Key Rating Issues in Synthetic CDO’s - Mayer BrownSynthetic CDO StructuresSynthetic CDO Structures...
Transcript of Key Rating Issues in Synthetic CDO’s - Mayer BrownSynthetic CDO StructuresSynthetic CDO Structures...
Key Rating Issues in Synthetic CDO’s
Key Rating Issues in Synthetic CDO’s
Tania Cunningham, CFA - Associate Director July 16, 2003
Tania Cunningham, CFA - Associate Director July 16, 2003
Types of SyntheticsTypes of Synthetics
Synthetic Index Trades
Eg. HYDI Credit Linked Trust
Synthetic Index Trades
Eg. HYDI Credit Linked Trust
Single Name CreditLinked Notes
Eg. TIERS Trust
Single Name CreditLinked Notes
Eg. TIERS Trust
nth to DefaultBaskets
Eg. (private)
nth to DefaultBaskets
Eg. (private)
Static InvestmentGrade Synthetics
Eg. Julia CDO, Ltd.
Static InvestmentGrade Synthetics
Eg. Julia CDO, Ltd.
Managed Synthetics
Eg. Shoreline
Managed Synthetics
Eg. Shoreline
High Yield Arbitrage CLOs
Eg. SERVES
High Yield Arbitrage CLOs
Eg. SERVES
SyntheticStructures
SyntheticStructures
Single TrancheCDOs
Eg. Private
Single TrancheCDOs
Eg. Private
Typical Synthetic CDO StructureTypical Synthetic CDO Structure
InvestorSPV
(Protection Seller)Swap Counterparty(Protection Buyer)
Reference Entities Collateral
Proceeds
Proceeds
CouponCDS Premium
Protection Payment
Collateral Economics
Credit Default Swap
Credit Linked Note
Synthetic CDO StructuresSynthetic CDO Structuresä Characteristics of a Credit Default Swap:
— Protection buyer/ protection seller
— Reference entities
— Credit events
— Obligations on which a credit event can be called
— Obligations on which loss settlement can be determined
— Settlement method: cash or physical settlement
— Valuation process (if cash settled)
ä Characteristics of a Credit Default Swap:
— Protection buyer/ protection seller
— Reference entities
— Credit events
— Obligations on which a credit event can be called
— Obligations on which loss settlement can be determined
— Settlement method: cash or physical settlement
— Valuation process (if cash settled)
Summary of the Rating ProcessSummary of the Rating Processä Sponsor/ manager motivations and expertise
examined
ä Probability of credit events determined
ä Likely recoveries in the event of default assessed
ä Structure reviewed:— collateral arrangements
— counterparty risk
ä Performance Analytics
ä Sponsor/ manager motivations and expertise examined
ä Probability of credit events determined
ä Likely recoveries in the event of default assessed
ä Structure reviewed:— collateral arrangements
— counterparty risk
ä Performance Analytics
Portfolio Default ProbabilityPortfolio Default Probabilityä Name-by-name analysis of the portfolio
ä Publicly rated names:
— Fitch rating or lowest public rating
— Market information (CDS spreads), potential downgrades and corporate analyst’s opinion may lead to rating adjustment
ä Default probability
— New criteria effective August 1, 2003
ä Portfolio Diversification
— Correlation analysis
ä Name-by-name analysis of the portfolio
ä Publicly rated names:
— Fitch rating or lowest public rating
— Market information (CDS spreads), potential downgrades and corporate analyst’s opinion may lead to rating adjustment
ä Default probability
— New criteria effective August 1, 2003
ä Portfolio Diversification
— Correlation analysis
Portfolio Default ProbabilityPortfolio Default ProbabilityTop Industry Concentrations
Insurance10.67%
Banking, Finance, and Real Estate10.67%
Building and Materials7.33%
Utilities6.67%
Broadcast and Media6.00%
Energy6.00%
Industrial/Manufacturing5.33%
Automobiles4.67%
Retail4.00%
13 Other Industries20.67%
Telecommunications8.67%
Computers and Electronics9.33%
Reference Entity RatingsRating Number % of Rating Category of Entities Portfolio Factor
‘AAA’ 2 1.33 1.3‘AA+’ — — 2.0‘AA’ 1 0.67 2.3‘AA–’ 5 3.33 3.3‘A+’ 15 10.00 4.0‘A’ 21 14.00 5.0‘A–’ 22 14.67 7.5‘BBB+’ 26 17.33 10.0‘BBB’ 35 23.33 14.0‘BBB–’ 15 10.00 20.0‘BB+’ 8 5.33 37.0‘BB’ — — 43.5‘BB–’ — — 46.5‘B+’ — — 50.0‘B’ — — 52.2‘B–’ — — 65.0‘CCC+’ — — 90.0‘CCC’ or Lower — — 100.0WA Rating Factor — — 11.32WA – Weighted average.
