LB - Abs Cdo Primer

ABS CDOs – A Primer
November 17, 2006 OVERVIEW
ABS CDO STRATEGY The ABS Collateralized Debt Obligation (CDO) market came into being in late 1999 and
has grown exponentially since then. Issuance in 2006 will exceed $150 billion, doubling
Michael Koss issuance in 2005, bringing total outstanding to more than $300 billion. We expect
212-526-8312 issuance to continue to grow briskly, as new assets get securitized and the synthetic ABS
mkoss@lehman.com
market expands over time. Almost by definition, ABS CDOs are complex, adding
Dan Mingelgrin structural complexity to underlying assets that are often themselves quite complex. As
212-526-7764 with any securitized sector, a better understanding of structure, collateral, and risks can
dmingelg@lehman.com help uncover value in the market. Our primer is designed to help investors increase their
understanding of these facets of ABS CDOs; in addition, we present a brief history of the
CASH CDO STRATEGY
ABS CDO market and discuss current trends therein. We focus in particular detail on the
risk factors and valuation considerations that investors face today.
Claude A. Laberge
212-526-5450
This report covers the following major areas:
claberge@lehman.com • What is an ABS CDO?
Lorraine Fan • Role of the Collateral Manager
212-526-1929
lfan@lehman.com • A Brief History of the ABS CDO Market
• Collateral Performance
STRUCTURED CREDIT
STRATEGY • Structure Features
Ashish Shah
• Valuation Considerations
212-526-9360 • Risk Factors
asshah@lehman.com
Gaurav Tejwani
212-526-4484
gtejwani@lehman.com
Bradley Rogoff
212-526-7705
brrogoff@lehman.com
Vikas Chelluka
212-526-3364
vchelluk@lehman.com
Ashish Keyal
44-20-710-29037
askeyal@lehman.com
PLEASE SEE ANALYST CERTIFICATIONS AND IMPORTANT DISCLOSURES STARTING AFTER PAGE 27
Lehman Brothers | ABS CDOs – A Primer
Table of Contents
What Is an ABS CDO? 3
Characteristics of CDOs 3
The Importance of CDOs 5
CDO Economics 101 5
Role of the Collateral Manager 8
Portfolio Selection 8
Eligibility Criteria 8
Surveillance 8
A Brief History of the ABS CDO Market 9
Issuance Trends 9
Diversified versus Real Estate Era 10
Spread History 11
Collateral Performance 11
Rating Transitions 11
Upgrade/Downgrade Statistics 12
Structural Features 13
Credit Enhancement 13
Coverage Tests 14
Priority of Payments 15
Optional Redemption and Auction Calls 16
Additional Features 17
Valuation Considerations 20
Rating Agency and Market Pricing and Stress Assumptions 20

An ABS Loan-Level Approach to Losses and Prepayments 20
Risk Factors 22
Credit 22
Basis Risk on Fixed-Rate Security Hedges 23
Available Funds Cap Risk (AFC) 23
Call Risk/Extension Risk 23
Conclusion 24
Glossary 25
November 17, 2006 2

What is an ABS CDO?

A collateralized debt obligation (CDO) is a bankrupcy remote special-purpose entity that
purchases a pool of securities and funds the purchase of those securities by issuing debt
liabilities collateralized by the underlying securities. The use of CDOs as a technology,
rather than an asset class, increases the universe of assets that may be purchased for a
CDO. Thus far, these asset classes include, but have not been limited to, leverage loans,
high grade credit, high yield credit, and ABS. In this report, we focus specifically on the
ABS CDO market, which now comprises about 65% of global CDO issuance.
Characteristics of CDOs
Before getting into the finer details of CDOs, some general understanding of CDO
characteristics and definitions will be necessary (also see Glossary).
• Collateral Ratings: The ABS CDO market may be divided into two general
categories, high grade and mezzanine. The assets in a high grade transaction tend to
have a single-A rating or higher, whereas assets in a mezzanine transaction tend to
be predominantly triple-B rated. The average high grade transactions issued in 2003-
2005 had 34% AAA, 39% AA, and 23% A exposure (Figure 1). 1 While mezzanine
transactions are composed of predominantly BBB (65%) assets, they also have
diversification of ratings in AAA (3%), AA (5%), A (19%), and BB (7%) assets. On
average, high grade and mezzanine transactions have a weighted average rating
factor (WARF) of 49 (Aa3/A1) and 413 (Baa2/Baa3), respectively (see Glossary for
more on WARF).
• Collateral Types: The breakdown of collateral types in 2003-2005 transactions is
heavily weighted to real estate through the purchase of HEL, RMBS, CMBS, and
CDOs. Digging deeper into HG and mezzanine collateral highlights some
differences (Figure 2). High grade transactions have less HEL exposure than
mezzanine transactions (52% vs. 60%), but have higher CDO 2 exposure (18% vs.
7%) (see our report on “ABS CDO exposure to CDOs of ABS” in the ABS Strategy
Weekly, August 25, 2006). RMBS exposure is similar to mezzanine exposure.
• Floating-rate and Fixed-rate Collateral: ABS CDOs purchase floating-rate and
fixed-rate securities for the portfolio. Since the CDO liabilities tend to be floating-
rate, the purchase of fixed-rate securities as collateral tends to be limited in order to
minimize basis risk. The fixed-rate concentration typically ranges from 0% to 30%.
Currently, ABS CDOs attempt to mitigate the cash flow sensitivity to interest rate
movements by purchasing amortizing swaps and caps, and in some circumstances
allowing for future reinvestment in fixed-rate collateral (see our report on “Fixed-
Rate Collateral Risk in HG ABS CDOs” in the ABS Strategy Weekly, June 2, 2006).
• Diversification: CDOs purchase collateral diversified by issuer, collateral type, and
rating. The level of diversification is one of the drivers of rating agency models
when determining the ratings of the debt liabilities.
• Managed and static: CDOs have the ability to buy and sell securities after the
closing date. Managed transactions are permitted to buy or sell assets and reinvest
principal payments from the assets. The collateral manager may buy and sell assets
1
This is an average of about 100 transactions. We assume that a portfolio with a BBB+ average rating or lower is
mezzanine (WARF of 260 or higher), and A- and higher is high grade.
2
Includes ABS CDOs, CLOs, HY CBOs, etc.
November 17, 2006 3

during the reinvestment period, which is typically the first three to five years of the
transaction’s life. In a static transaction, asset sales/purchases are typically not
permitted after the initial portfolio selection has occurred. The transaction’s portfolio
advisor performs the initial portfolio selection and any ongoing monitoring that must
be done. Some static transactions will allow the portfolio advisor discretionary sales
of assets that have appreciated or depreciated in value. “Lightly managed” or
“substitution” transactions permit some trading flexibility, but less than a typical
managed transaction. Going forward, we will use the term “collateral manager”
when referring to the manager or portfolio advisor of a managed and static
transaction, respectively.
• Cash versus Hybrid/Synthetic: Cash CDOs have assets that are fully funded cash
securities and pay a bond coupon (fixed or floating). Hybrid CDOs have a portion of
the assets in cash form and the remaining portion in synthetic form (which pays a
fixed premium). The synthetic portion is typically in the form of a credit default
swap or total return swap. In both cash and hybrid/synthetic transactions, mark-to-
market volatility of the underlying instrument does not directly impact the
transaction. On the other hand, market-value CDOs, another structure used in the
marketplace, may be affected by mark-to-market volatility if certain market value
triggers are breached. Although this is an important subset of the market, we will be
focusing mainly on static and managed cash/hybrid transactions without the market
value features throughout this primer.
Figure 1. ABS CDO Rating Composition Figure 2. ABS CDO Collateral Composition
100%
Rating High Grade Mezzanine
AAA 34% 3% 80%
AA 39 5
A 23 19 60%
BBB 3 65
BB 0 7 40%
WARF 49 (Aa3/A1) 413 (Baa2/Baa3)
20%
0%
High Grade Mezzanine
HEL CDO RMBS Other
Source: Lehman Brothers, Intex data. 2003-2005 transactions. Source: Lehman Brothers, Intex data. 2003-2005 transactions.
November 17, 2006 4

