated from individual credits to tranches „ Equity owner is in the first loss position „ Subordinated tranches (including Equity) are leveraged exposures „ Senior tranches (AAA, AAAA) are de-leveraged exposures

Each tranche exhibits different risk characteristics which reflect its expected share of portfolio losses as well as its sensitivity to changes in the expected distribution of losses within the portfolio

S Y N TH E T I C

C D O

An active risk management strategy allows banks to create and offer tranche protection or exposure on a standalone basis

VAIDYA NATHAN

50 9

Each tranche is defined in terms of pay-off Example Example of of a a mezzanine mezzanine tranche tranche Tranche loss

CDS Buyer of Protection

CDS Seller of Protection Fee

Tranche loss

Tranche size

S Y N TH E T I C

C D O

Tranche size

Portfolio loss

VAIDYA NATHAN

10 51

Sum of risks (losses) is equal to portfolio risk (loss) Tranche loss

Senior Eq. size

Mezz. size

Portfolio loss

Tranche loss Mezz. size Mezz. size

Portfolio loss

Tranche loss

Equity

S Y N TH E T I C

C D O

Eq. size Eq. size

Mezzanine

Portfolio loss

VAIDYA NATHAN

52 11

S Y N TH E T I C

C D O

Risk of a tranche is defined by its position in the capital structure Portfolio

Tranched structure

Individual credit default swaps

Third-loss piece (Senior piece)

De-leveraged

Second-loss piece (Mezzanine)

Medium risk

First-loss piece (Equity)

Highly leveraged

VAIDYA NATHAN

53 12

Synthetic first loss tranche is created by sourcing delta amount of CDS „ As in a CDO, risk is sourced from the CDS market; however, the amount of risk

sourced reflects only the risk of the first loss tranche „ By synthetically reproducing its specific risk profile, any tranche can be created

without the need to place the remainder of the capital structure

This risk is managed through active rebalancing of the 1st Loss delta

Risk

Market

Bank Spread

Risk on 1st Loss Tranche

Risk on 1st Loss Tranche

S Y N TH E T I C

C D O

Delta amount of CDS on individual credits

Delta is specific to the risk of the 1st Loss Tranche

Investor

SPV Spread

Spread

Collateral

VAIDYA NATHAN

54 13

Pricing framework for a synthetic tranche is based on expected loss „ At the portfolio level, spread compensates for expected loss: Portfolio Expected Loss = PV (Portfolio Credit Spread)

Risk

Compensation for Risk

„ Tranching reallocates losses within the portfolio „ The spread on an individual tranche must compensate for the tranche

expected loss

S Y N TH E T I C

C D O

Tranche Expected Loss = PV (Tranche Credit Spread)

VAIDYA NATHAN

55 15

To compute expected loss, obtain portfolio loss distribution Ingredients for the calculation of the loss distribution: „ Recovery rate for each name — use CDS market estimates „ Probability of default for each individual name — derived from the credit spread

and recovery rate estimate: „ PD = S/(1-Recovery Rate)

S Y N TH E T I C

C D O

„ Default correlation between names — derived from historical asset correlations

VAIDYA NATHAN

56 16

Estimating loss distribution requires tranching, spreads & recoveries A A simplified simplified example example of of a a portfolio portfolio containing containing only only one one name name Tranched structure

Portfolio: Spread (bps) 150

Notional 100

Maturity = 1 year

30%— Mezz

10%— Eq

S Y N TH E T I C

C D O

Recovery Rate Assumption = 50%

60%— Senior

VAIDYA NATHAN

57 17

Probability-weighting loss scenarios gives expected loss of tranches Portfolio Portfolio loss loss distribution distribution Eq

Mezz

0%

Senior

50%

100%

Tranche pricing pricing Tranche Probabilit y

Not ional default e d

Port folio loss

Port folio loss (%)

Equit y loss

Mezz loss

Sr loss

1

97. 00%

0

0

0%

0%

0%

0%

2

3. 00%

100

50

50%

10%

30%

10%

Whole port folio

Equit y

Mezz

Senior

Expe ct ed loss (% of Not )

1. 5%

0. 3%

0. 9%

0. 3%

Tranche Not ional

100%

10%

30%

60%

Tranche Spread (bps)

150

300

300

50

S Y N TH E T I C

C D O

Scenarios

VAIDYA NATHAN

58 18

To model loss distribution, we take into account default correlation

Credit Spreads

Asset Correlations

Recovery Rates

Market Implied Default Probabilities

Default Correlations

Model

S Y N TH E T I C

C D O

Loss distribution

VAIDYA NATHAN

59 19

Address default correlation in estimating the loss distribution? „ We model the correlation between names as a result of the correlation of each

name with a systemic market variable (similar to CAPM) „ Given a market environment, the names default independently „ However, the probability of default of any given name depends on the market

environment „ Integrating across all possible market environments yields a loss distribution which

S Y N TH E T I C

C D O

then incorporates the correlation between names

VAIDYA NATHAN

60 20

Probability of default depends on the market environment 1-year default default probability probability 1-year

Specific Specific market market scenarios scenarios Market environment Probability of market

Bad

Neutral

Good

1/3

1/3

1/3

1 year def. Prob for “Bad”

1 year def. prob for “Neutral”

1 year def prob for “Good”

(Average over markets) Name 1 1.0%

Name 1

1.90%

1.00%

0.10%

Name 2

2.0%

Name 2

3.80%

2.00%

0.20%

Name 3

0.5%

Name 3

0.50%

0.50%

0.50%

S Y N TH E T I C

C D O

Name 3 is uncorrelated with the market

VAIDYA NATHAN

61 21

Final loss distribution is calculated by averaging over all scenarios Probability Probability of of loss loss

Bad market

times 1/3

Loss Neutral market

times 1/3

Loss

S Y N TH E T I C

C D O

Good market

times 1/3

VAIDYA NATHAN

62 22

Loss distribution changes with spreads, recoveries, and correlation Probability Probability of of loss loss

Senior

when correlation increases

when correlation decreases

S Y N TH E T I C

C D O

when spread decreases

when spread increases

Loss

Average loss

VAIDYA NATHAN

63 23

Portfolio loss distributions for two portfolios

S Y N TH E T I C

C D O

Tranche Tranche loss loss different different for for same same tranche tranche size size for for two two different different portfolios portfolios

VAIDYA NATHAN

64 23

As loss distribution changes, delta of any specific tranche changes „ In a fully-sold synthetic CDO, the risk of changes in the loss distribution is passed

completely to investors „ The distribution of losses within the portfolio will fluctuate, but the portfolio

hedge is static „ In a synthetic tranche, bank’s hedge must offset the risk position of the specific

tranche „ Initial delta reflects the relative proportion of the portfolio loss distribution

which falls within the tranche „ Therefore, as the tranche expected loss changes, the delta hedge for that tranche

S Y N TH E T I C

C D O

must also be rebalanced

VAIDYA NATHAN

65 24

Risk management for synthetic tranches is rapidly evolving „ New technology allows for the creation of synthetic tranches with a range

of attractive characteristics beyond the original static „ Tranche exposure can be offered in local currency whereas risk is sourced

in a foreign currency „ Term of tranche credit exposure can differ from the term of the note or

swap „ Credit exposure of coupons can differ from the exposure of the principal „ Investor or asset manager can be offered limited portfolio management

S Y N TH E T I C

C D O

flexibility

VAIDYA NATHAN

66 26

Credit derivatives offer some unique characteristics Credit derivatives: „ unbundle credit risk from other aspects of ownership (tax,

accounting, liquidity, relationship) „ are similar in substance to many traditional credit instruments „ are not triggered by underlying price movements but only by default „ provide the only efficient short positioning vehicle „ frequently require explicit consideration of correlation risk „ are available on- or off-balance sheet and can provide leverage

S Y N TH E T I C

C D O

efficiently

VAIDYA NATHAN

67

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

VAIDYA NATHAN

68

Options on Credit Default Swaps „ Options on credit default swaps can be structured into trades that work for various

parties with different objectives „ Fund managers can sell options to earn premium while waiting for spreads to

widen to their target investment levels „ Hedge funds can sell volatility through options to earn premium „ Investors can buy callable CLNs/CDS to earn higher returns than on plain-vanilla

