CDO - Collateralised Debt Obligation

Synthetic & Cash Structure

December 4-7 , 2006

Per von Rosen, Nordea+46 (0)8 6149339

2

Contents

Introduction to the Securitisation Market (ABS)

Introduction to Collateralised Debt Obligation (CDO)

– Synthetic CDO

– Cash CDO

Summary

Introduction to the Securitisation Market (ABS)

Introduction to Collateralised Debt Obligation (CDO)

– Synthetic CDO

– Cash CDO

Summary

3

ABS - Asset-Backed Securities

The Asset-Backed Securitisation market started in the USA with securitisation of residential mortgages (1970/71).

The market has experiences high growth over the last years.

Asset-Backed Securities (ABS) are bonds secured on a portfolio of financial assets or receivables. ABS involves the repackaging of financial assets into debt securities.

ABS are generally issued by SPV (Special Purpose Vehicle), to delink the credit risk from the originator of the underlying assets.

The underlying credit risk of the portfolio is tranched into rated and unrated classes of notes and equity (each class rating is determined by it’s position in the priority of payments and other criteria's).

The average ABS has higher rating and rating stability than corporate bonds (Moody’s rating survey).

4

There are many different kinds of underlying assets within the securitisation market:

Example of Swedish ABS transactions:

– Europe Loan, SRM Inv (SBAB), Framtiden (City of Gothenburg), Midgaard (Nordea)

ABS - Underlying Assets

Asset-Backed Securities (ABS)Asset-Backed

Securities (ABS)

Trade Receivables

Credit Cards

Auto Loans

Consumer Loans

Collateralised Debt Obligation (CDO)

Collateralised Debt Obligation (CDO)

Residential (RMBS)

Commercial (CMBS)

Mortgage-Backed Securities (MBS)

Mortgage-Backed Securities (MBS)

Loans (CLO)

Bonds (CBO)

Synthetic (CDO)

5

ABS - Transaction Structure

Assets Tranches - Risk

PortfolioOf

Bond/loans

Senior

Mezzanine

Equity

Underlying Assets

Issuer (SPV)

True sales or synthetic through CDS.

Bonds or loans

Aaa

Aa2

Equity

Baa2

Prioritysenioritis

High

Low

Originator

Originator keeps the equity-piece.High risk

Low riskSecuritiesPotentially on Originator’sBalance Sheet.

6

What is a CDO 1 (2)

CDO (Collateralised Debt Obligation) emerged from the evolution of the capital markets in 1990.

A CDO is a security that is collateralised against a pool of financial assets. These assets can be bonds (from investment grade or high yield), loans and Asset-Backed Securities.

CDO involves slicing (tranching) of the credit risk of a portfolio into different classes of notes.

Highly efficient tool for re-allocating credit risk among market participants.

A’ la carte CDO Menu

The performance for a Cash flow CDO depends completely on the collateral cash flow, while a Market Value CDO depends on both the market value changes and the credit performance of the collateral.

Structure Purpose Transfer

Cash Flow CDO Balance Sheet or Arbitrage Synthetic or Cash

Market Value CDO Balance Sheet or Arbitrage Synthetic or Cash

7

What is a CDO 2 (2)

Cash Flow CDO – is a structured finance product that securities a diversified pool of debt into multiple classes of notes from the cash flows generated by the asset pool.

Market Value CDO - is a structured finance product that securities a diversified collateral of debt into multiple classes of notes from trading and sale of the collateral. The credit quality depends upon the managers ability to sell assets and pay-off debt classes principal and accrued interest.

The underlying collateral could either be static or managed.

The manager is the most important participant in any managed CDO and the manager has generally a long demonstrated experience (generally holds some equity in their own CDO).

Cash flow managers main task is to avoid defaults.

Market value managers main task is to achieve price appreciation.