The Importance of CDOs
CDOs serve a variety of functions within the capital markets. We describe a few of the
more important functions:
• Easy access to a portfolio of assets: CDOs provide investors the ability to gain
exposure to a portfolio of assets without doing all the heavy lifting that typically
accompanies that task. For instance, an investor does not have to go through the
process and large expenditure of building infrastructure to monitor performance,
sourcing the bonds on their own, or having specific in-house expertise for each
collateral type.
• Leveraged return on investment: In the traditional sense, CDOs were originally
conceived to provide an investor a leverage return on a portfolio of assets. Through
the purchase of equity in a transaction, the equity investor can gain leveraged
exposure to 100% of the assets, while only dedicating 1%-6% of the capital.
• Assets under management: For traditional money managers, insurance companies,
and hedge funds, being the manager of a CDO provides relatively stable fee income
on a pool of assets which are effectively “closed-end” funds for anywhere from three
to eight years.
• Term Financing: Some issuers, particularly hedge funds, issue CDOs to term fund
their assets rather than rely on the repo market for ongoing financing. This allows
the issuer to lock in a financing rate for the term of the CDO rather than run the risk
of monthly changes in their repo rates and margin calls.
• Arbitrage Vehicles: We will discuss this in more detail in the next section, but a
CDO may also be thought of as an arbitrage vehicle. CDOs will typically be issued
when the average spread of the assets less the average cost of liabilities is high
enough to produce an attractive return to the equityholder.
CDO Economics 101
So how is the CDO able to purchase assets and make payments to liabilities and manager
fees and still have money left to pay the equityholders? Or more simply put, how is there
an arbitrage in the first place and all investing parties are still satisfied? Let us first use a
simplified example of the economics behind the CDO arbitrage and then look at the
factors that drive it.
We begin with a typical example of a managed mezzanine ABS CDO capital structure
diagram (Figures 3). We will come back to this example throughout the primer. The
CDO contains $1 billion in ABS assets and has issued CDO liabilities in the form of a
class A (AAA-rated), class B (AA-rated), class C (A-rated), class D (BBB-rated), and
equity 3. The assets are assumed to generate an average interest payment of L+180 bp.
Figure 3 provides additional information regarding the class sizes and their stated
coupon.
3
We chose not to further tranche the class A in this example for now, but it is important to note that the class A may
be split into “super-senior” and “mezzanine” AAA-rated securities, where the super-senior is senior in the waterfall
to the mezzanine AAA to the extent losses are above the class A level. In other words, if losses were to reach the
class A level, the mezzanine AAA tranche would suffer 100% loss before the super-senior tranche experienced its
first dollar of loss. In some high grade transactions, the super-senior is funded through the commercial paper (CP)
market rather than through the term market. The CP is rolled on a periodic basis (1-6 months) and may experience
liability spread tightening or widening over the life of the transaction.
November 17, 2006 5

To estimate the CDO arbitrage, we calculate the gross excess spread as the difference
between the average asset spread and the average liability spread. For example, Figure 4
looks at a simplistic CDO arbitrage calculation where the gross excess spread of 131 bp
comes from subtracting the average liability spread of 49 bp from the average asset
spread of 180 bp. Further subtracting transaction fees such as manager fees (25 bp) and
administrative costs (trustee and rating agency fees; 2 bp) from the gross excess spread
leaves us with the net excess spread (104 bp), or CDO arbitrage. We approximate the
equity internal rate of return (IRR) by multiplying the net excess spread (104 bp) by the
equity leverage (20x) and get a gross 20.8% IRR to the equity 4.
Having reviewed the economics behind the CDO arbitrage, we can look at why the
arbitrage exists in the first place.
• Liability Spreads: The tighter pricing of liabilities relative to the spread on the
collateral is a function of a number of factors. First, the ability to credit tranche the
capital structure from AAA to equity classes allows for more efficient distribution of
the underlying risk to investors with varying risk profiles. Second, as discussed
earlier, CDO investors value the ability to easily source ABS portfolios without
having to have the expertise or infrastructure to purchase and monitor securities.
Third, investors value the collateral manager’s ability to manage the deal over time
and/or their security selection expertise.
• Equity Returns: The equity hurdle rate, or the minimum equity return required by
investors to participate in a transaction, helps determine what the CDO arbitrage
level must be at any given point in time. An investor’s hurdle rate will depend on the
perceived risk, attractiveness of returns versus other benchmarks, and so on. It may
also depend on the motivation behind issuing a CDO. For example, as discussed
earlier, hedge funds using CDOs to lock in term financing might retain the equity
and could be willing to retain it at IRRs that are lower than what might be acceptable
to a different investor.
• Asset Spreads: As long as asset spreads are high enough to support the liability cost
(plus other fees and costs) and meet equity hurdle rates, a transaction may be issued.
When CDOs are the marginal buyers of the specific assets (as is often the case with
BBB HELs, for example), the asset spreads will adjust to ensure the arbitrage is high
enough (see our report, “ABS CDO Demand: Outlook and Implications for HELs”
of July 21, 2006). However, when CDOs are not the marginal buyer of the specific
asset, the asset may or may not be desirable for the transaction and will heavily
depend on the overall spread of the asset and whether it meets CDO investor and
rating agency constraints. For example, CDOs might purchase BBB CMBS
securities at tighter spreads than BBB HELs even though doing so reduces the
average asset spread of the transaction. It might still make sense to purchase the
security if it provides more diversification and less correlation within rating agency
models (makes the capital structure more efficient) and/or investors are willing to
buy CDO liabilities at tighter spread levels (for greater diversification, for example).
4
Note: this is a simplification of the equity IRR calculation. In practice, we would need to consider the changes in
liability costs over time and deleveraging of the equity. Additionally, this assumes no losses and does not account
for upfront underwriting fees and hedging fees.
November 17, 2006 6

Figure 3. Typical Mezzanine ABS CDO Capital Structure
ABS CDO Assets ABS CDO Liabilities
Class A (80%)
AAA
ABS Assets 3L+32 bp
Ongoing P&I
$1,000,000,000
Average Coupon of Assets

L+180 bp
Issuance
Proceeds Class B (8%)
AA
3L+55 bp
Class C (2%)
A
3L+135 bp
Class D (5%)
BBB
3L+325 bp
Equity (5%)
Source: Lehman Brothers
Figure 4. Simple CDO Arbitrage Calculation

Average Asset Spread (bp) 180
Average Cost of Liabilities (bp) 49
Gross Excess Spread (bp) 131
Collateral manager Fees (bp) 25

Trustee, Admin, Rating Agency Fees (bp) 2
CDO Arbitrage (Net Excess Spread) (bp) 104
Equity Leverage (=100%/5%) 20
Approximate Gross Equity IRR (%) (=1.04%*20) 20.8
November 17, 2006 7

Role of the Collateral Manager

The collateral manager has a number of roles. To name just a few, the collateral manager
selects the credits in the portfolio, manages to certain eligibility criteria, and conducts
ongoing surveillance of the transaction.
Portfolio Selection
The collateral manager is first and foremost responsible for selecting the initial portfolio
of assets. Historically, the “ramp” period for selecting and purchasing assets for a CDO
ranged from eight to 12 months. With the advent of single-name CDS, the ramp period
has been dramatically shortened for synthetic/hybrid mezzanine transactions, as the
majority of the portfolio may be purchased in a matter of weeks. High grade transactions
still have extended ramp periods because there is limited availability of single-name CDS
referencing HG bonds. As a result, HG managers are constrained to purchasing assets
through the new-issue and secondary cash markets.
The initial portfolio selection is perhaps the most important job of the collateral manager.
Avoiding securities from the outset that have a high likelihood of being downgraded or
incurring losses can help avoid discount sales of assets in the future. Future sales of
discounted assets can have ramifications for the equityholders and could also reduce
overcollateralization and interest coverage levels, thereby increasing the likelihood of
trigger failures (see sections on credit enhancement and coverage tests below).
Eligibility Criteria
Managed transactions have security eligibility criteria which must be met during the
initial selection of the portfolio and when managing the portfolio during the reinvestment
period. The eligibility criteria are typically established in order to provide some
boundaries for the manager in terms of what they can and cannot purchase. The criteria
incorporate maximum and minimum composition requirements pertaining, but not
limited, to:
• Rating of securities
• Fixed-rate securities
• Specified types of securities (e.g., RMBS, CMBS, equipment lease)
• Average life of securities
• Weighted average life of the portfolio
• PIKable securities, and more
In general, the purchase or sale of assets during the reinvestment period may be
permitted as long as the eligibility criteria are satisfied, or if the criteria are not satisfied,
to the extent the criteria are “maintained or improved.”
Surveillance
An important role of the collateral manager is to provide ongoing surveillance for the life
of the transaction. This may incorporate surveillance of the underlying securities as well
as overall transaction performance. To the extent managers are able to spot assets that are
underperforming and may be susceptible to future downgrades and writedowns,
replacing them within the transaction might be warranted. Of course, the earlier a
manager can catch underperformance, the better off the trust will be when selling the
asset, making strong surveillance an important aspect of managing transactions.
November 17, 2006 8