CLNs/CDS „ Investors can buy principal protected notes that are linked to spread tightening

or spread widening options or their combinations higher returns or reduce their risk profile on their investment

CR E D IT

DE R I VAT I VE

OP TI ON S

„ Investors can combine CDS options with interest rate or equity products to earn

VAIDYA NATHAN

69

Options terminology „ A call on a credit default swap is the right to buy risk/sell protection (receiver

option) „ A put on a credit default swap is the right to sell risk/buy protection (payer option) „ A straddle is a combination of a put and a call at the same strike „ ATM denotes an option struck at-the-money

CR E D IT

DE R I VAT I VE

OP TI ON S

„ OTM denotes an option struck out-of-the-money

VAIDYA NATHAN

70

Options payout diagrams Bullish Strategy

Bullish Strategy

Decreasing Price Increasing Spreads

80 bps

Increasing Price Decreasing Spreads

50 bps

Long call

Decreasing Price Increasing Spreads

20 bps

80 bps

(unlimited upside when spreads tighten

CR E D IT

DE R I VAT I VE

OP TI ON S

downside limited to premium paid)

Bearish Strategy

(unlimited upside when spreads widen downside limited to premium paid) Decreasing Price Increasing Spreads

80 bps

Increasing Price Decreasing Spreads

50 bps

20 bps

50 bps

20 bps

Short put (unlimited downside when spreads widen, upside limited to premium earned)

Bearish Strategy

Long put

Increasing Price Decreasing Spreads

Short call (unlimited downside when spreads tighten, upside limited to premium earned)

Decreasing Price Increasing Spreads

80 bps

Increasing Price Decreasing Spreads

50 bps

20 bps

VAIDYA NATHAN

71

Options on single name credit default swaps „ Underlying: 5yr CDS on a single Reference Entity „ Maturity: 3 months „ Premium: prices are quoted in cents, premium is paid T+3 business days „ Strike: At-the-money-spot spread of the 5yr CDS derived from the current trading

level at the time of the trade „ Exercise: European (only at the maturity of the option)

CR E D IT

DE R I VAT I VE

OP TI ON S

„ Settlement: „ Physical: into the 5yr CDS upon exercise at the pre set strike „ Cash: option is unwound upon exercise and cash paid out to client „ Knock out: „ Both calls and puts knock out in case of a Credit Event on the Reference Entity

VAIDYA NATHAN

72

Options on TRAC-X Europe „ TRAC-X Europe options are a liquid, standard product to trade credit volatility on a

macro basis „ Underlying: TRAC-X Europe 99 Swaps with 20 September 2011 maturity „ Maturity: 3 months or 6 months „ Premium: prices are quoted in cents, premium is paid T+3 business days „ Strike: Preset strikes of 30, 35, 40, 45, 50 and 55, chosen to reflect ATM and OTM

CR E D IT

DE R I VAT I VE

OP TI ON S

forward levels in increments of 5 bps „ Exercise: European (only at the maturity of the option) „ Settlement: Physical, into TRAC-X Europe 99 Swaps upon exercise at the pre set

strike „ Knock out: No knock out in case of a Credit Event in TRAC-X Europe 99 Swaps

VAIDYA NATHAN

73

CR E D IT

DE R I VAT I VE

OP TI ON S

TRAC-X

VAIDYA NATHAN

74

Trade Idea 1: Long credit trade to earn premium „ A fund manager who finds France Telecom too tight compared to his target investment levels

could sell an OTM put on France Telecom to earn premium

15 ¢

CR E D IT

DE R I VAT I VE

OP TI ON S

Decreasing Price Increasing Spread

BE: 80 bps

Put strike: 76 bps

ATM: 66 bps

Manager receives 15 cents to sell OTM put Increasing Price Decreasing Spread

„ Breakeven analysis: „ Investor receives 15 cents to sell a 3-month European put on France Telecom struck at 76 bps „ If France Telecom widens by more than 4 bps (= 15 bps / 4.3 duration) from the strike in 3

months, the investor will lose money on the option (I.e., if France Telecom widens beyond 80 bps, the investor will lose more on the option than the option premium earned) — However, the manager will be put into France Telecom risk at a strike corresponding to his target investment level VAIDYA NATHAN

75

Trade Idea 2A: Short credit trade „ An investor who believes the market will widen in the short to medium term could buy an ATM

put on TRAC-X Europe to benefit from spreads widening

Decreasing Price Increasing Spread

Put strike: ATM: 45 bps BE: 53 bps

36 ¢

Increasing Price Decreasing Spread Investor pays 30 cents to buy ATM put

CR E D IT

DE R I VAT I VE

OP TI ON S

Breakeven analysis: „ Investor pays 36¢ to buy a 3-month maturity European put on TRAC-X Europe struck at 45 bps „ If TRAC-X Europe widens by more than 8 bps (= 36¢ / 4.3 duration) from the strike in 3 months, the

investor will make money from exercising his ATM put (i.e., if TRAC-X Europe widens beyond 53 bps, the investor will make more on the option than the option premium paid) „ If TRAC-X Europe widens to 60 bps in 3 months, the investor will make 64.5¢ (= (60-45) bps x 4.3

duration) from exercising his ATM put. The investor’s payout ratio in this case will be 1.79 (= 64.5¢ payout / 36¢ option premium paid) „ If TRAC-X Europe widens to 70 bps in 3 months, the investor will make 1.1% (= (70-45) bps x 4.3

duration) from exercising his ATM put. The investor’s payout ratio in this case would be 2.99 (= VAIDYA NATHAN 1.1% payout / 36¢ option premium paid)

76

Trade Idea 2B: Short credit trade with capped upside „ If the same investor finds the ATM put too expensive and thinks market spreads will widen but

not by too much, he could buy an ATM put and sell an OTM put to subsidise cost of the ATM put

Bear spread payout

CR E D IT

DE R I VAT I VE

OP TI ON S

Decreasing Price Increasing Spread

Put strike: 60 bps

BE: 45 bps

Put strike: ATM: 45 bps 30 ¢

Increasing Price Decreasing Spread Investor pays 21 cents to buy bear spread

Breakeven analysis: „ Investor pays 30¢ net to buy a 3-month maturity European put on TRAC-X Europe struck at 45 bps and sell a 3-month European put on TRAC-X Europe struck at 60 bps „ If TRAC-X Europe widens by more than 7 bps (= 30¢ / 4.3 duration) from 50 bps in 3 months, the investor will make money on the ATM put he bought (i.e., the upside from buying the option will be higher than the option premium paid) „ Maximum payout on this trade will be 64.5¢, if TRAC-X Europe goes to 60 bps or wider (= {60 bps – 45 bps} x 4.3 duration). The investor’s payout ratio in this trade is 2.15 (= 64.5¢ / 30¢ option premium paid) „ If investor thinks the market will not widen beyond 60 bps, he should put on the bear spread (payout ratio of 2.15) instead of buying the ATM put outright (payout ratio of 1.79 when spreads are at 60 bps) VAIDYA NATHAN

77

Trade Idea 3: Short volatility trade „ A hedge fund does not have a strong credit view but believes that credit market will

experience little volatility in the short to medium term could trade volatility and earn premium by selling a 3-month ATM straddle on TRAC-X Europe 36 ¢ Investor receives 36 cents to sell straddle B/E on put: 53 bps

CR E D IT

DE R I VAT I VE

OP TI ON S

Decreasing Price Increasing Spread

ATM: 45 bps

B/E on call: 37 bps Increasing Price Decreasing Spread

„ Breakeven analysis: „ Hedge fund receives 36 cents to sell a 3-month maturity European straddle on

TRAC-X Europe struck at 45 bps „ If TRAC-X Europe tightens or widens by more than 8 bps (= 36¢ / 4.3 duration)

from the strike in 3 months, the fund will lose money on the option (i.e., the downside from selling the option will be higher than the option premium earned) VAIDYA NATHAN

78

Trade Idea 4: Callable CDS „ A CDS investor with investment targets that are higher than current market levels could sell 10yr France