8

CDO - Diversifications and Asset Class

CDOs are governed by diversification guidelines

– Asset Class

– Diversification by issuer

– Industry

– Region

Different Asset Class

– CDOs of High yield and Investment grade bonds (CBO)

– CDOs of Investment grade bank loans (CLO)

– CDOs of Asset-Backed Securities (CDOs of ABS)

– CDOs of Mortgage-Backed Securities (CDOs of MBS)

– CDOs of debt by other securities issued CDOs (CDOs of CDOs)

9

CDO - Asset risk

For a Cash Flow CDO the credit risk of the debt tranches is determined by:

– Default probability

– Default correlation

– Default severity (recovery)

– Attachment points & tranche width

For a Market value CDO the credit risk of the debt tranches is determined by the underlying pool and by the managers ability to sell asset and achieve price appreciations.

10

Synthetic CDO - Introduction

0

100

200

300

400

500

600

700

1995 1996 1997 1998 1999 2000 2001 2002 2003

USD

Bill

ion

Cash CDOs Synthetic CDOs

Global CDO Growth: 1995-2003The first synthetic CDO transaction was done in December 97 by JP Morgan.

In a synthetic CDO the credit risk is transfer through Credit Default Swaps (CDS).

Development of standardised synthetic CDOs in the form of tradable CDS indices (DJ iTRAXX).

Prerequisite – Standardisation

- CDS of IG Corporate- CDS of HY Corporate- CDS of ABS (RMBS/MBS)?

Source: Bank of America

The CDO market has experienced high growth over the last past years

11

Synthetic CDO - Mechanics

Super SeniorTranche

Extremely low Default probability

Credit Portfolio

Mezzanine[AAA – BBB]

Low Default probability

Equity

Highest Default probability

CDOs are tranched into several default probability segments to reflect both investors who are risk averse and those who can accept higher risk with a high return.

In tranching, the nominal risk of the portfolio will be segmented into more risky and less risky layers (Super Senior, Mezzanine, Equity).

Somebody will assume the risk of the first default from the portfolio, some others risk of the next default up to a defined limit and so on (i.e. subordination).

The most junior tranche (Equity or “First loss piece”) resumes the first losses and is generally paying a high return (about 15 - 20%)

The super senior tranche is a vital driver behind the economics of a synthetic CDO (and sometimes of a cash CDO, as well).

The Super-Senior tranche (normally 80-90% of the capital structure) is either kept by the originator or sold as a unfunded note, normal to monolines. It’s generally paying a low marginal (8 - 15 bp).

90%

7%

3%

12

0.0%2.5%5.0%7.5%

10.0%12.5%15.0%17.5%20.0%22.5%

25.0%27.5%30.0%

0.0% 2.3% 4.4% 6.5% 8.6% 10.7% 12.8% 14.9%

Portfolios loss distribution

Synthetic CDO - Tranching 1 (2)

A credit portfolio is sliced into layers with different level of risk and sold to investors with different risk appetite.

The tranching is done in combination with rating agency auditing (rating).

The expected rating for each tranche is a function of the underlying portfolios and each tranche’s expected loss.

The expected loss for the Equity tranche is extremely high.

The expected loss for the Super senior tranche is extremely low.

Super SeniorEquity

Mezza-nine

Exp. Loss%

Pro

babi

lity

%

13

Synthetic CDO - Tranching 2 (2)

The tranching of CDOs allows the investors to leverage their investment and also to take exposure of the default correlations.

A synthetic CDO is either tranched as

– Single tranche CDOs

– Multi tranche CDOs

– Full capital CDO structures

A single tranche CDO is created much more quickly than a full capital CDO structure and could also easier be tailor made to meet investors demand for risk/return profile.

A CDO could also be tranched as Combination notes (Combo notes, are rated as a principal-only rating). This notes are created by combining two or more note classes (usually equity & mezzanine).

There are two types of combo notes; income combo notes (inv. in par capital) and capital enhance combo notes (inv. in par and zcpn capital).

14

Synthetic CDOMulti Tranche Structure – Mezzanine & Equity

• SPVs are commonly used as a issuer platform, CLNs could be used if investors are willing to accept the additional credit risk of the CLN issuer.