A Brief History of the ABS CDO Market

Issuance Trends
The ABS CDO market came into being in late 1999 and has experienced exponential
growth since. Today, we estimate the total ABS CDO amounts outstanding to be more
than $300 billion. Initially, the ABS CDO market comprised mostly mezzanine
transactions (Figure 5). From 2000 to 2003, only $4.5 billion of high grade transactions
priced, compared to about $25 billion mezzanine transactions. However, in 2004,
issuance of high grade transactions surpassed mezzanine transaction issuance, a trend
that continues today.
The growth in ABS CDO issuance may be attributed to a number of factors:

• Underlying ABS market growth: As the underlying ABS markets have grown,
especially the subprime mortgage market, CDO vehicles have had more than enough
assets to choose from.
• Low default history: Given the low number of bond defaults experienced in the ABS
markets, investors have sought to increase their ABS exposure through CDOs.
• Synthetics market: The growth of the single-name CDS market for subprime and
CDOs provided additional supply, helping to fuel issuance of mezzanine
transactions.
• Growing investor base: The growth in investor sponsorship for the asset class across
the capital structure has created more demand for CDO securities and equity.
• Search for yield: As spreads across a variety of asset classes have compressed in
recent years, investors looking for incremental spread have turned to the ABS CDO
market.
• Increase in secondary market liquidity: Liquidity within the secondary market has
increased in recent years as more dealers became active, providing increased
comfort to investors.
Figure 5. ABS CDO Issuance
$bln
80
60
40
20
0
2000 2001 2002 2003 2004 2005 2006 YTD
Issuance Year
Source: Lehman Brothers. Through 9/30/06.
November 17, 2006 9

Diversified versus Real Estate Era
When ABS CDOs were first issued in 1999, transactions were diversified across a
number of securitized sectors, spanning from credit cards and home equity loans to
aircraft-leasing securitizations (Figure 6). The average transaction pre-2003 had 37%
HEL, 24% CDO 5, 10% CMBS, and 29% “other” exposures (Figure 7a). After a
tumultuous period in late 2002 and early 2003, when the ABS CDO market endured
distress in sectors to which it was heavily exposed, including manufactured housing,
aircraft, and franchise loans, the CDO market emerged with an entirely different look and
purpose. Rather than staying diversified across sectors, issuers gravitated toward real
estate (e.g., subprime, resi-A, and CMBS) sectors. In stark contrast to pre-2003, the
average transaction in 2003-2005 had 55% HEL, 19% RMBS, 14% CDO, 5% CMBS,
and only 7% “other” exposures (Figures 7b).
Figure 6. ABS CDO Timeline
Diversified Era Real-Estate Era

ABS CDO market
recovers after
Aircraft ABS Multitude of severe downgrades High Grade Single-Name "Hybrid" CDOs
incurs downgrades in in underlying CDO market CDS is become the
First ABS significant manufactured sectors, emerges begins its standardized in norm in
CDO issued downgrades housing (e.g. with more exponential late-2005 mezzanine
in late-1999 and losses Conseco residential growth
following 9/11 Finance, mortgage
Oakwood, etc.) concentration
2000 2001 2002 2003 2004 2005 2006 2007
Figure 7a. Diversified Era (2000-2002) Figure 7b. Real Estate Era (2003-2005)
Other
CMBS
7%
Other 5%
29% HEL
37% CDO
14%
HEL
55%
CMBS
10% RMBS RMBS
0% 19%
CDO
24%
Source: Lehman Brothers, Intex data Source: Lehman Brothers, Intex data
5
Includes ABS CDOs, CLOs, HY CBOs, etc.
November 17, 2006 10

Spread History
Following the market disruption in late 2002, increased sponsorship by investors for
ABS CDOs led to spread tightening across the capital structure (Figure 8). Still, the ABS
CDO sector has traditionally offered wider spreads compared to other spread sectors
such as subprime, RMBS, CMBS, and CLOs (Figure 9). This is partly due to the
complex nature of the structures and collateral, less transparency in prices of the
underlying securities, less liquidity in the secondary market, and greater leverage to
weakening credit.
Collateral Performance
Rating Transitions
As an overall asset class, structured finance (inclusive of ABS, CMBS, RMBS, and
CDOs) has been a steady outperformer compared to corporates (when measured by
rating stability). Specifically, global structured finance AAA-, AA-, and A-rated
securities have experienced lower downgrade ratios than similarly rated global
corporates (Figure 10). Triple-B structured finance securities, on the other hand, have
experienced slightly higher downgrades than corporates (6.2% vs. 6.1%).
Figure 8. ABS CDO Spread History
bp bp
160 400
120 300
80 200
40 100
0 0
Jun-00 Jun-01 Jun-02 Jun-03 Jun-04 Jun-05 Jun-06
AAA AA A (RHS) BBB (RHS)
Source: Lehman Brothers. Data averages high grade and mezzanine spreads.
Figure 9. Floating-Rate Spread Comparison, bp

Rating ABS CDOs HEL RMBS CMBS CLO
A 140 42 40 36 72
BBB 325 110 100 80 150
Source: Lehman Brothers. As of 9/30/06.

Upgrade/Downgrade Statistics
While overall structured finance rating stability has been strong, a closer look at
subsectors found in more recent ABS CDOs may show mixed results (Figure 11). RMBS
and CMBS experienced very strong upgrade/downgrade ratios throughout their history,
while HELs and CDOs experienced more downgrades than upgrades. CDOs stand out as
the worst performing sector by far, with an 11.8 downgrade/upgrade ratio since 1996.
This is mostly owing to the poor performance of HY CBOs, which experienced
significant downgrades up until 2005, as well as many ABS CDOs issued in 2000-2002.
Figure 10. Average Annual Rating Transition Matrices
Original Rating Aaa Aa A Baa Upgrade Stable Downgrade

Aaa – SF 98.7 0.9 0.2 0.1 0.0 98.7 1.3
Aaa – Corp 92.6 7.1 0.3 0.0 92.6 7.4
Aaa – Diff 6.1 -6.2 -0.1 0.1 0.0 6.1 -6.1
Aa – SF 4.9 91.6 2.4 0.7 4.9 91.6 3.5
Aa – Corp 0.9 91.2 7.6 0.2 0.9 91.2 7.9
Aa – Diff 4 0.4 -5.2 0.5 4.0 0.4 -4.4
A – SF 1 3.1 92.4 2.3 4.1 92.4 3.5
A – Corp 0.1 2.6 91.1 5.4 2.7 91.1 6.2
A – Diff 0.9 0.5 1.3 -3.1 1.4 1.3 -2.7
Baa – SF 0.4 0.5 2.4 90.5 3.3 90.5 6.2
Baa – Corp 0.1 0.3 5.2 88.3 5.6 88.3 6.1
Baa – Diff 0.3 0.2 -2.8 2.2 -2.3 2.2 0.1
Source: Moody’s. The 1-yr rating transitions for each of the years 1984-2004 are averaged to compute the numbers
shown. The 1-yr rating transition records any transition from the current rating at the beginning of the year for all
outstanding securities to the final rating at the end of the 1-yr period. Each year’s transition numbers are then
weighted by the outstanding number of rated bonds at the beginning of the given year. Half of the total withdrawn
ratings for the given year are subtracted from the outstanding number of rated bonds for weighting purposes.
Figure 11. Upgrade/Downgrade Statistics by Asset Class
Sector Category 2005 2004 1996-2005