Credit EventsCredit Events
ä Failure to Pay
ä Bankruptcy
ä Restructuring
ä Repudiation/Moratorium
ä Obligation Acceleration
ä Obligation Default
ä Failure to Pay
ä Bankruptcy
ä Restructuring
ä Repudiation/Moratorium
ä Obligation Acceleration
ä Obligation Default
Credit EventsCredit Eventsä Market Practice
— For developed markets– Failure to Pay
– Bankruptcy
– Some form of Restructuring
— For emerging markets and sovereigns– Failure to Pay
– Some form of Restructuring
– Repudiation/Moratorium
ä Market Practice
— For developed markets– Failure to Pay
– Bankruptcy
– Some form of Restructuring
— For emerging markets and sovereigns– Failure to Pay
– Some form of Restructuring
– Repudiation/Moratorium
Credit EventsCredit Events
ä Restructuring
— Old Restructuring
— Mod Restructuring
— Mod-Mod Restructuring
— No Restructuring
ä Restructuring
— Old Restructuring
— Mod Restructuring
— Mod-Mod Restructuring
— No Restructuring
Credit EventsCredit Events
ä Obligation Acceleration or Obligation Default
— Alternative Credit Events
— Used Primarily in “one off” trades for which the events may be relevant
ä Obligation Acceleration or Obligation Default
— Alternative Credit Events
— Used Primarily in “one off” trades for which the events may be relevant
Restructuring and/or Obligation AccelerationRestructuring and/or Obligation Acceleration
Examples: Both the Xerox restructuring (June 2002) and the Conseco restructuring (Oct. 2002)highlighted ‘soft credit event’ risk and ‘cheapest to deliver’ risk
Examples: Both the Xerox restructuring (June 2002) and the Conseco restructuring (Oct. 2002)highlighted ‘soft credit event’ risk and ‘cheapest to deliver’ risk
Since the inclusion of Restructuring and/or Obligation acceleration as a Credit Event increases the probability of loss to investors, Fitch increases Its base probability when rating a synthetic CDO.
Since the inclusion of Restructuring and/or Obligation acceleration as a Credit Event increases the probability of loss to investors, Fitch increases Its base probability when rating a synthetic CDO.
Fitch applies a default adjustment factor to ReferenceEntities whether old-R, mod-R or mod-mod R is included asa Credit Event.
Fitch applies a default adjustment factor to ReferenceEntities whether old-R, mod-R or mod-mod R is included asa Credit Event.
Fitch Rating ImplicationsFitch Rating Implications
Fitch pays close attention to the choice of Credit Events, thescope of protection based Obligation selection and ObligationCharacteristics included in the terms of the CDS.
Fitch pays close attention to the choice of Credit Events, thescope of protection based Obligation selection and ObligationCharacteristics included in the terms of the CDS.
When analyzing the default risk of synthetic CDOs, Fitchwill, when applicable, take account of the Credit Eventsdefined and Obligations specified in the underlying CDS.
When analyzing the default risk of synthetic CDOs, Fitchwill, when applicable, take account of the Credit Eventsdefined and Obligations specified in the underlying CDS.
Loss DeterminationLoss Determination
ä Recovery Rate assumptions
— New criteria effective August 1, 2003
— Loss settlement method (cash versus physical)
— Valuation process (cash settlement)
ä Recovery Rate assumptions
— New criteria effective August 1, 2003
— Loss settlement method (cash versus physical)
— Valuation process (cash settlement)
Convertible, Exchangeable and Accreting ObligationsConvertible, Exchangeable and Accreting Obligations
ä The 2003 Definitions essentially roll in the Convertible, Exchangeable and Accreting Obligations Supplement
ä The 2003 Definitions essentially roll in the Convertible, Exchangeable and Accreting Obligations Supplement
Example: CSFB had refused to accept delivery of Railtrack exchangeable bonds from protection buyer Nomura.