Telecom protection callable by bank in 5 years that pays more than the 10yr plain-vanilla CDS Callable CDS Premium for 10yr FRTEL CDS callable in 5yrs If 5yr FRTEL CDS < 95 bps in 5yrs

Bank

Bank calls the CDS from the investor

Investor

If 5yr FRTEL CDS > 95 bps in 5yrs

CR E D IT

DE R I VAT I VE

OP TI ON S

Investor holds the CDS contract for 10yrs

„ Breakeven analysis: „ 5yr France Telecom CDS: 66 bps, 10yr France Telecom CDS: 89 bps „ 10yr France Telecom CDS callable in 5 years pays 95 bps (WHAT IS THE CATCH HERE?) „ If Bank calls the CDS in 5 years, investor will have earned 29 bps (44%) more running than the

5yr CDS, and FRTEL 5yr CDS would have to tighten by more than 29 bps for investor to lose the 29 bps earned in the first 5 years to enter into a new France Telecom CDS „ If Bank does not call the CDS, investor will have earned 6 bps (7%) more running than selling

plain-vanilla 10yr protection VAIDYA NATHAN

79

Binary (Digital) Credit Swaps

Z bps per annum Protection Buyer

Protection Seller Contingent Payment (Par)

CR E D IT

DE R I VAT I VE

OP TI ON S

Clients looking for leverage opportunities on a single name can sell protection in a binary swap „ In a plain-vanilla credit swap, the protection seller would receive the recovery value of the

Reference Credit in a credit event „ In a binary swap, the protection seller receives no recovery value in a credit event „ Client receives a higher spread to compensate for zero recovery in a credit event

VAIDYA NATHAN

80

Digital (Binary) Default Swaps „ Digital default swaps will demand a higher premium than a standard default swap „ Its price will be sensitive to the recovery rate that has been assumed in the

calibration procedure „ The higher the recovery rate in the calibration, the higher the calibrated hazard

rates will be „ Since a digital default swap depends only on the hazard rates and not on the

CR E D IT

DE R I VAT I VE

OP TI ON S

recovery rate, it has a price that effectively increases with the assumed recovery

VAIDYA NATHAN

81

Digital Default Swaps „ Based on run

above the quote on ABY offers a recovery bidoffer of 35-45 with the underlying ABY 5yr CDS spread at 297bp

CR E D IT

DE R I VAT I VE

OP TI ON S

„ An investor who

believes the recovery rate on ABY will be below 35 in the future should sell ABY recovery at 35

VAIDYA NATHAN

82

Digital Default Swaps Executing this position will result in two trades: „ Investor will be short protection on a DDS (Digital Default Swap) with a recovery

rate fixed at 35, a spread of 297bps, 5yr maturity „ Investor will be long protection on a standard CDS, spread of 297bps, same maturity

date „ The net of the two positions has zero carry

CR E D IT

DE R I VAT I VE

OP TI ON S

Profiting in Default: „ If ABY defaults and bonds are trading at 20 (i.e. actual recovery is 20) investor

earns 15 points on the combined position: „ On short protection position (DDS) investor loses 65 (100 – fixed recovery rate of 35) „ On the long protection position (regular CDS) investor earns 80 (100 less cost to buy

bond in market at 20)

VAIDYA NATHAN

83

Calculate MTM on DDS Calculate the MTM after one year assuming DDS is now trading at 20 instead of 35 Original positions: „ Short fixed recover position at 35 (Loss on default is 100 – 35 = 65) „ Long regular CDS position (Gain on default is 100 – R)

Unwind Positions: „ Long fixed recovery position at 20 (Gain on default is 100 – 20 = 80)

CR E D IT

DE R I VAT I VE

OP TI ON S

„ Short regular CDS position (Loss on default is 100 – R) „ Notional: USD 10 million, ignore discount rates

Year

Survival Probability

2

0.95

3

0.90

4

0.85

5

0.80

VAIDYA NATHAN

84

Cancelable Credit Swaps

Y bp per annum Protection Buyer

Protection Seller Contingent Payment (Par — Recovery)

CR E D IT

DE R I VAT I VE

OP TI ON S

Right to cancel

„ Seller receives a fee in return for making a Contingent Payment if there is a Credit Event of the

Reference Credit „ Buyer decreases exposure to Reference Credit(s), but assumes contingent (“two-name”)

exposure to Seller „ Buyer has the right to cancel the trade (European or American option)

VAIDYA NATHAN

85

Yield Pickup on structured CDS Vanilla Mitsubishi Corp. CDS „ 3yrs @ 14bps, 5yrs @ 18bps, 10yrs @ 30bps

Cancelable Mitsubishi Corp. CDS „ 3yrs @ 15.5bps, 5yrs @ 20.5, 10yrs @ 33.5

Binary Mitsubishi Corp. CDS

CR E D IT

DE R I VAT I VE

OP TI ON S

„ 3yrs @ 18.6bps, 5yrs @ 24bps, 10yrs @ 40bps

VAIDYA NATHAN

86

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

VAIDYA NATHAN

87

Credit Spreads

Credit Spreads

Recovery Rate

C D S

VAL U A T I O N

Default Probability

Credit Spreads = Default Probability x (1-Recovery Rate)

VAIDYA NATHAN

88

Credit Default Swaps transfer default risk

X bp per annum Protection Buyer

Protection Seller Contingent Payment (Par — Recovery)

„ Via a Credit Default Swap, the Protection Buyer transfers risk that the

Reference Entity will default „ Protection Seller receives a fee, similar to an insurance premium, and

C D S

VAL U A T I O N

assumes the risk that the Reference Entity will default „ Upon a default or Credit Event, protection seller makes a contingent

payment to the protection Buyer calculated according to the actual losses (e.g. if Reference Entity recovers at 60%, protection seller pays 40%)

VAIDYA NATHAN

89

Caselet - EuroAutos

No Default

Broker/Dealer Protection Seller

1.6% p.a., Q, Act/360 Zero

XYZ Bank Protection Buyer

Quarterly Payments Stop

Default

Broker/Dealer Protection Seller

Notional

XYZ Bank Protection Buyer

C D S

VAL U A T I O N

Senior Unsecured Obligations with notional amount of $10mn

VAIDYA NATHAN

90

Measure of default probability is required to price risky cash flow „ Two sources exist: „ Historical Analysis

i)

Proprietary Assumptions

„ Company/Sector expertise „ Fundamental Analysis

ii)

Market Data

„ Calibrate probability against

market levels

„ Since we will be hedging with market instruments, it is essential that we

derive our default probabilities from the same instruments

C D S

VAL U A T I O N

„ As hedging instruments, credit default swaps offer superior liquidity,

lowest cost, and maximum flexibility

VAIDYA NATHAN

91 3

Specifications of credit markets are similar to interest rate markets Yield Curve

Market Details Currency Trade Date

16-May-06

Days to Spot

2

Value Date

18-May-06

VAL U A T I O N

Conventions

C D S

Spd to Mid

USD Swap #

Maturity

2

Maturity

Actual

Mid

Spreads

Dates

Times

yield

Bid

Mid

Swap

Zero

Ask

rates

rates

1

1D

19-May-06

0.003

3.500%

2.00

0.00

2.00

3.50%

3.56%

2

1M

19-Jun-06

0.088

3.700%

2.00

0.00

2.00

3.70%

3.76%

3

2M

18-Jul-06

0.167

3.785%

2.00

0.00

2.00

3.79%

3.85%

Swap Basis

360

4

3M

18-Aug-06

0.252

3.870%

2.00

0.00

2.00

3.87%

3.93%

Swap Days

B

5

4M

18-Sep-06

0.337

3.909%

2.00

0.00

2.00

3.91%

3.96%

MMkt DCC

Act/365

6

5M

18-Oct-06

0.419

3.948%

2.00

0.00

2.00

3.95%

3.99%

Swap DCC

Act/365

7

6M

20-Nov-06

0.510

3.987%

2.00

0.00

2.00

3.99%

4.03%

Swap cpns PA

2

8

9M

20-Feb-07

0.762

4.065%

2.00

0.00

2.00

4.07%

4.08%

Swap BDC

M

9

1Y

18-May-07

1.000

4.130%

2.00

0.00

2.00

4.13%

4.13%

MMkt Basis

360

10

18M

19-Nov-07

1.507

4.177%

2.00

0.00

2.00

4.18%

4.22%

MMkt freq.