HEL Downgrade Rate 1.8% 2.0% 2.1%
Upgrade Rate 2.0% 1.5% 1.7%
Downgrade/Upgrade ratio 0.9 1.4 1.3
RMBS Downgrade Rate 0.9% 0.1% 1.2%
CDOs Downgrade Rate 3.0% 5.6% 9.7%
CMBS Downgrade Rate 3.5% 5.7% 3.6%
Upgrade Rate 16.4% 8.8% 8.7%
Source: Moody’s Investors Service

Structural Features
In this section, we discuss a number of structural features common to many ABS CDOs.
• Credit enhancement: The senior/subordinate capital structure, overcollateralization,
and excess spread provide the majority of protection to the bondholders.
• Coverage tests: In traditional ABS CDO structures, additional protection may come
from coverage tests. Passing or failing the overcollateralization coverage tests (OC)
or interest coverage tests (IC) will affect how cash flows are allocated to the
liabilityholders.
• Priority of Payments: A transaction has payment rules that determine how cash flow
is allocated to liabilityholders. The priority of payments will typically be determined
by the deal age, status of performance triggers, and how quickly the collateral
balance factors down.
• Optional Redemption and Auction Calls: Transactions are issued with call features,
allowing for the liquidation of the assets and return of principal to bondholders and
equityholders.
• Additional Features: Structures will vary deal-to-deal. In addition, structural
nuances such as a class D turbo of principal, the presence and timing of performance
triggers (or lack thereof), and swaps and reserve accounts for synthetic hybrid
transactions may also be present.
Credit Enhancement
• Senior-subordinate capital structure: ABS CDOs are issued as senior/subordinate

transactions, where subordination acts as the primary form of credit enhancement for
bonds that are more senior in priority. In our earlier example of a mezzanine
transaction, the class B, C, and D (each with a rating of AA, A, and BBB,
respectively) provide protection from losses to the class A. The class C and D
provide protection to the class B, and so on. The size of each class will vary from
one transaction to the next, depending on such factors as the collateral composition,
rating composition, and structure. (See Figure 12 for a typical high grade and
mezzanine capital structure.)
• Overcollateralization: All the debt classes begin with a certain amount of
overcollateralization from the outset of the transaction. This overcollateralization
comes from having the first loss piece, or equity tranche, subordinate to the debt
classes. For HG transactions, equity provides about 1% subordination to the class D 6
and about 5% for mezzanine transactions. If the class D turbo is present in a
transaction, this is a feature that serves to increase overcollateralization over time. (In
the “Additional Features” section, we will discuss the class D turbo in more detail.)
• Excess spread: The difference between the interest cash flow from the assets and the
interest paid to the CDO liabilities (including ongoing fees in the transaction),
known as excess spread, acts as a form of additional credit enhancement available to
protect debt tranches from losses. In the absence of losses, excess spread flows
through to the equityholder. To the extent losses are realized, the excess spread may
be redirected to paying down liabilities or “building par” if performance triggers
begin to fail (more on this in the performance trigger section). Prior to failing
6
BB-rated securities may also be issued, in which case the size of the equity would be lower.

triggers, however, excess spread will continue to flow to the equityholder and any
losses are effectively absorbed by the equity.
Certain managed and static transactions also have the ability to sell assets. In managed
transactions, this may occur during the reinvestment period. In some static transactions,
the collateral manager may be able to sell highly appreciated or discounted assets 7.
While the ability to sell assets is not credit enhancement in the strictest sense, providing
the manager with some flexibility to sell assets could benefit bondholders to the extent
that a manager is able to identify assets that may have problems down the road and sell
before the market fully prices in those risks.
Figure 12. Typical ABS CDO Capital Structure, % of transaction size

AAA - Super Senior 80-85% 65-70%
AAA - Mezzanine 8-10% 10-15%
AA 2-3% 7-9%
A 1-2% 1-2%
BBB 1-2% 3-5%
Equity 1% 4-6%
Coverage Tests
Prior to analyzing the coverage tests, priority of payments, and call provisions in
transactions, we expand on our earlier mezzanine example with some additional
assumptions in Figure 13. We assume a reinvestment period of four years, an optional
redemption period between the fourth and eighth year, and an auction call beginning in
year 8. We also provide details on the coverage tests and their required levels.
Traditional ABS CDO transactions have two performance-related coverage tests: the
overcollateralization (OC) and interest coverage (IC) tests. Unlike certain collateral
composition tests (e.g., WARF, weighted average spread), these are traditionally the only
tests that have direct consequences on how cash flow is allocated through the CDO
waterfall. As discussed earlier, a typical transaction will have class A, class B, class C,
and class D. Each class has a coverage test associated with it, with the exception of class
A and B, which share the class A/B coverage test. To the extent that one of the coverage
tests is breached, cash flow may be diverted away from the class of bonds junior to the
test (see the Glossary for a description of how coverage tests are calculated and how
rating downgrades may affect the OC calculation).
7
In traditional static transactions, asset sales may occur and proceeds are applied toward paying principal. In lightly
managed or substitution transactions, a collateral manager may be able to substitute collateral for assets in the pool
to protect the pool from deterioration. The substituted amount is typically capped (i.e. 30% of the initial collateral
balance may be substituted over the first 5 years) and must conform to the eligibility criteria.

Figure 13. Additional Managed Mezzanine Transaction Assumptions

Reinvestment Period First 4 Years
Optional Redemption Period Between the 4th and 8th Year
Auction Call Beginning of 8th Year
Coverage Tests Required

Class A/B OC Ratio Test 109%
Class C OC Ratio Test 104%
Class D OC Ratio Test 102%
Class A/B IC Ratio Test 112%

Class C IC Ratio Test 108%
Class D IC Ratio Test 105%
Haircut Provisions % of Par

Ba1, Ba2, Ba3 in excess of 10% 90%
B1, B2, B3 80%
Caa1 or lower 50%
Priority of Payments
The payments of interest and principal to CDO liabilities follow specific rules that are
dependent on 1) coverage tests passing or failing, 2) whether the transaction is in its
reinvestment period or post-reinvestment period (for managed transaction), and 3) the
collateral factor (how much of the original collateral balance is paid down). Figure 14
shows a typical interest and principal waterfall for a transaction with cash collateral. The
waterfall for hybrid transactions are slightly more detailed in that the super-senior swap
payments and CDS termination payments must also be considered (see the “Additional
Features” section).
• Interest proceeds (coverage tests are passing): Certain taxes, fees, and issuer
expenses are paid prior to the debtholders receiving any interest. Once those
payments are made, if coverage tests are passing, interest (accrued, unpaid and
deferred) is paid sequentially to the class A, B, C, and D. Any remaining interest
proceeds are distributed to the class D as principal (see the “Additional Features”
section), subordinate management fees and then to the equity.
• Interest proceeds (coverage tests are failing): If coverage tests are failing, interest is
paid sequentially up to the level of the failed class. Remaining interest collections
are redirected to paying principal to more senior classes. Using our mezzanine
example, if the class C coverage tests fail, then any remaining interest collections
after paying steps 1-8 of the interest proceeds waterfall (Figure 14) would be
allocated to paying class A, B, and C principal sequentially until the class C
coverage tests pass. When the class C coverage tests pass again, interest payments
may be made to class D 8.
8
If a CDO class has an interest shortfall, the shortfall will accrue at the coupon rate and such amounts may be
repaid when interest and principal proceeds are sufficient according to the priority of payment rules. The shortfall
amount is added to the outstanding principal balance of the bond for calculating current interest.