NordeaMarketsNordeaMarkets SPVSPV

Delta hedgeDelta hedge

Class A (Aa3)Class A (Aa3)

Class B (Baa1)Class B (Baa1)

Class C (Ba2)Class C (Ba2)

Equity (Nr)Equity (Nr)

Credit risk

Notes (tranche)

High risk

Low risk

Interest/Premium

[3M EUR + 0,70%]

[3M EUR + 1,95%]

[3M EUR + 3,80%]

[3M EUR + 16,5%]

CollateralCollateral

Interest Capital

6,80 - 9,10%

4,60 - 6,80%

2,70 - 4,60%

0,00 - 2,70%

Subordinations

Premium

Prio

rity

seni

ority

CDO transaction (Mermaid 5) arranged & structured by Nordea Markets, October, 2005

15

Synthetic CDO - Pricing & Correlations

The pricing of the different note classes of a CDO is depending of a number of factors (portfolio spreads, attachment, detachment, recovery, correlation, rating and structuring costs etc).

For a full capital structure the average spreads, for all note classes, should approx. equal the average underlying portfolio spreads minus rating and structuring costs.

Correlation plays an important role in the pricing of a synthetic CDO.

CDOs could be view as a correlation product and a natural extension of the First-to-Default (FTD) basket concept.

Correlation (default) – the likelihood of default of one credit heightening the likelihood of default of another credit.

Correlation could either be calculated through internal price models or could be estimated through the Credit Derivative Market (tradable tranched CDS indices, FTD basket or CDOs).

A change in the correlation will have a great impact of the market value for the different tranches.

16

Synthetic CDO - Correlations 1 (2)

Low correlation - there is little chance of high numbers of defaults within the portfolio.

High correlation – there is a great chance of low numbers of defaults within the portfolio.

The more defaults within a portfolio gets correlated, the more the portfolio behaves like a single credit.

How are the different tranches sensitive to default correlations:

– Senior notes are short correlation (sensitive to increase in correlation)

– Mezzanine notes are typical short correlation (sensitive to decrease in correlation)

– Equity are long correlations (sensitive to decrease in correlation)

The equity tranche is the most sensitive tranche for changes in correlations.

CDOs are also sensitive for changes in the underlying credit spreads.

17

Synthetic CDO - Correlations 2 (2)

Portfolio Default Distribution

A shift in correlation will have a large effect on the equity tranche.

As correlations increases toward 100%, the probability that the equity is being wiped out becomes more similar to the probability of the most senior tranche being wiped out.

Source. Fitch Ratings

18

Cash CDO - Introduction

Cash CDOs are transactions that are structured using underlying financial assets that are normally sold by the Originator to a SPV (normally transferred through a true sale).

All payments to the investors depend on the proceeds from the collateral assets (interest, principal payments).

If the underlying assets don’t have a bullet maturity, there will be a uncertain pay down profile.

Cash CDOs are normally paid down through amortisation and a soft bullet.

A Cash CDO requires a period in which the SPV finds the assets to invest into (“ramp-up period”).

Cash CDOs are using a number of different Credit Enhancements (CE) compare to synthetic CDOs.

A cash CDO could be viewed as a “standard alone transaction” issued by an SPV, and there is no recourse to the Originator.

A CDO cannot be impaired by default of any other CDO or the Originator/structure (not valid in case of a synthetic securitisation)

19

Cash CDO - Structure 1 (2)

OriginatorBank

OriginatorBank

Financial AssetsFinancial Assets

Credit DefaultSwap (CDS) SPVSPV

Aaa to Baa Notes

Aaa to Baa Notes

Subordinated Notes

Subordinated Notes

When it’s not possible, or economically preferable, the financial assets are not sold to a Special Purpose Vehicle (SPV) but retained by the originator.

Instead, the credit risk of the underlying assets is transferred using Credit Derivatives, which exactly mirror the payments of the collateral assets (hence, no standard Credit Default Swaps).

Here the risk is sold to an SPV, which then issues bonds to pass the risk onto investors.

Synthetic Securitisation

20

Cash CDO - Structure 2 (2)

OriginatorBank

OriginatorBank

Financial AssetsFinancial Assets

True sales of financial assets SPVSPV

AaaBondsAaa

Bonds

Junior BondsJunior Bonds

Baa1Bonds

True Sales

Baa1Bonds

First Cash flow

Second Cash flow

Residual Cash flow

Transfer

Cash flow

Principal & interest

The financial assets are sold from the originator to a Special Purpose Vehicle (SPV).