4

11

2Y

19-May-08

2.005

4.207%

2.00

0.00

2.00

4.21%

4.25%

0.00

12

3Y

18-May-09

3.003

4.254%

2.00

0.00

2.00

4.25%

4.30%

1

13

4Y

18-May-10

4.003

4.284%

2.00

0.00

2.00

4.28%

4.33%

GBP

14

5y

18-May-11

5.003

4.319%

2.00

0.00

2.00

4.32%

4.37%

Mean Reversion Interpolation type Libor Holidays

VAIDYA NATHAN

92 3

Same as money market conventions

C D S

VAL U A T I O N

Market Conventions CCY

AUD

EUR

GBP

HKD

JPY

SGD

USD

USR

Days2Spot

1

1

1

1

1

1

1

1

Float Cnv

Act/365F

Act/360

Act/365F

Act/365F

Act/360

Act/365F

Act/365

Act/360

Fixed Cnv

Act/365F

30/360

Act/365F

Act/365F

Act/365F

Act/365F

Act/365

30/360

MM Basis

365

360

365

365

360

365

360

360

MM Freq

2

2

2

4

2

2

4

4

Swap CPA

2

1

2

4

2

2

2

2

Swap Days

A

B

A

A

A

A

B

B

Swap Basis

365

360

365

365

365

365

360

360

Swap Bad Day Conv

M

M

M

M

M

M

M

M

LIBOR Hol

AUD

EUR

GBP

HKD

JPY

SGD

GBP

USR

VAIDYA NATHAN

93 3

Additional aspects USD

Credit Reference Parameters

Trade Date

16-May-06

Fee

Pay

Days to Spot

3

Interval

Accrued

Reference Details Currency

Settlement Date

Index

19-May-06 Q

FALSE

Q

Payment

Payment

Mean

Spread

Recovery

Day Count

Bad Day

Reversion

Interpolation

Rate

Convention

Convention

Act/365

M

(%) 0.00

1

50.00%

Credit Spreads Tenors

1D

1M

3M

6M

1Y

2Y

3Y

Dates

22-May-06

19-Jun-06

21-Aug-06

20-Nov-06

21-May-07

19-May-08

19-May-09

Bid-Offer Spread

3.0

3.0

3.0

3.0

3.0

3.0

3.0

Bid Spread

98.50

98.50

118.50

138.50

148.50

Mid Spread

100.0

100.0

120.0

140.0

150.0

Ask Spread

101.50

101.50

121.50

141.50

151.50

2.005%

2.005%

2.417%

2.836%

3.049%

Clean Spreads

2.005%

2.005%

C D S

VAL U A T I O N

Duration

0.254

0.496

0.967

1.860

2.689

Credit Reference Volatility Structure Tenors

1D

1M

3M

6M

1Y

2Y

3Y

Dates

20-May-06

19-Jun-06

19-Aug-06

19-Nov-06

19-May-07

19-May-08

19-May-09

15.00%

15.00%

15.00%

15.00%

ATM Volatilities

VAIDYA NATHAN

94 3

Trade Specifications Trade Summary

CDS References

CDS Price (Deterministic)

372,020.31

CDS Fee

CDS Price (Tree)

360,488.06

Trade Recovery Rate

Opt. Premium (Tree)

0.00

Maturity

2Y

Opt. Premium(Deterministic)

0.00

Coupon Interval

Q

Hedge Cost

0.00

First Fixing Date 19-May-2006

360,488.06

Next Regular Fixing Date 19-May-2006

Total Trade Inputs Trade Notional

VAL U A T I O N

50.00%

Maturity Date 19-May-2008 100,000,000.00

Fee Day Count

ACT/365

Trade Date

16-May-2006

Accrual Bad Day

M

Sttlement Date

19-May-2006

Payment Bad Day

M

Forward Start Time(eg. 1M, 1Y)

C D S

120.00 bp

1y

CDS Structure

Call / Put

Call

Pay Accrued Fee on Default

FALSE

CDS (Long / Short)

Long

Pay at Maturity

FALSE

Option (Long / Short)

Long

CDS Fee payment tlll maturity

FALSE

Option References

Option Exercise Schedule

Option Maturity

1Y

Exercise Interval

Q

Exercise Effective Date

19-May-2006

Maturity Date

21-May-2007

Exercise

Exercise

Strike

#

Unadjusted

Adjusted

Date(s)

0

19-May-06

19-May-06

19-May-06 VAIDYA NATHAN

95 3

CDS Cashflow due from Buyer and Seller of Protection Buyer of protection pays quarterly premium to seller until the earlier of a credit event or maturity X

3m

X

X

X

X

6m

9m

12m

15m

At inception: PV of both legs are equal

3m

6m

18m

...

60m

...

60m

Credit Event

9m

12m

15m

18m

C D S

VAL U A T I O N

100 - R The seller of protection pays par less recovery to the protection buyer if there is a credit event during the life of the contract

VAIDYA NATHAN

96

CDS Cashflow due from Buyer and Seller of Protection

N

Risky PVFIXED = ∑ S .DFi .SPi .α i i =1

S is the per-annum CDS spread N is the number of coupon periods DFi is the riskless discount factor from time t0 to ti SPi is the Survival Probability of the reference entity from time t0 to ti αi is the accrual factor from ti-1 to ti R is the recovery rate on the delivered obligation

C D S

VAL U A T I O N

N

Risky PVFLOATING = ∑ (1 − R ).DFi .(SPi −1 − SPi ) i =1

VAIDYA NATHAN

97

Valuation of any risky cash flow is based on concept of risky PV „ As default risk increases, the PV of a risky cash flow decreases „ This corresponds to discounting a risky cash flow with a risky discount

factor „ A risky discount factor is alternatively expressed as the product of a

risk-free discount factor and a survival probability „ To price a “risky” instrument we therefore need „ Payment structure „ Risk free discount curve

C D S

VAL U A T I O N

„ Default probability curve

VAIDYA NATHAN

98 2

Deriving default probability from CDS spreads A A 1-year 1-year default default swap swap with with annual annual coupon coupon

ND ( 0 ,1)

P

S

1 year annual CDS

0

(1 − P( ND 0 ,1 ) )

-(1-R)

Period 1

C D S

VAL U A T I O N

0=

1 (1 + r risk free

⎡P ND x S - (1 − P ND ) x (1 - R)⎤ ⎥⎦ ) ⎢⎣

VAIDYA NATHAN

99 4

By bootstrapping, a term structure of default prob can be estimated Default Default probability probability tree tree construction construction

P( 1ND ,2 )

S

S ND ( 0 ,1)

P

(1 − P( 1ND ,2 ) )

0

(1 − P( ND 0 ,1 ) )

(100-R)

Period 1

Period 2

P(0ND ,1)

solved in Period 1 to deduce

P(1ND ,2)

C D S

VAL U A T I O N

Using bootstrapping method we can use in Period 2

(100-R)

5 V A I D Y A N A T H A N 100

Useful Rule of Thumb „ Market equilibrium should ensure that expected loss is equal to the PV of

any spread paid in compensation for bearing the risk:

Expected loss

Spread

S = P D (1 − R) „ Rearranging gives a simple expression for default probability in terms of

CDS spread and recovery rate

S P = (1 − R)

C D S

VAL U A T I O N

D

1 1− ( 1 + s(t : t,T )n ) p(t,T ) = 1− θ

6 V A I D Y A N A T H A N 101

Calculation of default prob is complicated by a number of factors „ Recovery assumptions „ Recovery amount cannot be known in advance; therefore an assumption

must be made „ Assumptions about what will be recovered can vary:

— Recover a flat cash amount — Recover a percentage of risk free PV — Recover a percentage of outstanding notional plus accrued interest „ Interpolation „ Credit default swaps typically pay fee quarterly „ Fee accrual