• Principal proceeds (coverage tests are passing): Principal proceeds are first used to
ensure that steps 1-5 of the interest proceeds waterfall (Figure 14) are met. If
coverage tests are passing, principal proceeds (after ensuring class A, B, C, and D
accrued and deferred interest is paid) are used to purchase new assets if the
transaction is still in its reinvestment period (first four years in our example). After
the reinvestment period ends, the remaining principal proceeds are used to pay the
class A, B, C, and D note principal pro rata (if the outstanding collateral balance is
greater than 50% of the original balance) or sequentially (if the outstanding
collateral balance is less than 50% of the original balance). Once the class A, B, C,
and D note principal is paid down (typically through the call); any additional
principal proceeds flow to subordinate collateral management fees and the equity.
• Principal proceeds (coverage tests are failing): If coverage tests are failing and are
not cured through the interest proceeds waterfall, the principal distributions are used
to pay principal to the classes senior to the trigger level until the tests are cured. All
coverage tests must be cured before any principal proceeds may be reinvested (if
during reinvestment period) or normal principal distributions are made.
Now we consider cases that tie in both the coverage tests and priority of payments. In our
mezzanine transaction example, the equity cash flow could be diverted when the class D
OC falls below 102% or the class D IC falls below 105% (Figure 13). Depending on the
structure, diverted cash flow is paid sequentially as principal starting with the most
senior class down to the failed class until the test is cured. Using the same example,
interest proceeds would be used to pay the class A, B, C, and D sequentially until the
class D OC increases above 102% and the class D IC increases above 105%. These tests
are somewhat unique compared to most ABS transactions in that curing can occur in the
same period. In those cases, the remaining cash flow available after tests are cured would
be allocated according to the “passing” priority of payment waterfall.
If more senior coverage tests are breached (e.g., class A/B, class C), junior classes will
typically PIK (cease paying interest), and interest and principal proceeds would be used
to pay the bonds sequentially through the failed class. In our example, if the class C OC
falls below 104% or the class C IC falls below 108%, then the class D would PIK and
interest and principal proceeds would be used to pay the class A, B, and C until the class
C OC increased above 104% and the class C IC increased above 108%.
Optional Redemption and Auction Calls
In most transactions, the equityholder has the right to call the transaction during the
optional redemption period (Figures 13 and 15). The optional redemption period permits
the equityholders (with a majority vote) to instruct the manager to liquidate the assets in
the trust and pay down the class A, B, C, and D outstanding principal balances at par.
Cash that remains after the repayment of par on the debt classes and fees is then allocated
to paying subordinate manager fees and equityholders. The transaction may not be called
if the liquidation proceeds are insufficient to pay the class A, B, C, and D their par
amount. In our mezzanine example, the optional redemption period begins after the
reinvestment period ends (year 4) and lasts until the auction call date (year 8).
If the transaction is not called during the optional redemption period, the trustee will call
the transaction at the auction call date. In our example, this occurs eight years after the
pricing date. At the auction call date, if the value of the assets equals or exceeds the
amount of outstanding debt of the class A, B, C, and D, the trustee will instruct the
manager to liquidate the assets. The transaction will not be called if the class A, B, C,

and D outstanding principal balances are not fully repaid. If the auction does not succeed,
the trustee will attempt another auction every quarter until the notes are redeemed.
Additional Features
Our mezzanine example provides the basic structural features typically found in the market,
but clearly there are features that will vary from deal to deal. For instance, some transactions
provide additional principal payments to the class D, called a “BBB turbo”; some transactions
have an trigger holiday or no triggers at all; and many have a structure that permits the
majority of collateral to be sourced synthetically. We summarize these features below:
• Class D or BBB Turbo: A BBB turbo feature may be used to pay down a portion of
the class D principal early in the life of the transaction and/or over the entire life of the
transaction. Some turbo features allow 10%-15% of the initial balance of the class D to
be paid down during the first three to four years of the transaction’s life. A certain
percentage of the excess spread may be used to pay class D principal throughout the
life of the transaction. The turbo feature is typically subordinate in the interest
proceeds waterfall (see item 15 in Figure 14). The class D may also turbo if the class D
OC or IC test fails. In that event, rather than paying the class A, B, C, and D
sequentially until the test passes, the class D is paid. However, the class D does not
typically turbo principal if the class A/B and/or class C tests are failing as well.
• Trigger Holiday: Transactions will not have the traditional OC or IC triggers during
the first four to five years of the transaction’s life. This typically coincides with the
reinvestment period. Once the reinvestment period ends, the OC and IC triggers
become active. This is a feature that has been used in only managed mezzanine
transactions to date.
• Triggerless: Transactions will not have the traditional OC or IC triggers throughout
the life of the transaction. Therefore, there is no mechanism in the waterfall to divert
interest and principal from the class C, D, and equity. However, as a protection to
more senior bondholders in the event the OC declines above a certain level, there is
a trigger that allocates all principal proceeds sequentially rather than pro rata,
whether the transaction is still in its reinvestment period or not. This is a feature that
has been used in only managed mezzanine transactions to date.
• Hybrid or Synthetic Transactions: Most mezzanine transactions issued since late 2005
have a combination of cash assets and synthetic assets. When the majority of assets are
synthetic, the transactions are referred to as “hybrid” or “synthetic” CDOs. The addition
of synthetic assets through CDS necessitates some structural changes to the traditional
cash CDO model. First and foremost, the top 65%-70% of the capital structure, typically
known as the “super-senior” is issued as an unfunded swap 9, where the trust pays the
swap provider a fixed premium each month and receives cash payments from the swap
provider to the extent the super-senior becomes undercollateralized (Figure 16). The
notes, which are issued subordinate in the capital structure (typically the bottom 30%-
35%) to the super-senior, are usually fully funded and pay a monthly or quarterly
floating-rate coupon. The proceeds are used to purchase cash assets and establish a
reserve account to be used in the event payments must be made on the CDS
positions (such as writedown payments to the buyer of protection).
9
This may also be issued as a variable funding note (VFN) rather than in a synthetic swap form.

Figure 14. Example Interest and Principal Waterfall
Interest Proceeds Principal Proceeds
1. Items 1-5 of Interest

1. Taxes, Fees, Expenses
Proceeds, if not paid in full
2. Class A, then Class B If Class AB Coverage Tests

2. Senior Management Fee
Principal (sequential) are not met
3. Hedge Fees & Payments, 3. Class C Accrued &
if any Unpaid Interest*
4. Class A Accrued and 4. Class A, B, and C If Class C Coverage Tests
Unpaid Interest Principal (sequential) are not met
5. Class B Accrued and 5. Class C Accrued and
Unpaid Interest Unpaid Deferred Interest
If Class AB Coverage Tests 6. Class A, then Class B 6. Class D Accrued and
are not met Principal (sequential) Unpaid Interest*
7. Class C Accrued and 7. Class A, B, C and D If Class D Coverage Tests
Unpaid Interest Principal (sequential) are not met
8. Class C Accrued and 8. Class D Accrued and
Unpaid Deferred Interest Unpaid Deferred Interest
If Class C Coverage Tests 9. Class A, B, and C 9. Reinvestments (during
are not met Principal (sequential) Reinvestment Period)
10. After Reinvestment
Period & when Collateral
10. Class D Accrued and
Balance >50% of Original
Unpaid Interest
Balance: Class A, B, C and
D Principal Pro-Rata
11. After Reinvestment
Period & when Collateral
11. Class D Accrued and
Balance <50% of Original
Unpaid Deferred Interest
Balance: Class A, B, C and
D Principal Sequential
If Class D Coverage Tests 12. Class A, B, C and D 12. Subordinate Collateral
are not met Principal (sequential) Management Fee
13. Subordinate Collateral 13. Additional Trustee and If Interest Proceeds did not
Management Fee Other Fees pay in full
14. Additional Trustee and 14. Additional Hedge
Other Fees Payments
15. Subordinated Notes
15. Class D Principal Turbo
(Equity)
16. Additional Hedge
Payments
17. Subordinated Notes
(Equity)
* To the extent not paid from interest proceeds, and only to the extent such payments do not cause a senior OC Ratio to fail.