The SPV issues bonds to finance the acquisition and to pass the risk of the underlying assets onto investors.

The cash flow from the underlying assets is use for paying interest and principle to investors, accord-ing to priority.

The junior investor will receive the residual cash flow after that all expenses has been paid.

21

Cash CDO - Credit Enhancements

The rating of each note class is a function of the quality of the underlying asset pool, but there are a number of ways to improve the quality of each notes.

Subordination – total par amount of liabilities that are subordinated to a particular note class (amount of loss protection).

Excess spread – interest that are available to the equity investors (after interest payments on the CDOs other liabilities have been made). E.g. the CDO has coupon of 4% but the assets pay a coupon of 5%. The excess spread is to cover any defaults.

Overcollateralisation – excess of the par amount of the collateral available to secure one or more note classes over the par amount of those notes.

Priority of payments (Waterfall) – determine in which way interest, amortisation and principal should be made to each note class. “Structural Enhancement”.

External credit enhancements – guarantee payments by third party insurers (monoline credit insurance companies, “Monoline Wrap”).

Reserves –available for the benefit of the senior notes.

22

Cash CDO - Coverage test

Early amortisations trigger are designed to protect the rated notes of Cash Flow CDO structures, i.e. to maintain a min. level of credit quality and therefore protections for note-holders.

Overcollateralisation Coverage test (OC), require the transaction to maintain a preset min. overcollateralisation ratio for each rated note class.

Interest Coverage test (IC), require the transaction to maintain a min. interest coverage ratio for each rated note class.

Most CDO transactions requires, in the event of the failure of one or more tests, the senior most outstanding class of notes to be paid down first (the notes be repaid sequentially). “Structural Enhancements”.

23

Summary – Synthetic vs. Cash CDO

CashSynthetic

87,50%

3,75%

2,25%2,75%

3,75% 7,00%

Super senior tranche (AAA+)

Senior tranche (AAA)

Mezzanine tranche (Aa2)

Mezzanine tranche (Baa2)

Equity tranche (Nr)

88,00%

2,25%2,75%

Synthetic CDOs are easy to understand, to structure and ramp-up.

Cash CDO are more complex transactions involving a number of different credit enhancements.

Synthetic CDOs are normally bullet transactions and involves a collateral.

Cash CDO are paid down through amortisation and a soft bullet (uncertain regarding Weighted Average Life, “WAL”)

Synthetic CDO allows investors to take exposure on correlations risk.

Cash CDOs are using “actual” default triggers compare with synthetic CDOs that are using defined credit events (according to the CDS market).

24

Summary - Benefits of buying into CDOs

CDOs offer enhanced return compare with the cash market.

Diversification through access to market segment that are not easily accessible.

CDO offers exposure to an asset with low correlation to other securities such as vanilla bonds or equities.

CDOs represent the ultimate instrument for earning liquidity premium and to access a diversified pools of credit risk.

Leverage/Deleverage. CDO technique can be used for creating deleverage investment or to add leverage for higher return expectations.

CDO is flexible and can be designed to satisfy a specific appetite of investor (risk/return profile).

Returns in the CDO market are considerably higher than in other securitisation and cash instrument, with a long track records.

25

Default Performance and Rating Volatility - ABS versus Corporates

OriginalRating Corporates ABS Corporates ABS Corporates ABSAaa 0.14% 0.00% 15% 0.28% NA NAAa 0.36% 0.00% 15.19% 1.51% 3.40% 1.95%A 0.50% 0.00% 12.35% 1.14% 7.65% 3.19%Baa 1.67% 0.20% 12.13% 6.59% 12.38% 1.04%

Cumulative Cumulative Cumulative Default Percentage Downgrade Percentage Upgrade Percentage

Source: Moody's Investors Service, Feb 2002

Risk comparison of ABSs vs Corporate Bonds

26

Moody’s: The MARVEL example

Found at www.moodys.com

Step 1 – we input the expected loss and the uncertainty of this estimate into the model found from historical and market data

Step 1 - Input details of the Assets

AssetsExpected Loss 5.00%Standard Deviation 1.44%

27

Moody’s MARVEL example cont’

We specify the detailsStep 2 - Input the Credit Enhancement for the Notes being issued