C D S

VAL U A T I O N

„ In a standard CDS, only the accrued spread is usually paid at time of

default

7 V A I D Y A N A T H A N 102

Effect of recovery assumption on implied survival probability

Higher Default Probability High Recovery Assumption

Lower Survival Probability

Higher Survival Probability

Low Recovery Assumption

C D S

VAL U A T I O N

Lower Default Probability

V A I D Y A N A T H A N 103

Survival probability with time for different recovery rates Survival Survival probability probability with with time time for for different different recovery recovery rates rates (CDS (CDS Spread Spread = = 100 100 bps) bps) 100.00%

Survival Probabilities with Time

90.00% 80.00% 70.00% 60.00% 50.00%

R= 90%

R= 50%

R= 10%

40.00% 30.00% 20.00% 10.00% Time (in y rs)

C D S

VAL U A T I O N

0.00% 0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

For low recovery rate assumptions, survival probability decreases approximately linearly over time. For high recovery rate assumptions, this relationship is more ‘convex’ V A I D Y A N A T H A N 104

Decline in survival probability with higher recovery rates Survival Survival probability probability with with time time for for different different recovery recovery rates rates (CDS (CDS Spread Spread = = 100 100 bps) bps) Maturity

R= 90%

R= 50%

R= 10%

0

99.99%

100.00%

100.00%

1

91.12%

98.09%

98.93%

2

83.02%

96.21%

97.87%

75.64%

94.37%

96.82%

4.0027

68.92%

92.56%

95.78%

5.0027

62.80%

90.79%

94.75%

6.0027

57.21%

89.05%

93.74%

7.0027

52.13%

87.35%

92.73%

8.0055

47.50%

85.67%

91.73%

9.0055

43.28%

84.04%

90.75%

10.0055

39.43%

82.43%

89.78%

11.0055

35.93%

80.85%

88.82%

12.0082

32.73%

79.30%

87.86%

13.0082

29.82%

77.78%

86.92%

14.0082

27.17%

76.29%

85.99%

15.0082

24.76%

74.83%

85.07%

16.011

22.55%

73.40%

84.15%

17.011

20.55%

72.00%

83.25%

18.011

18.73%

70.62%

82.36%

19.011

17.06%

69.27%

81.48%

20.0137

15.55%

67.94%

80.60%

C D S

VAL U A T I O N

3

V A I D Y A N A T H A N 105

Caselet: a typical problem of front & middle office folks Client entered into the following CDS trade a year back

Tenors

„ Reference Entity: AT&T Corporation

1D

98.5

100

101.5

1M

98.5

100

101.5

2M

98.5

100

101.5

„ Notional: USD 10 million

3M

98.5

100

101.5

„ Contract Spread: 150 bps

4M

98.5

100

101.5

„ Current bid/offer for AT&T is as below

5M

98.5

100

101.5

6M

98.5

100

101.5

9M

98.5

100

101.5

1Y

98.5

100

101.5

18M

98.5

100

101.5

2Y

98.5

100

101.5

3Y

98.5

100

101.5

4Y

98.5

100

101.5

5Y

98.5

100

101.5

„ Maturity: 5 years

VAL U A T I O N

„ Calculate the MTM on the trade

C D S

Bid Mid Ask Spread Spread Spread

V A I D Y A N A T H A N 106

Survival Probabilities As Weighting Factors

K

MTM = ∑ (Current CDS − ContractCDS).DFi .SPi i =1

Current CDS is the CDS spread currently prevailing in the market Contract CDS is the CDS spread at which the trade was entered K is the number of coupon periods DFi is the riskless discount factor from time t0 to ti

C D S

VAL U A T I O N

SPi is the Survival Probability of the reference entity from time t0 to ti

V A I D Y A N A T H A N 107

Conceptualising CDS Mark-to-Market ORIGINAL X

3m

X

6m

X

9m

X

12m

X

15m

X

18m

X

...

60m

=

NO CREDIT EVENT X-Y

X-Y

15m

18m

X-Y

...

OFFSET

C D S

VAL U A T I O N

3m

6m

9m

12m

15m

18m

Y

Y

...

60m

Y

V A I D Y A N A T H A N 108

60m

But the Cash Flows are risky … ORIGINAL X

X

X

X

X

Nominal Value

3m

6m

9m

12m

15m

18m

...

CREDIT EVENT

60m

Defaulted Obligation

Credit Event

Defaulted Obligation

OFFSET

C D S

VAL U A T I O N

3m

6m

=

9m

12m

15m

Y

18m

...

60m

X-Y

X-Y

15m

18m

X-Y

...

60m

Annuity Cancelled Default Payments Net off

Nominal Value

V A I D Y A N A T H A N 109

Forward default probability for different recovery rates Forward Forward default default probability probability with with time time for for different different recovery recovery rates rates (CDS (CDS Spread Spread = = 100 100 bps) bps) Forward Default Probabilities with Time

10.00% 9.00% 8.00% 7.00% 6.00% 5.00%

R= 90%

R= 50%

15

17

R= 10%

4.00% 3.00% 2.00% 1.00%

C D S

VAL U A T I O N

0.00%

Time (in yrs) 1

2

3

4

5

6

7

8

9

10

11

12

13

14

16

18

19

20

For low recovery rate assumptions, forward default probability decreases approximately linearly over time. For high recovery rate assumptions, it decreases exponentially V A I D Y A N A T H A N 110

Increased forward default probability with higher R Forward Forward default default probability probability with with time time for for different different recovery recovery rates rates (CDS (CDS Spread Spread = = 100 100 bps) bps)

C D S

VAL U A T I O N

Maturity

R= 90%

R= 50%

R= 10%

1

8.87%

1.91%

1.07%

2

8.10%

1.88%

1.06%

3 4.0027

7.38% 6.72%

1.84% 1.81%

1.05% 1.04%

5.0027 6.0027 7.0027

6.12% 5.58% 5.08%

1.77% 1.74% 1.70%

1.03% 1.02% 1.01%

8.0055 9.0055

4.63% 4.22%

1.67% 1.64%

0.99% 0.98%

10.0055

3.85%

1.61%

0.97%

11.0055 12.0082

3.50% 3.20%

1.58% 1.55%

0.96% 0.95%

13.0082 14.0082 15.0082

2.91% 2.65% 2.41%

1.52% 1.49% 1.46%

0.94% 0.93% 0.92%

16.011 17.011 18.011 19.011 20.0137

2.20% 2.00% 1.82% 1.66% 1.52%

1.43% 1.40% 1.38% 1.35% 1.33%

0.91% 0.90% 0.89% 0.88% 0.88%

V A I D Y A N A T H A N 111

Issuer-Weighted Recovery Rate Descriptive Statistics

Mean (1982 2003)

Mean (2004)

Senior Secured

57.4%

80.8%

Senior Unsecured

44.9%

50.1%

Senior Subordinated

39.1%

44.4%

Subordinated

32.0%

NA

Junior Subordinated

28.9%

NA

All Bonds

42.2%

54.3%

C D S

VAL U A T I O N

Priority in Capital Structure

V A I D Y A N A T H A N 112

Effect of recovery assumption on risky duration Higher Default Probability High Recovery Assumption

Lower Survival Probability Lower “Risky Duration”

Higher “Risky Duration” Higher Survival Probability

C D S

VAL U A T I O N

Low Recovery Assumption Lower Default Probability

V A I D Y A N A T H A N 113

Risky Duration for different credit spreads Risky Risky Duration Duration for for constant constant term term structure structure of of credit credit spreads spreads 8

Risky duration for different credit spreads

7

Spread 50 bps flat

Spread 100 bps flat

Spread 200 bps flat

6

5

4

3

C D S

VAL U A T I O N

2

1

Maturity 0 1D

1M

2M

3M

4M

5M

6M

9M

1Y

18M

2Y

3Y

4Y

5Y

6Y

7Y

8Y

9Y

10Y

V A I D Y A N A T H A N 114

Nonlinearity of risky duration Nonlinearity Nonlinearity of of risky risky duration duration for for half half and and double double credit credit spreads spreads

C D S

VAL U A T I O N

Maturity 1D 1M 2M 3M 4M 5M 6M 9M 1Y 18M 2Y 3Y 4Y 5Y 6Y 7Y 8Y 9Y 10Y

Mat times

Spread 50 bps flat

Duration

Spread 100 bps flat

Duration

Spread 200 bps flat

Duration

0.003 0.077 0.170 0.244 0.329 0.416 0.496 0.748 1.000 1.496 2.000 3.008 4.005 5.003 6.003 7.005 8.014 9.011 10.008