Figure 15. Paydown of CDO Liabilities (to Auction Call)
1,000
Millions
800
Auction Call
Reinvestment Optional
600 Period Redemption
Period
400
200
0
0 24 48 72 96 120 144
Deal Age (mos.)
Equity BBB A AA AAA
Figure 16. Synthetic or Hybrid Transaction Structure
ABS CDO Assets ABS CDO Liabilities
Credit Default Swaps

Senior Swap Agreement (70%)
$700,000,000
15 bp
Ongoing P&I
Class A (10%)
Issuance AAA
Proceeds 3L+45 bp
Cash Collateral Debt Securities
$200,000,000 Class B (8%)
AA
3L+55 bp
Class C (2%)
A
3L+135 bp
Cash Reserve Account
$100,000,000 Class D (5%)
BBB
3L+325 bp
Equity (5%)

Valuation Considerations
In this section, we summarize how transactions are rated by the agencies and typically
analyzed by the market, and provide our more robust bottom-up loan-level approach to
valuing ABS CDOs.
Rating Agency and Market Pricing and Stress Assumptions
In rating ABS CDO liabilities, the rating agencies tend to split the task into two separate
analyses. The first analysis uses default, recovery rate, correlation, and diversification
assumptions to arrive at an expected loss for the portfolio. The second analysis applies
those losses as well as specific cash flow stresses to the specific CDO structure in order
to determine whether the CDO tranches can withstand the stresses.
Each rating agency uses its own methodology for rating CDOs. This may incorporate a
host of techniques 10: Moody’s uses its “correlated binomial method (CBM)” to produce
the expected portfolio loss based on default, recovery, and correlation inputs through
CDOROM. Standard & Poor’s CDO Evaluator incorporates the probability of default by
asset types and correlations among, and within, asset types to arrive at an expected loss.
Fitch’s VECTOR uses Monte Carlo simulation of defaults, losses, and correlations to
produce expected losses and loss distributions.
Investors, however, tend to assess transactions by analyzing cash flows through a “top-
down” approach originally developed by the corporate CDO market. Default, recovery,
and prepayment assumptions are used to price and stress transactions as follows:
• Prepayment rate: Each underlying security in the portfolio is priced using its pricing
speed or six-month average historical prepayment rate (for seasoned securities).
Each security is assumed to pass its triggers and be called at its call date. This
generates the transactions aggregate collateral cash flow to which top-level defaults
are applied.
• Default rate: A constant default rate, or investor defined curve, is applied to the
aggregate balance of the portfolio as determined under the pricing speed
prepayments assumptions discussed above.
• Recovery rate: A recovery rate is applied to the defaulted balance with a lag
(typically 12 months). For example, a 60% recovery rate is often used for mezzanine
transactions, but the assumption may differ from one transaction to the next
depending on the asset mix.
An ABS Loan-Level Approach to Losses and Prepayments
We believe a more realistic method for developing prepayment and loss assumptions is
warranted in order to capture the “ABS” nature of the collateral. We prefer to use a
“bottom-up” approach that begins with a loan-level analysis of the specific securities in
the transaction, accounting for the intricacies of the underlying loans and bond structures.
Using an approach of this nature can more accurately reflect the timing of principal
paydowns of the collateral as well as the timing of losses which could impact the average
life and evaluation of risk in the CDO liabilities.
For example, let us consider a “normal” and “stress” credit scenario whereby we analyze
the collateral of the underlying HEL securities (i.e., the subprime loans themselves)
10
This summary of the rating agency process is an oversimplification of their methodologies. We should note that in
addition to the collateral analysis, rating agencies are also very involved in the legal and structural features that are
present in CDOs. We encourage investors to visit the rating agency websites for more details.

within a generic static mezzanine CDO by applying different voluntary prepayment,

default, severity, and delinquency vectors to the HEL loans in order to project how HEL
securities perform over time. We find that writedowns of the HEL tranches (and hence
CDO collateral losses) tend to be more back-ended (by at least 12 months) than typical
CDO-style stress assumptions imply (Figures 17a and 17b). (For our more detailed report
on the topic, see “Putting the “ABS” in ABS CDOs” in the U.S. ABS Weekly Outlook,
April 21, 2006.)
Turning to the prepayment assumptions currently used in the market, we find that typical
prepayment assumptions at pricing are generally consistent with the realistic
prepayments in “normal” credit environments, but can overstate the rate of paydowns in
a “stress” case (Figure 18 – we use a static mezzanine transaction for simplicity). The
reason prepayments on ABS CDO collateral could slow in a “stress” case more than
what is implied by typical stress cases is that the underlying HEL performance triggers
could begin to fail and cause the securities to extend significantly and underlying
transactions are less likely to be called.
Figure 17a. CDO Cumulative Losses – Normal Figure 17b. CDO Cumulative Losses - Stress
Cum Loss Cum Loss

2.0% 10%
1.5% 8%
1.0% 5%
0.5% 3%
0.0% 0%
0 12 24 36 48 60 72 84 96 108 120 0 12 24 36 48 60 72 84 96 108 120
Deal Age (Months) Deal Age (Months)
Normal CDO Style Stress CDO Style
Source: Lehman Brothers Source: Lehman Brothers
Figure 18. Static Mezzanine CDO Collateral Balance Factor
Balance Factor
1.0
0.8
0.6
First Auction
0.4 Call Date
0.2
0.0
0 24 48 72 96 120 144 168 192
Deal Age (Months)
Pricing Normal Stress
Source: Lehman Brothers. Assumes a seven year auction call.

Risk Factors
Credit
A key risk in ABS CDOs is leverage. Rating downgrades and losses on CDO liabilities
could increase as a result of a worsening in credit of the underlying collateral securities.
While this is not unique to CDOs, the CDO liabilities can be more leveraged to
worsening credit than ABS, MBS, or CMBS.
For example, consider a mezzanine and high grade CDO transaction where, for
simplicity sake, the portfolio is comprised of 100% HEL securities and the mezzanine
has BBB HELs and high grade has AA and A HELs. If we project CDO cumulative
losses in different credit scenarios by stressing voluntary prepayment, default, severity,
and delinquency rates on the underlying HEL loans of HEL transactions, we find that at a
certain point, the CDO collateral losses will begin to increase at a more rapid pace than
the HEL loans (Figure 19).
This is fairly intuitive if one thinks about the timing of how HEL loan losses would
permeate through the HEL securities and into CDOs in a high stress scenario. HEL loan
losses must first be high enough to result in a writedown to the most subordinate HEL
securities (say the BBs) before reaching the HEL BBBs. If losses were to stop there, the
CDO collateral losses would be zero, while the HEL collateral losses would be greater
than zero. If losses continued to increase and were high enough to result in HEL BBB
writedowns, the CDO would begin to incur its first collateral losses. If HEL loan losses
continue to increase until the HEL BBBs are completely written down, this would result
in 100% cumulative losses to the mezzanine CDO (assuming no paydown of principal
occurred). If HEL loan losses increase further and begin writing down the HEL AA and
A tranches, the high grade CDO begins to incur collateral losses. If loan losses stopped
increasing at that point, they would not be high enough to reach the HEL AAAs, but
would result in 100% writedown to the AAA, AA, A, and BBB classes of the mezzanine
CDO and a significant portion of the high grade classes.
Figure 19. ABS CDO and HEL Collateral Cumulative Losses
Cum Losses
50%
40% Mezzanine CDO Collateral Losses

High Grade CDO Collateral Losses
30% HEL Collateral Losses
20%
10%
0%
Meltdow n Stress Normal