External credit enhancementCash Reserve 2.00%Captured Excess Spread 2.00%Total 4.00%

Total credit enhancementEnhancement by

Tranche Average Class Specific Subordinated External Credit TotalTranche Size Life (years) Enhancement Tranches Enhancement Enhancement

- A 90.30% 2 9.7% 4.00% 13.7%- B 4.40% 3.5 5.3% 4.00% 9.3%- C 2.70% 3.7 2.6% 4.00% 6.6%- D 2.60% 3.8 4.00% 4.0%Total (should be 100%) 100%

28

Moody’s MARVEL example cont’

We read off the results from the Moody’s lognormal model, i.e. the implied Moody’s rating expected by us in this example.

Step 3 - Read off the Notes' results

Actual loss NoteTranche (as a % of tranche) Results

- A 0.000110% Aaa- B 0.209798% A3+- C 5.085894% Ba3- D 44.199504% Ca+

29

Moody’s MARVEL example cont’The probability distribution of the expected loss

0.00%

0.05%

0.10%

0.15%

0.20%

0.25%

0.30%

0.35%

0.0% 2.5% 5.0% 7.5% 10.0% 12.5% 15.0% 17.5% 20.0% 32.5% 45.0%

30

Moody’s MARVEL example cont’

Again, all this is a zero sum game apart from the credit enhancements being added by the originator to obtain a balance sheet relief.

If we take away the credit enhancements we have

EXCESS RETURN≡EXPECTED LOSS

The credit enhancement is spread out evenly to lower the expected loss and thus increasing the rating implying a lower spread offering.

31

Source: ”The Binomial Expansion Method Applied to CBO/CLO Analysis”, December 13, 1996, Moody’s Investor Services

Rating 1 2 3 4 5 6 7 8 9 10

Aaa 0,00005% 0,00020% 0,00071% 0,00180% 0,00291% 0,00400% 0,00520% 0,00660% 0,00820% 0,01000%Aa1 0,00057% 0,00300% 0,01000% 0,02100% 0,03100% 0,04200% 0,05400% 0,06700% 0,08200% 0,10000%Aa2 0,00136% 0,00800% 0,02600% 0,04700% 0,06800% 0,08900% 0,11100% 0,13500% 0,16400% 0,20000%Aa3 0,00302% 0,01900% 0,05900% 0,10091% 0,14200% 0,18300% 0,22700% 0,27200% 0,32700% 0,40000%A1 0,00581% 0,03700% 0,11700% 0,18900% 0,26100% 0,33000% 0,40600% 0,48000% 0,57300% 0,70000%A2 0,01087% 0,07000% 0,22200% 0,34500% 0,46700% 0,58300% 0,71000% 0,82900% 0,98200% 1,20000%A3 0,03885% 0,15000% 0,36000% 0,54000% 0,73000% 0,91000% 1,11000% 1,30000% 1,52000% 1,80000%Baa1 0,09000% 0,28000% 0,56000% 0,83000% 1,10000% 1,37000% 1,67000% 1,97000% 2,27000% 2,60000%Baa2 0,17000% 0,47000% 0,83000% 1,20000% 1,58000% 1,97000% 2,41000% 2,85000% 3,24000% 3,60000%Baa3 0,42000% 1,05000% 1,71000% 2,38000% 3,05000% 3,70000% 4,33000% 4,97000% 5,57000% 6,10000%Ba1 0,87000% 2,02000% 3,13000% 4,20000% 5,28000% 6,25000% 7,06000% 7,89000% 8,69000% 9,40000%Ba2 1,56000% 3,47000% 5,18000% 6,80000% 8,41000% 9,77000% 10,70000% 11,66000% 12,65000% 13,50000%Ba3 2,81000% 5,51000% 7,87000% 9,79000% 11,86000% 13,49000% 14,62000% 15,71000% 16,71000% 17,66000%B1 4,68000% 8,38000% 11,58000% 13,85000% 16,12000% 17,89000% 19,13000% 20,23000% 21,24000% 22,20000%B2 7,16000% 11,67000% 15,55000% 18,13000% 20,71000% 22,65000% 24,01000% 25,15000% 26,22000% 27,20000%B3 11,62000% 16,61000% 21,03000% 24,04000% 27,05000% 29,20000% 31,00000% 32,58000% 33,78000% 34,90000%Caa 26,00000% 32,50000% 39,00000% 43,88000% 48,75000% 52,00000% 55,25000% 58,50000% 61,75000% 65,00000%