50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0

0.0027 0.0764 0.1684 0.2410 0.3245 0.4100 0.4869 0.7297 0.9693 1.4316 1.8893 2.7721 3.5954 4.3756 5.1179 5.8222 6.4956 7.1196 7.7092

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.0027 0.0764 0.1681 0.2404 0.3237 0.4087 0.4851 0.7261 0.9634 1.4194 1.8687 2.7330 3.5253 4.2687 4.9708 5.6303 6.2596 6.8308 7.3654

200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0 200.0

0.0027 0.0763 0.1675 0.2392 0.3220 0.4061 0.4816 0.7190 0.9517 1.3955 1.8285 2.6604 3.3929 4.0663 4.6947 5.2733 5.8261 6.3039 6.7428

V A I D Y A N A T H A N 115

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

V A I D Y A N A T H A N 116

Framework „ Every credit default swap is documented in a contract based on the ISDA format -

called the Confirmation „ The terms used in the Confirmation are defined in the 2003 Credit Derivatives

Definitions (formerly 1999 Definitions) „ A high level of standardisation of documentation exists in the market „ Standardization makes credit default swaps easier to trade, creates transparency

CR E D IT

DE R I VAT I VE S

R O ADMAP

and facilitates market participation

V A I D Y A N A T H A N 117

Key Contract Terms „ Reference Entity - the entity that credit protection covers „ Obligations - Borrowed Money, Bonds or Loans are types of obligations that the protection covers

„ Credit Events - the triggers are Bankruptcy, Repudiation/Moratorium, Failure to Pay and Restructuring, Obligation Acceleration „ Bankruptcy – covers insolvency, appointment of administrators/liquidators, creditor

arrangements, etc

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ Failure to Pay – on one or more Obligations after expiration of any applicable grace period „ Restructuring – agreement between Reference Entity and holders of any Obligation (and

such agreement is not provided for under the terms of that Obligation) with respect to reduction of interest or principal, postponement of payment of interest or principal, change of currency (other than “Permitted Currency”) and subordination

„ Deliverable Obligations - settle contracts with Bonds or Loans with predefined characteristics

V A I D Y A N A T H A N 118

Key Contract Terms - Deliverable Obligations „ If a Credit Event

occurs, the Buyer of protection can deliver Deliverable Obligations to the Seller

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ Deliverable

Obligations are not the same as Obligations — they are more narrowly defined

Deliverable Obligation Categories: No No No No No Yes

Payment Borrowed Money Reference Obligation(s) Only Bond Loan Bond or Loan

Deliverable Obligation Characteristics: Yes Yes No No No No Yes No Yes Yes No No Yes 30 years No Yes

Not Subordinated Specified Currency Standard Specified Currencies Not Sovereign Lender Not Domestic Currency Not Domestic Law Listed Not Contingent Not Domestic Issuance Assignable Loan Consent Required Loan Direct Loan Participation Indirect Loan Participation Qualifying Participation Seller Transferable Maximum Maturity Accelerated or Matured Not Bearer

V A I D Y A N A T H A N 119

2003 Definitions - Why Introduce and Key Changes „ Why Introduce „ To consolidate market experience - the 2003 Definitions represent a development

of the 1999 Definitions. „ Too many supplements (now incorporated) „ Modified Restructuring was not adopted in Europe „ Time to overhaul and clean up definitions

„ Modified Modified Restructuring for Europe „ New Settlement Fallbacks „ New Guarantee provisions

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ Key Changes

V A I D Y A N A T H A N 120

CR E D IT

DE R I VAT I VE S

R O ADMAP

CDS Structural Roadmap REFERENCE ENTITY

Underlying credit risk transferred?

CREDIT EVENT

Types of "default" covered

OBLIGATIONS

Default on which instruments qualify

PROTECTION PERIOD

Occurance of Credit Event

REFERENCE OBLIGATION

Seniority of exposure transferred

DELIVERABLE OBLIGATION

Instruments used for settlement

PHYSICAL SETTLEMENT

CDS settlement V A I D Y A N A T H A N 121

Caselet: Armstrong World Industries US US company company Armstrong Armstrong World World Industries Industries missed missed payments payments on on its its debt debt

„ US company Armstrong World Industries missed payments on its debt,

which triggered credit default swaps „ Its parent company Armstrong Holdings however, did not default „ Many market participants had treated the parent and principal subsidiary

interchangeably and had hedged positions with offsetting contracts in the other entity

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ The lesson here is that there may be substantial credit basis risk between

different entities in the same group „ Worse still, certain contracts in the market had referenced simply

Armstrong without clarifying to which specific entity the contract referred

V A I D Y A N A T H A N 122

Caselet: National Power National National Power Power PLC PLC demerged demerged certain certain assets assets and and subsidiaries subsidiaries into into two two entities entities

„ In November 2000, National Power PLC of the UK demerged certain assets

and subsidiaries into two entities: Innogy and International Power „ In consideration for the transfer of assets to Innogy, shareholders were

given holdings in the new entity „ National Power then changed its name to International Power

„ This demerger prompted substantial debate as to whether Innogy had

become a Successor

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ Innogy also assumed certain debt obligations of National Power

V A I D Y A N A T H A N 123

Non Sovereign Decision Tree Non-Sovereign Non-Sovereign Successor Successor Summary Summary Decision Decision Tree Tree Does New Entity have >75% of Obligations?

It is the Sole Successor for the for the Entire Credit Derivative Transaction

YES NO

CR E D IT

DE R I VAT I VE S

R O ADMAP

YES Does New Entity have <25% of Obligations?

YES Does Reference Entity still exist

No Change to Contract

NO NO Each Successor Assigned New Credit Derivative Transaction – Includes Reference Entity >25%

New Entity Taking Largest % of Relevant Obligations is Successor

V A I D Y A N A T H A N 124

Caselet: Xerox Corporation Xerox Xerox extended extended the the date date for for repayment repayment of of principal principal

„ In the summer of 2002, as part of a wider agreement with its banks,

Xerox extended the date for repayment of principal „ This was in respect of a major syndicated bank facility that was due for

repayment in September „ However, market participants entered a legal dispute about whether this

was a result of a deterioration in creditworthiness reasonably have occurred

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ And over what period prior to Restructuring such deterioration could

V A I D Y A N A T H A N 125

Caselet: Argentina Obligation Obligation Exchange Exchange requirements requirements

„ Obligation Exchange requirements became the subject of legal disputes „ Argentina was facing a tight liquidity situation „ It “requested” local investors to exchange $50bn of bonds for new issues

with lower coupons

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ In question was the meaning of “mandatory” in such circumstances

V A I D Y A N A T H A N 126

Caselet: Railtrack Bankruptcy Bankruptcy Credit Credit Event Event

„ On 7 October 2000, Railtrack plc was placed by the UK government into

Special Railways Administration „ This constituted a Bankruptcy Credit Event „ The announcement date was a Saturday „ Investors who bought credit default swap protection on the Wednesday,

„ Under the current conventions, however, such risks are considerably

reduced

CR E D IT

DE R I VAT I VE S

R O ADMAP

Thursday or Friday of the previous week would have not been covered for this Credit Event

V A I D Y A N A T H A N 127

Caselet: Railtrack CTD Obligation “Widows “Widows and and orphans” orphans” clause clause

„ Following the Railtrack Bankruptcy Credit Event in 2000, the cheapest-to-

deliver obligation was the 3.5% of 2009 exchangeable bond „ Most of the market took the view that, provided the bond is

exchangeable or convertible at the option of the holder, the bondholder should be the beneficiary and the exchange or conversion option within its control

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ One further complication in the Railtrack case was the inclusion of a so

called “widows and orphans” clause in the exchangeable bond which gave the trustee the right to force conversion of the bond on the holder in certain circumstances where it was viewed as being in the interests of the investor „ After a protracted legal dispute, in February 2003, UK courts ruled in

favour of deliverability

V A I D Y A N A T H A N 128

Caselet: Marconi Somewhat Somewhat unusual unusual guarantee guarantee structure structure