Basis Risk on Fixed-Rate Security Hedges
When a portion of CDO collateral is fixed-rate but CDO liabilities are all floating-rate,
the trust is exposed to interest rate risk (either positive or negative, assuming there are no
hedges in the transaction). When short interest rates rise, the average CDO liability cost
increases while only a portion (the floating rate portion) of the collateral interest
increases. This results in a decrease in excess spread. The reverse occurs when short rates
rally. While the trusts typically purchase amortizing swaps and caps to match the fixed-
rate notional over time in an attempt to hedge this risk, mismatches can occur when
prepayments deviate from the pricing speed assumption.
An additional risk exists whereby the weighted average coupon of the fixed-rate
collateral drifts higher or lower throughout the life of the transaction. This “coupon drift”
impacts the excess spread, as there may be more or less interest cash flow coming into
the transaction than initially expected. The coupon drift is dependent on the interest rate
scenario, the types of fixed-rate collateral, payment priority, and whether underlying
transaction triggers are passing or failing (to name a few), all variables that increase the
uncertainty of fixed-rate interest cash flows in the transaction (see our analysis of “Fixed-
Rate Collateral Risk in ABS CDOs” in the ABS Strategy Weekly, June 2, 2006). Of
course, the “coupon drift” is not just a risk for the fixed-rate securities in the transaction,
but for the entire portfolio, as the average spread on floating-rate securities will also drift
based on similar variables.
Available Funds Cap Risk (AFC)
Because upwards of 50% of the collateral in more recent vintage ABS CDOs are HEL
securities, available funds cap (AFC) risk in HELs is also a risk for ABS CDOs. The risk
to ABS CDO investors is that 1) a decline in HEL excess spread increases the chance of
HEL writedowns, 2) HEL securities are downgraded because of their lower
enhancement, and 3) AFC interest shortfalls on HELs result in less excess spread to the
CDO.
AFC risk stems from the fact that the underlying HEL securities are exposed to rising
interest rates due to the basis risk between floating-rate bonds and a combination of their
fixed-rate and hybrid collateral (which are subject to periodic and lifetime caps on the
rate of the mortgage). As interest rates rise, the fixed-rate loans in the pool are unable to
reset higher and hybrid loans may hit their caps, resulting in excess spread being
squeezed. This can create problems later in the life of a transaction when losses are
higher. The HEL trusts may purchase swaps and/or caps which partly offset the interest
rate risk inherent in HEL transactions.
Call Risk/Extension Risk
Under normal circumstances, transactions will be called during the optional redemption
period or at the auction call date. The auction call will be exercised to the extent the
liquidation value of the portfolio is 100% or higher. However, to the extent the
transaction is not called, bondholders are susceptible to extension risk. The amount of
extension risk will vary depending on the underlying portfolio characteristics and credit
scenario. For example, in a normal credit scenario, AAA to BBB securities could have
about two to three years of extension risk if the transaction is not called (Figure 20).
However, if one applies a credit scenario stressful enough to result in underlying HEL
trigger failure and CDO losses, the extension risk could be much greater.

Figure 20. Managed Mezzanine Liability Average Lives, to Call and to Maturity
Average Life (Years,
To Auction Call) Average Life (Years, To Maturity)
Tranche Normal Normal Stress
AAA – SS 6.8 9.2 11.4
AAA – Mezz 6.8 9.4 16.5
AA 6.8 9.4 18.3
A 6.8 9.4 21.1
BBB 6.3 8.6 1.6*

* The average life is significantly shortened because the BBB incurs a writedown in this example.
Conclusion
The tremendous growth of the ABS CDO market has created the second largest ABS
sector by outstandings and has accounted for more than 60% of global CDO issuance
YTD. ABS CDOs add additional complexity to what are already complex underlying
sectors, but offer attractive nominal spreads compared to investing directly in the ABS,
MBS, CMBS, or CLO investments. As with any securitized sector, a better
understanding of structure, collateral, and risks can help uncover value in the market.

Glossary
• Auction Call: The trustee instructs the manager to liquidate all the assets in the
portfolio as long as all the debt classes receive par. This typically occurs seven or
eight years after the pricing date. If the auction fails, the trustee will attempt another
auction every quarter.
• Coupon Drift: Any deviation in the average coupon or spread of the collateral from
the initial collateral spread. With respect to fixed-rate collateral, to the extent the
average coupon on fixed-rate bonds drifts from the rate of the amortizing swap,
excess spread could be affected.
• Coverage Tests: Performance-related triggers such as interest coverage (IC) or
overcollateralization (OC) tests that can affect the payment priority through the
waterfall.
The calculation of the OC is as follows:
Class A/B OC = outstanding collateral balance (plus paydowns received during the
period) / (A + B)
Class C OC = outstanding collateral balance (plus paydowns received during the
period) / (A + B + C)
Class D OC = outstanding collateral balance (plus paydowns received during the
period) / (A + B + C + D)
where A, B, C, and D refer to the outstanding balance of their respective classes at
the beginning of the period.
The calculation of the IC is as follows:

Class A/B IC = (S – costs – fees) / (a + b)
Class C IC = (S – costs – fees) / (a + b + c)
Class D IC = (S – costs – fees) / (a + b + c + d)
where S is the scheduled interest proceeds on the collateral debt securities and a, b,
c, and d refer to the scheduled interest plus interest on deferred interest amounts of
the respective classes. Costs and fees refer to hedging and trustee costs and manager
fees.
An important consideration when calculating OC is that the outstanding collateral
balance is determined by the par value of the assets rather than the market value.
Therefore, appreciation or depreciation in the portfolio’s value will not affect the OC
calculation for trigger purposes. However, when an asset is sold at a premium or
discount, the gain or loss on the sale will increase or decrease the OC level as a
result.
Rating migration may also influence coverage tests. Rather than assuming a par
value for all the assets in the portfolio when calculating OC, bonds that have
incurred downgrades might incur a haircut to the par amount. In our mezzanine
example, if the total Ba1, Ba2, and Ba3 exposure becomes greater than 10%, the
difference between the actual exposure and 10% is haircut to 90% (Figure 13). In
other words, if 15% of the portfolio is rated Ba1, Ba2, and Ba3, 5% of the portfolio
(15% minus 10%) will be haircut by 10%, or effectively carried at 90% of the par

value. This would lead to an OC decline of 0.5% when calculating the OC coverage.
Additionally, any B1, B2, and B3 exposure will be haircut by 20%, or carried at
80% of the par value and any Caa1 or lower exposure will be haircut by 50%, or
carried at 50% of the par value.
• Excess Spread: The difference between the interest cash flow received from the
assets and the interest paid on the liabilities (minus ongoing fees).
• Hybrid Transaction: When a portion of the CDO collateral is synthetic,
transactions may be referred to as hybrids.
• Loan-to-Value (LTV): The aggregate debt outstanding divided by the value of the
collateral.
• Managed transaction: A manager is paid a fee to select the initial CDO portfolio of
assets and make investment decisions during the reinvestment period.
• Optional redemption period: For managed transactions, the optional redemption
period begins after the reinvestment period ends and lasts until the auction call date.
For static transactions, the optional redemption period begins after the “no-call”
period. During this time period, the equityholders can call the transaction with a
majority vote as long as the debt is paid back at par.
• Overcollateralization: The notional of the collateral outstanding divided by the
debt outstanding. For instance, if a transaction has $103 of collateral and $100 of
debt outstanding, the OC is 103%.
• PIK (payment-in-kind): If a debt class does not receive its full stated interest
payment, the bond is considered to be PIKing.
• Reinvestment period: The period during a managed transaction, typically the first
three to five years, when principal proceeds received on the underlying collateral
may be invested for substitute collateral and/or the manager has some flexibility in
buying and selling collateral securities.
• Sequential Pay Period: When the collateral balance is less than 50% of the original
collateral balance, or the auction call is not successful, the principal proceeds are
used to pay down the debt sequentially by order of seniority.
• Static transaction: A portfolio advisor is paid a fee to select the initial CDO
portfolio and conduct ongoing surveillance. The portfolio is typically fully ramped at
closing.
• WARF: The weighted average rating factor, or WARF, is a concept originated by
Moody’s. It assigns a number to each rating which represents the expected
cumulative default rate over a 10-year period (Figure 23). The cumulative default
rate is rating-specific, rather than sector-specific. For CDOs, WARF is useful in
providing one number to help describe the average credit of the portfolio.