Fitch may apply a haircut to its normal recoveryRate assumption when either convertible bondsand/or consent required loans constitutedeliverable obligations
Fitch may apply a haircut to its normal recoveryRate assumption when either convertible bondsand/or consent required loans constitutedeliverable obligations
Structural FeaturesStructural Featuresä Counterparty exposure
— Achieve structural delinkage (reserve accounts and downgrade triggers)
— Immunize “breakage costs”
ä Collateral arrangements
— Overcollateralization and marking-to-market
— Put option/repurchase agreement
ä Counterparty exposure
— Achieve structural delinkage (reserve accounts and downgrade triggers)
— Immunize “breakage costs”
ä Collateral arrangements
— Overcollateralization and marking-to-market
— Put option/repurchase agreement
SummarySummary
ä Area of growth and development
ä New rating issues raised with innovative structures
ä New data used to refine assumptions
ä New criteria report: ‘Global Rating Criteria for CDOs’
ä Area of growth and development
ä New rating issues raised with innovative structures
ä New data used to refine assumptions
ä New criteria report: ‘Global Rating Criteria for CDOs’
www.fitchratings.com
Rating nth-to-Default Basket CreditLinked Notes
Rating nth-to-Default Basket CreditLinked Notes
Tania Cunningham, CFA - Associate Director
July 16, 2003
Tania Cunningham, CFA - Associate Director
July 16, 2003
Defining the characteristics of an nth-to-default contractDefining the characteristics of an nth-to-default contract
ä Correlation trade
ä Type of OTC credit derivative
ä Payoffs depend on the occurrence of the nth credit event in an underlying basket of bonds
ä Correlation trade
ä Type of OTC credit derivative
ä Payoffs depend on the occurrence of the nth credit event in an underlying basket of bonds
Cash Flows of nth to Default Basket SwapsCash Flows of nth to Default Basket Swaps
Credit Event on nth Asset No Credit Event on nth Asset
TRUST INVESTOR
Proceeds
Notes/Certificates
1 Initial Exchange
TRUST INVESTOR
PAR
2 Early TerminationRecovery amountor physical security
TRUST INVESTOR
Proceeds
Notes/Certificates
1 Initial Exchange
TRUST INVESTOR
Note
2 At Maturity
PAR
Maturity
Closing
Time Line
Types of nth-to-defaultsTypes of nth-to-defaults
First loss pieces in traditional CDOs
First loss pieces in traditional CDOs
First-to-defaultFirst-to-default
Third-to-defaultThird-to-default
Linear basketsLinear baskets
Fifth-to-defaultFifth-to-default
TYPESTYPES
The rating is ultimately based onThe rating is ultimately based on
ä Creditworthiness of individual reference entities
ä Number of reference entities
ä Default correlation between reference entities
ä Number of defaults to trigger payment
ä Term of the deal
ä Credit event language, legal opinions, etc.
ä Creditworthiness of individual reference entities
ä Number of reference entities
ä Default correlation between reference entities
ä Number of defaults to trigger payment
ä Term of the deal
ä Credit event language, legal opinions, etc.
Rating nth to Default Basket SwapsRating nth to Default Basket SwapsRating of 1st to Default Basket
Ratings of the reference entities form the boundaries for the rating of the basket.
§ Upper boundary – weakest credit
§ Lower boundary – sum of the default probabilities of each credit
§ Lower boundary assumes no creditsdefaults at exactly the same time
0 10020 40 60 80
Def
ault
Pro
babi
lity
Correlation %
Ratings (reference entities, swap counterparty and collateral)
Associated default probabilities
Default correlation (reference entities, swapcounterparty and collateral)
Number of reference entities
Term of the deal
Credit event language, legal options, etc.
Number of defaults to trigger payment
Lower Boundary
Upper Boundary
Advantages of Monte CarloAdvantages of Monte Carlo
ä The number of assets— Typically ranges from 2 to 25
ä The number of assets— Typically ranges from 2 to 25
Distorts the results if we use a scenario model based on historical averages.Distorts the results if we use a scenario model based on historical averages.
Advantages of Monte CarloAdvantages of Monte Carlo
ä Correlation between the assets— Plays a crucial role in this type of deal
ä Correlation between the assets— Plays a crucial role in this type of deal
The MC approach provides a fairly simple way to factor in correlation.The MC approach provides a fairly simple way to factor in correlation.
Advantages of Monte CarloAdvantages of Monte Carlo
ä Dynamics of transaction— Captured using MC simulation better than any other method
ä Dynamics of transaction— Captured using MC simulation better than any other method
In particular MC is the only way to model the tail end of the distribution: It is possible to compute the mean and the variance without using a MC approach but not the 99th percentile unless you assume a specific statistical distribution.
In particular MC is the only way to model the tail end of the distribution: It is possible to compute the mean and the variance without using a MC approach but not the 99th percentile unless you assume a specific statistical distribution.
Effect of Correlation: three name first-to-default basket Credit Default SwapsEffect of Correlation: three name first-to-default basket Credit Default Swaps
A B
C
A B
C
A B C
Negative Correlation Low Correlation High Correlation
Effect of CorrelationEffect of Correlation
ä First-to-default transactions benefit from it because it reduces the likelihood of one or more defaults
ä The effect of correlation for all other nth to default transactions is more variable
ä First-to-default transactions benefit from it because it reduces the likelihood of one or more defaults
ä The effect of correlation for all other nth to default transactions is more variable
Effect of CorrelationEffect of Correlation
0%5%
10%20%
30%40%
50%60%
70%80%
90%99%
First-to-DefaultSecond-to-Default
Third-to-DefaultFourth-to-Default
Fifth-to-Default
Sixth-to-Default
Seventh-to-Default
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Def
ault
Pro
bab
iliti
es
Correlation levels
Correlation Chart
First-to-Default
Second-to-Default
Third-to-Default
Fourth-to-Default
Fifth-to-Default
Sixth-to-Default
Seventh-to-Default
The CommitteeThe Committee
ä Results based on a number of correlation assumptions presented
ä Where necessary input sought on individual bi-variate correlations from Fitch corporate analysts
ä Consistency checks by reference to comparable deals
ä US and European attendance on committees is normal
ä Results based on a number of correlation assumptions presented
ä Where necessary input sought on individual bi-variate correlations from Fitch corporate analysts
ä Consistency checks by reference to comparable deals
ä US and European attendance on committees is normal
www.fitchratings.com