„ The Marconi group had a somewhat unusual guarantee structure „ The holding company Marconi PLC provided lenders and bondholders of

subsidiary Marconi Corporation PLC with a guarantee „ Although the bond guarantees were stated to be “unconditional” they

contained a provision that they would fall away upon the repayment of certain other guaranteed obligations

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ In 2002 a Bankruptcy Credit Event occurred in relation to Marconi, and

the approach of market participants was to deliver loans instead of bonds, so as to avoid the risk that the guarantee structure would render the bonds undeliverable (under 1999 Definitions) „ The main exception to this, was where the bond in question was stated as

the Reference Obligation since in most circumstances this is deliverable

V A I D Y A N A T H A N 129

Caselet: Xerox Syndicated Bank Loan extension Pressure Pressure on on Mod-R Mod-R

„ Mod-R worked pretty well in the US till it came under pressure „ In summer 2002, Xerox extended maturities of a syndicated bank loan „ In this case the maturity limitation requirements of Mod-R did not really

insulate Sellers of protection from the “cheapest-to-deliver” risk „ This was because, although not long dated, Xerox’s yen bonds were

CR E D IT

DE R I VAT I VE S

R O ADMAP

trading about 15-20 points lower than where the dollar bank loans were quoted

V A I D Y A N A T H A N 130

Modified Modified Restructuring New features: „ The Restructured Obligation must be a Multiple Holder Obligation (i.e.

more than three holders)

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ If Buyer triggers the contract

— Deliverable Obligations subject to a maturity cap of 60 months for the Restructured Bond or Loan, 30 months for others AND must be Conditionally Transferable Obligations (matches LMA standard) — Buyer may elect to partially settle

V A I D Y A N A T H A N 131

Modified Modified Restructuring „ In 2005, a customer buys 5 year protection on EnergyCo „ June 2006, EnergyCo enters legally binding agreement to restructure

certain of its bonds „ You can deliver (1) restructured bonds maturing before mid 2011 and

(2) non restructured bonds maturing before 2010 60 month Maturity Cap

Maturity Floor

CR E D IT

DE R I VAT I VE S

R O ADMAP

30 month Maturity Cap

2005 Effective

2006

2007

2008

Legally effective date of Restructuring

2009

2010

2011

2012

STD

V A I D Y A N A T H A N 132

Modified Modified Restructuring - US and Europe Feature:

CR E D IT

DE R I VAT I VE S

R O ADMAP

Maturity Cap:

MR (US standard) 30 month cap Floored at STD

MMR (European Standard) • 30 month cap for non

restructured obligations • 60 month cap for restructured obligations • Floored at STD

Transferability:

Must be transferable to extensive list of entities without consent.

Must be transferable to entities regularly engaged in loan and securities markets with consent not to be unreasonably withheld

Obligations covered:

At least three holders and requires a 2/3 majority to implement restructuring.

At least three holders and in the case of a Loan, requires a 2/3 majority to implement restructuring

V A I D Y A N A T H A N 133

New Settlement Fallbacks „ To avoid failed contracts, parties now have an indefinite period of time to settle

the contract „ Buyer must attempt scheduled settlement but if this fails, fallbacks will apply „ Buyer may continue to attempt delivery „ Seller may close out by buying in the Deliverable Obligation or nominating an

CR E D IT

DE R I VAT I VE S

R O ADMAP

alternative for delivery

V A I D Y A N A T H A N 134

New Settlement Fallbacks - Bond Delivery „ Buyer has 30 calendar days to deliver a Settlement Notice „ Buyer has 30 Business Days + 5 Business Day fallback to effect delivery of the Bonds „ If Delivery has not occurred by this date, Seller may buy the Bond in at the lowest

offer. Seller has 4 Business Days to complete the process „ If Seller fails to complete the process, Buyer may continue to attempt delivery.

Buyer can not deliver whilst the buy-in process is in operation „ Seller may try to buy the Bond in again, but must wait at least 8 Business Days

CR E D IT

DE R I VAT I VE S

R O ADMAP

before restarting the process

V A I D Y A N A T H A N 135

New Settlement Fallbacks - Bond Delivery

Buy-in may start any time

CR E D IT

DE R I VAT I VE S

R O ADMAP

30 days Event Determination Date

Last day to deliver Notice of Physical Settlement

30 + 5 BD

Buy-in max 4 BD

≥ 8 BD

Buy-in max 4 BD

Last day for delivery of Bonds before fallbacks begin Buyer may continue to attempt to deliver between buy-in attempts by Seller

V A I D Y A N A T H A N 136

New Settlement Fallbacks - Loan Delivery „ Buyer has 30 calendar days to deliver a Settlement Notice „ Buyer has 30 Business Days + 5 Business Day fallback to effect delivery of the Loan „ If Delivery has not occurred by this date, Buyer may deliver a Transferable Bond

or an Assignable Loan instead provided that Buyer provides certification from a Managing Director that reasonable efforts were used to get consent

„ If Buyer has not delivered anything for a further 15 Business Days, Seller may

CR E D IT

DE R I VAT I VE S

R O ADMAP

nominate an Assignable Loan or a Transferable Bond and require Buyer to purchase and deliver

V A I D Y A N A T H A N 137

New Settlement Fallbacks - Loan Delivery

CR E D IT

DE R I VAT I VE S

R O ADMAP

Buyer can deliver any other Assignable Loan or Transferable Bond

30 days Event Determination Date

Last day to deliver Notice of Physical Settlement

30 + 5 BD

15 BD

Seller can nominate an available Loan or Bond

Last day for delivery of loans before fallbacks begin

V A I D Y A N A T H A N 138

New Guarantee Provisions „ Users can now select what type of guarantees can trigger the contract and what is

deliverable „ Guarantee has to be a “Qualifying Guarantee” i.e. a written instrument where

Reference Entity irrevocably agrees to make payment „ Upstream, downstream, side-stream and third party guarantees may be identified

and treated separately „ Europe “All Guarantees” will be adopted. In US “Qualifying Affiliate Guarantees” „ Qualifying Affiliate Guarantee – Reference Entity guarantees debt of an affiliate

where it owns more than 50 percent of the voting shares of that affiliate

CR E D IT

DE R I VAT I VE S

R O ADMAP

will be adopted

V A I D Y A N A T H A N 139

New Guarantee Provisions

Parent Company

Europe and Asia – covers all these Qualifying Guarantees

50% 50%

CR E D IT

DE R I VAT I VE S

R O ADMAP

Operating Company

Guarantee

US - the only Qualifying Affiliate Guarantee – downstream with a holding of at least 50%

50%

Guarantee Reference Entity Guarantee

Guarantee Third Party

Subsidiary

V A I D Y A N A T H A N 140

Other Key Changes „ CHF is now a standard deliverable currency „ Notice of Intended Physical Settlement becomes Notice of Physical Settlement

(“NPS”) „ Pari Passu Ranking becomes Not Subordinated „ Minor amendments to Restructuring definition, Successor definitions „ Not Contingent definition revised to remove need for a coupon

„ Public Sources broadened to include Australian Financial Review et al plus the main

source of business news in country of Reference Entity „ Scheduled Termination Date no longer subject to adjustment

CR E D IT

DE R I VAT I VE S

R O ADMAP

„ Convertibles language broadened to accommodate wider range of deliverables

V A I D Y A N A T H A N 141

Other Key Changes „ Modified Following replaced with Following as standard convention for all trades „ Valuation provisions restructured so that all Firm Bids are used regardless of

Quotation Size, with zero only deemed for the no bid portion „ Repudiation/Moratorium redrafted to require a subsequent Failure to Pay (in any

CR E D IT

DE R I VAT I VE S

R O ADMAP

size) before the earlier of (a) 60 days and (b) the next payment date for the instrument. Buyer must also deliver a notice of a Potential Repudiation/Moratorium

V A I D Y A N A T H A N 142

Outline Page

C R E D I T

D E R I V A T I V E S

Credit Derivatives Overview

1

CLN & Linear Basket

14

Basket Products

26

Synthetic CDO

49

Credit Derivative Options

68

CDS Valuation

87

Credit Derivatives Roadmap

116

Credit Derivatives Update

143

V A I D Y A N A T H A N 143

Applications for credit derivatives in the global market Motivations Motivations for for using using Credit Credit Derivatives Derivatives