Figure 21. Rating Factor of Debt Securities
Rating Factor Rating of Debt Security Rating Factor Rating of Debt Security
1 Aaa 940 Ba1
10 Aa1 1,350 Ba2
20 Aa2 1,766 Ba3
40 Aa3 2,220 B1
70 A1 2,720 B2
120 A2 3,490 B3
180 A3 4,770 Caa1
260 Baa1 6,500 Caa2
360 Baa2 8,070 Caa3
610 Baa3 10,000 Ca or lower
Source: Moody’s Investor Service

Explanation of the Lehman Brothers Mortgage Model
The Lehman Brothers Mortgage Valuation Model allows investors to analyze mortgage-backed (MBS), asset-backed (ABS) and
commercial mortgage-backed securities (CMBS). The model collects pertinent and material information needed to evaluate and
calculate the risk measures of the security. The model provides option-adjusted spreads and durations along with other risk
measures using Lehman Brothers' Prepayment, Default, and Term Structure Models.
Analyst Certification
The views expressed in this report accurately reflect the personal views of Michael Koss, the primary analyst responsible for this
report, about the subject securities or issuers referred to herein, and no part of such analyst's compensation was, is or will be
directly or indirectly related to the specific recommendations or views expressed herein.
Important Disclosures
Lehman Brothers Inc. and/or an affiliate thereof (the "firm") regularly trades, generally deals as principal and generally provides
liquidity (as market maker or otherwise) in the debt securities that are the subject of this research report (and related derivatives
thereof). The firm's proprietary trading accounts may have either a long and / or short position in such securities and / or
derivative instruments, which may pose a conflict with the interests of investing customers.
Where permitted and subject to appropriate information barrier restrictions, the firm's fixed income research analysts regularly
interact with its trading desk personnel to determine current prices of fixed income securities.
The firm's fixed income research analyst(s) receive compensation based on various factors including, but not limited to, the
quality of their work, the overall performance of the firm (including the profitability of the investment banking department), the
profitability and revenues of the Fixed Income Division and the outstanding principal amount and trading value of, the profitability
of, and the potential interest of the firms investing clients in research with respect to, the asset class covered by the analyst.
Lehman Brothers generally does and seeks to do investment banking and other business with the companies discussed in its
research reports. As a result, investors should be aware that the firm may have a conflict of interest.
To the extent that any historical pricing information was obtained from Lehman Brothers trading desks, the firm makes no
representation that it is accurate or complete. All levels, prices and spreads are historical and do not represent current market
levels, prices or spreads, some or all of which may have changed since the publication of this document.
Lehman Brothers' global policy for managing conflicts of interest in connection with investment research is available at
www.lehman.com/researchconflictspolicy.
To obtain copies of fixed income research reports published by Lehman Brothers please contact Valerie Monchi
(vmonchi@lehman.com; 212-526-3173) or clients may go to https://live.lehman.com/.
Company-Specific Disclosures
As an administrator of; and frequent issuer, structurer and dealer in structured and securitized credit products, you should
assume that Lehman Brothers has a significant financial interest in one or more of the products that are discussed in this
publication. As a frequent dealer in debt securities you should also assume that Lehman Brothers has a significant financial
interest in the underlying assets of the products that are discussed in this publication.
Legal Disclaimer
This material has been prepared and/or issued by Lehman Brothers Inc., member SIPC, and/or one of its affiliates ("Lehman
Brothers"). Lehman Brothers Inc. accepts responsibility for the content of this material in connection with its distribution in the
United States. This material has been approved by Lehman Brothers International (Europe), authorised and regulated by the
Financial Services Authority, in connection with its distribution in the European Economic Area. This material is distributed in
Japan by Lehman Brothers Japan Inc., and in Hong Kong by Lehman Brothers Asia Limited. This material is distributed in
Australia by Lehman Brothers Australia Pty Limited, and in Singapore by Lehman Brothers Inc., Singapore Branch ("LBIS").
Where this material is distributed by LBIS, please note that it is intended for general circulation only and the recommendations
contained herein do not take into account the specific investment objectives, financial situation or particular needs of any
particular person. An investor should consult his Lehman Brothers' representative regarding the suitability of the product and
take into account his specific investment objectives, financial situation or particular needs before he makes a commitment to
purchase the investment product. This material is distributed in Korea by Lehman Brothers International (Europe) Seoul Branch.
Any U.S. person who receives this material and places an order as result of information contained herein should do so only
through Lehman Brothers Inc. This document is for information purposes only and it should not be regarded as an offer to sell or
as a solicitation of an offer to buy the securities or other instruments mentioned in it. With exception of the disclosures relating to
Lehman Brothers, this report is based on current public information that Lehman Brothers considers reliable, but we do not
represent that this information, including any third party information, is accurate or complete and it should not be relied upon as
such. It is provided with the understanding that Lehman Brothers is not acting in a fiduciary capacity. Opinions expressed herein
reflect the opinion of Lehman Brothers' Fixed Income Research Department and are subject to change without notice. The
products mentioned in this document may not be eligible for sale in some states or countries, and they may not be suitable for all
types of investors. If an investor has any doubts about product suitability, he should consult his Lehman Brothers representative.
The value of and the income produced by products may fluctuate, so that an investor may get back less than he invested. Value
and income may be adversely affected by exchange rates, interest rates, or other factors. Past performance is not necessarily
indicative of future results. If a product is income producing, part of the capital invested may be used to pay that income.
Lehman Brothers may, from time to time, perform investment banking or other services for, or solicit investment banking or other
business from any company mentioned in this document. No part of this document may be reproduced in any manner without the
written permission of Lehman Brothers.
© 2006 Lehman Brothers. All rights reserved. Additional information is available on request. Please contact a Lehman Brothers'
entity in your home jurisdiction.

LB - Abs Cdo Primer

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

LB - Abs Cdo Primer

Uploaded by

Copyright:

Available Formats

ABS CDOs – A Primer

November 17, 2006 OVERVIEW

Role of the Collateral Manager 8

A Brief History of the ABS CDO Market 9

Rating Agency and Market Pricing and Stress Assumptions 20

November 17, 2006 2

What is an ABS CDO?

November 17, 2006 3

HEL CDO RMBS Other

November 17, 2006 4

The Importance of CDOs

CDO Economics 101

November 17, 2006 5

November 17, 2006 6

Figure 3. Typical Mezzanine ABS CDO Capital Structure

ABS CDO Assets ABS CDO Liabilities

Average Coupon of Assets

Source: Lehman Brothers

Figure 4. Simple CDO Arbitrage Calculation

Collateral manager Fees (bp) 25

Equity Leverage (=100%/5%) 20

Approximate Gross Equity IRR (%) (=1.04%*20) 20.8

Source: Lehman Brothers

November 17, 2006 7

Role of the Collateral Manager

November 17, 2006 8

A Brief History of the ABS CDO Market

The growth in ABS CDO issuance may be attributed to a number of factors:

Figure 5. ABS CDO Issuance

Source: Lehman Brothers. Through 9/30/06.

November 17, 2006 9

Diversified versus Real Estate Era

Figure 6. ABS CDO Timeline

Diversified Era Real-Estate Era

2000 2001 2002 2003 2004 2005 2006 2007

Source: Lehman Brothers

November 17, 2006 10

Figure 8. ABS CDO Spread History

AAA AA A (RHS) BBB (RHS)

Figure 9. Floating-Rate Spread Comparison, bp

Source: Lehman Brothers. As of 9/30/06.

November 17, 2006 11

Figure 10. Average Annual Rating Transition Matrices

Original Rating Aaa Aa A Baa Upgrade Stable Downgrade

Figure 11. Upgrade/Downgrade Statistics by Asset Class

Sector Category 2005 2004 1996-2005

Source: Moody’s Investors Service

November 17, 2006 12

• Senior-subordinate capital structure: ABS CDOs are issued as senior/subordinate

November 17, 2006 13

Figure 12. Typical ABS CDO Capital Structure, % of transaction size

Source: Lehman Brothers

November 17, 2006 14

Figure 13. Additional Managed Mezzanine Transaction Assumptions

Coverage Tests Required

Class A/B IC Ratio Test 112%

Haircut Provisions % of Par

Source: Lehman Brothers

November 17, 2006 15

November 17, 2006 16

November 17, 2006 17

Figure 14. Example Interest and Principal Waterfall