Appendix: Moody’s Idealized Probability of Default Rates

32

Rating 1 2 3 4 5 6 7 8 9 10

Aaa 0,00003% 0,00011% 0,00039% 0,00099% 0,00160% 0,00220% 0,00286% 0,00353% 0,00451% 0,00550%Aa1 0,00031% 0,00165% 0,00550% 0,01155% 0,01705% 0,02310% 0,02970% 0,03685% 0,04510% 0,05500%Aa2 0,00075% 0,00440% 0,01430% 0,02585% 0,03740% 0,04850% 0,06105% 0,07425% 0,09020% 0,11000%Aa3 0,00161% 0,01045% 0,03245% 0,05555% 0,07810% 0,10065% 0,12485% 0,14960% 0,17985% 0,22000%A1 0,00320% 0,02035% 0,06435% 0,10395% 0,14355% 0,18150% 0,22330% 0,26400% 0,31515% 0,38500%A2 0,00598% 0,03850% 0,12210% 0,18975% 0,25685% 0,32065% 0,39050% 0,45595% 0,54010% 0,66000%A3 0,02137% 0,08250% 0,19800% 0,29700% 0,40150% 0,50050% 0,61050% 0,71500% 0,83600% 0,99000%Baa1 0,04950% 0,15400% 0,30800% 0,45650% 0,60500% 0,75350% 0,91850% 1,08350% 1,24850% 1,43000%Baa2 0,09350% 0,25850% 0,45650% 0,66000% 0,86900% 1,08350% 1,32550% 1,56750% 1,78200% 1,98000%Baa3 0,23100% 0,57750% 0,94050% 1,30900% 1,67750% 2,03500% 2,38150% 2,73350% 3,06350% 3,35500%Ba1 0,47850% 1,11100% 1,72150% 2,31000% 2,90400% 3,43750% 3,88300% 4,33950% 4,77950% 5,17000%Ba2 0,85800% 1,90850% 2,84900% 3,74000% 4,62550% 5,37350% 5,88500% 6,41300% 6,95750% 7,42500%Ba3 1,54550% 3,03050% 4,32850% 5,38450% 6,52300% 7,41950% 8,04100% 8,64050% 9,19050% 9,71300%B1 2,57400% 4,60900% 6,36900% 7,61750% 8,86600% 9,83950% 10,52150% 11,12650% 11,68200% 12,21000%B2 3,98000% 6,41850% 8,55250% 9,97150% 11,39050% 12,45750% 13,20550% 13,83250% 14,42100% 14,96000%B3 6,39100% 9,13550% 11,56650% 13,22200% 14,87750% 16,06000% 17,05000% 17,91900% 18,57900% 19,19500%Caa 14,30000% 17,87500% 21,45000% 24,13400% 26,81250% 28,60000% 30,38750% 32,17500% 33,96250% 35,75000%

Source: The Binomial Expansion Method Applied to CBO/CLO Analysis, December 13, 1996, Moody’s Investor Services

Appendix: Moody’s Idealized Cumulative Expected Loss Rates

33

Using historical and market data we obtain expected default rates

These default rates are magnified by stressfactors given by S&P

AAA – 5, AA – 4 , A – 3, BBB – 2 etc

Stress ratio=Default ratio*Stressfactor

The stress ratio determines the rating

Appendix: S&P and Fitch old models uses stresstestingof the assets

34

Appendix: S&P – the new CDO evaluator

Uses monte carlo simulation

http://www.standardandpoors.com/emarketing/structuredfinance/copyof111201_evaluator.html

35

Appendix: Fitch – the new model

The Derivative Fitch Default VECTOR model

http://www.derivativefitch.com/vector.cfm

36

Moody’s: Uses statistical models

Marvel – lognormal example

Binominal Expansion Model(BET) or Monte Carlo(MC)

On the frontier in research – investigating using different copulas, i.e. concerning the correlation problem in assets