1. Trading/ market making

2. Product structuring 3. Hedging trading instruments

CR E D IT

DE R I VAT I VE S

U PDAT E

4. Active portfolio/ asset management 5. Management of individual credit lines 6. Management of regulatory capital 7. Management of economic capital

V A I D Y A N A T H A N 144

Rankings for application of Credit Derivatives

CR E D IT

DE R I VAT I VE S

U PDAT E

Applications for credit derivatives

Rankings 2003

Rankings 2006 (E)

Trading/market making

1

1

Product structuring

2

2

Hedging trading instruments

3

4

Active portfolio/asset management

4

3

Management of individual credit lines

5

5

Management of regulatory capital

6

7

Management of economic capital

7

6

V A I D Y A N A T H A N 145

Credit Derivatives positions Credit Credit Derivatives Derivatives positions positions

1997 1998 1999 2000

CR E D IT

DE R I VAT I VE S

U PDAT E

2001 2002 2003 2004 2006 (E)

180

350 586

893 1189

1952 3548

5021 8206

V A I D Y A N A T H A N 146

Comparative Interest rate derivatives growth Interest Interest Rate Rate Growth Growth (USD (USD billion) billion) 141,991

121,799

101,658 89,955 77,568

50,015

54,072

60,091

64,125

64,668

Jun-00

Dec-00

67,465

CR E D IT

DE R I VAT I VE S

U PDAT E

42,368

Jun-98

Dec-98

Jun-99

Dec-99

Jun-01

Dec-01

Jun-02

Dec-02

Jun-03

Dec-03

V A I D Y A N A T H A N 147

Breakdown of market participation Market Market Composition Composition

24% 42%

CR E D IT

DE R I VAT I VE S

U PDAT E

34%

Intermediary / Market Maker

Buyer

Seller V A I D Y A N A T H A N 148

Institutions using credit derivatives to buy protection Buyers Buyers of of credit credit protection protection

2006 (Expected)

Other

3% 3% 1%

2003

Pension fundsAgencies 2% Mutual funds 5%

7%

Banks

3%

Securities houses Hedge funds

5%

Corporates

Insurance

16%

Companies

51%

Mutual funds

9%

Pension funds

Banks Corporates

Other Agencies

43%

16%

4%

DE R I VAT I VE S

U PDAT E

1999

CR E D IT

Insurance Companies

6%

7%

1% 1% 1% Banks Securities houses

3%

Hedge funds

Hedge funds Corporates

17%

Insurance Companies Mutual funds

18%

Securities houses 15%

63%

Pension funds Other Agencies

V A I D Y A N A T H A N 149

Sellers of credit protection Institutions Institutions using using credit credit derivatives derivatives to to sell sell protection protection

2006 (Expected)

4%

2003

Other

4% 1% Banks

Pension fundsAgencies 1% Mutual funds 6%

Securities houses Hedge funds

6%

38%

20%

Corporates Insurance Companies

Banks

Mutual funds

2%

34%

Pension funds

Insurance

DE R I VAT I VE S

U PDAT E

Companies

CR E D IT

Other Agencies

15% 16%

1999

21%

2% 3% 1% Banks Securities houses Hedge funds

23%

Corporates

Corporates

3% Hedge funds 15%

Securities houses 14%

47% 3%

Insurance Companies Mutual funds Pension funds

5%

Other Agencies 16% V A I D Y A N A T H A N 150

Credit Derivatives by region Credit Credit Derivatives Derivatives by by region region in in 2006 2006 (Expected) (Expected)

3%

43%

U PDAT E

39%

CR E D IT

DE R I VAT I VE S

10%

London

Europe ex-London

5%

Asia/Australia

US

Other

V A I D Y A N A T H A N 151

Credit Derivatives by region in 2003 & 1999 Credit Credit Derivatives Derivatives by by region region

2003

1999

45%

44%

41%

CR E D IT

DE R I VAT I VE S

U PDAT E

40%

9%

8%

6%

5% 1%

London

Europe exLondon

Asia/Australia

US

Other

1% London

Europe exLondon

Asia/Australia

US

Other

V A I D Y A N A T H A N 152

Credit Derivatives booked by region Credit Credit Derivatives Derivatives booked booked by by region region

14% 6% 14%

CR E D IT

DE R I VAT I VE S

U PDAT E

66%

London

Europe ex-London

Asia/Australia

US

V A I D Y A N A T H A N 153

Credit Derivatives Market Size (US$ bn) 2003

2004

2006

Global market size

3,548

5,021

8,206

London market size

1,586

2,230

3,563

Americas market size

1,459

2,000

3,173

Asia/Australia market size

287

446

858

Other Europe/Rest of World

216

345

612

CR E D IT

DE R I VAT I VE S

U PDAT E

Credit Derivatives Market Size by region

V A I D Y A N A T H A N 154

Global Credit Derivatives Product Usage Global Global Credit Credit Derivatives Derivatives Product Product Usage Usage

4%

2% 1% 1%

4%

51% Single-name credit default swaps

4%

Synthetic CDOs – full capital Synthetic CDOs – partial capital

6%

Full index trades Tranched index trades

CR E D IT

DE R I VAT I VE S

U PDAT E

2%

Credit linked notes Total return swaps Basket products

9%

Asset swaps Credit spread options Swaptions Equity linked credit products

10% 6%

V A I D Y A N A T H A N 155

Current Product Usage Global Global Credit Credit Derivatives Derivatives Product Product Usage Usage –– 2006 2006 (Expected) (Expected)

3%

3% 3% 1% 42%

5%

Single-name credit default swaps Synthetic CDOs – full capital

4%

Synthetic CDOs – partial capital Full index trades

6%

Tranched index trades

CR E D IT

DE R I VAT I VE S

U PDAT E

Credit linked notes Total return swaps

5%

Basket products Asset swaps Credit spread options Swaptions

12%

Equity linked credit products 11%

5% V A I D Y A N A T H A N 156

Underlying reference entity Category Category of of underlying underlying reference reference entity entity

2006 (Expected)

2003

2% 5%

Corporate assets

7%

4% 2%

Financials

8%

Sovereign assets (emerging markets)

22% 64%

22%

Other

64%

U PDAT E DE R I VAT I VE S CR E D IT

Sovereign assets (nonemerging markets)

1999

3% 6%

Corporate assets

9%

Corporate assets Financials

22%

Financials

60%

Sovereign assets (emerging markets)

Sovereign assets (emerging markets)

Sovereign assets (nonemerging markets)

Sovereign assets (non-emerging markets)

Other

Other V A I D Y A N A T H A N 157

Credit rating of underlying reference entity Credit Credit rating rating of of the the underlying underlying reference reference entity entity

2006 (Expected)

AAA – AA

2003

AAA – AA

19%

Below B 16%

A – BBB 13%

63%

4%

A – BBB 17%

1999

BB – B AAA – AA

17%

BB – B

Below B

66%

0%

A – BBB 17%

CR E D IT

DE R I VAT I VE S

U PDAT E

Below B

65%

3%

BB – B

V A I D Y A N A T H A N 158

Tenor distribution Maturity Maturity

2006 (Expected)

2%

2003

7%

Over 10 years

16%

2%

Under 1 year Under 1 year

1 – 5 years

7%

5 years 21%

5 – 10 years

5 – 10 years

21%

54%

CR E D IT

DE R I VAT I VE S

U PDAT E

1999

5%

Over 10 years

9%

9% 5 years 18%

Under 1 year 1 – 5 years

1 – 5 years

52%

5 years 36%

41%

5 – 10 years Over 10 years

V A I D Y A N A T H A N 159

Market Constraints Constraints Constraints in in using using Credit Credit Derivatives Derivatives 1. Lack of client knowledge of the product 2. Regulatory constraints

3. Systems / Infrastructure

CR E D IT

DE R I VAT I VE S

U PDAT E

4. Pricing – lack of data 5. Lack of agreed accounting conventions 6. Lack of homogenous documentation 7. Lack of market liquidity and depth

V A I D Y A N A T H A N 160