Lehman - Auto ABS Primer

U.S. Securitized Products FIXED INCOME INCOME RESEARCH RESEARCH | U.S. SECURITIZED SECURITIZED PRODUCTS | WEDNESDAY, APRIL 30, 2008

An Introduction to Auto ABS Consumer ABS

OVERVIEW

Brian Zola 212-526-8311 [email protected]

Gross issuance of auto ABS has averaged around $80 billion since 2002, with current total outstandings of around $150 billion. The widening of spreads has generated renewed interest in the sector, from both traditional securitized products participants and cross-over investors who have traditionally focused on credit. As a result, we are releasing a new auto ABS primer.

Kumar Velayudham 212-526-8311 [email protected]

Quantitative Research Gaetan Ciampini 212-526-5751 [email protected]

The primer is organized into four sections. In the first section, we provide a broad overview of the auto ABS sector. We look at the different collateral types—prime, near prime, and subprime—as well as the key participants and trends within each of the sectors. In the second section, we analyze the collateral characteristics of auto ABS. The focus is on examining the key variables (collateral and macroeconomic) that drive prepayment and credit behavior. In the third section, we look at auto ABS securitization structures, detailing some of the features unique to auto ABS, and compare deal structures and credit enhancements across issuers. In the fourth and final section, we provide a few basic frameworks for analyzing auto ABS, providing a framework for both credit and prepayment analysis of auto ABS securities.

Sue Li 212-526-6681 [email protected]

PLEASE SEE ANALYST CERTIFICATIONS AND IMPORTANT DISCLOSURES STARTING AFTER PAGE 22

Lehman Brothers | U.S. Securitized Products Research

TABLE OF CONTENTS Overview............... Overview............................. ............................ ............................ ............................ ............................ ............................ ............................ ........................... ............. 1

Auto ABS Market

3

Size of the Market Market ........................... ......................................... ............................ ........................... ........................... ............................ ............................ ................ .. 3 Collatera Collaterall Types ........................... ........................................ ........................... ............................ ............................ ........................... ........................... .................... ...... 3 Prime Prime versus versus Subprime Subprime ........................... ......................................... ........................... ........................... ........................... ........................... ................ .. 3 Captive Finance versus Independent Finance Issuers ..................... .......... ...................... ...................... ................... ........ 4 Pool Characteristics Across Sectors ...................... ........... ..................... ..................... ...................... ...................... ...................... ............. .. 5

Collateral Performance

6

Auto Loan Characteri Characteristics.................. stics................................ ........................... ........................... ........................... ........................... ........................... ............... 6 Auto ABS Conventions................ Conventions.............................. ............................ ........................... ........................... ............................ ............................ ................... ..... 6 Voluntary Prepayment Characteristics ..................... .......... ..................... ..................... ...................... ...................... ...................... ................. ...... 7 Loan-Level Loan-Level Factors Factors .......................... ........................................ ............................ ........................... ........................... ............................ ...................... ........ 7 Deal-Level Deal-Level Factors Factors .......................... ........................................ ............................ ............................ ............................ ............................ ...................... ........ 9 Macroecon Macroeconomic omic Factors Factors .......................... ........................................ ........................... ........................... ........................... ........................... ................ .. 9 Credit Credit Character Characteristic isticss .......................... ........................................ ............................ ........................... ........................... ............................ ........................11 ..........11 Loan-Level Loan-Level Factors Factors .......................... ........................................ ............................ ........................... ........................... ............................ .................... ...... 11 Deal Level Factors Factors ............................ .......................................... ............................ ............................ ............................ ............................ .................. .... 12 Macroecon Macroeconomic omic Factors Factors .......................... ........................................ ........................... ........................... ........................... ...........................13 ..............13

Securitization Structure

14

Key Characteri Characteristics stics .......................... ........................................ ............................ ........................... ........................... ............................ ...........................14 .............14 Credit Enhanceme Enhancement nt ........................... ......................................... ............................ ............................ ............................ ............................ ........................15 ..........15 Credit Credit Enhanceme Enhancement nt across across Deals..................................... Deals.................................................. ........................... ........................... ............... .. 16

Security Valuation

17

Deal Level Analysis—Ford 2007-A............... 2007-A..... ..................... ...................... ..................... ..................... ...................... ..................... .............. .... 17 Credit Credit Enhanceme Enhancement nt ........................... ......................................... ............................ ............................ ............................ ............................ ................. ... 17 Historical Historical Performan Performance........... ce........................ ........................... ............................ ............................ ........................... ........................... ................. ... 18 AAA Prepayment Prepayment Analysis Analysis ........................... ........................................ ........................... ........................... ........................... ........................... ............... .. 18 Premium Premium Bonds.............................. Bonds........................................... ........................... ............................ ........................... ........................... ....................... ......... 18 Discount Discount Bonds....................................... Bonds.................................................... ........................... ............................ ............................ ...........................19 .............19 Prepayments and Subordinates............. ..................... ........... ..................... ..................... ..................... ...................... .................. ....... 19 Subordinat Subordinatee Security Security Credit Credit Analysis Analysis .......................... ........................................ ........................... ........................... ...........................20 .............20 Constant Constant CDRs.......................... CDRs........................................ ........................... ........................... ............................ ............................ ...........................20 .............20 Using Loss Multpiles Multpiles ........................... ......................................... ........................... ........................... ........................... ........................... ................. ... 21 Deleveragi Deleveraging ng Effect ............................ ......................................... ........................... ............................ ............................ ........................... ................. .... 21

April 30, 2008

2


AUTO ABS MARKET SIZE OF THE MARKET Gross annual issuance of auto ABS has averaged $80 billion in recent years

Since the first deal in 1985, auto ABS has grown into one of the market’s core sectors. The auto ABS market showed steady growth until 2002, when issuance reached a record $101 billion. 1 Since then, gross issuance has averaged around $80 billion (not counting floor plans, leases, and motorcycles), with total outstandings of around $150 billion. This decline in growth rates can be attributed to the maturing of the sector, lower auto sales for some of the traditional securitizers, and some participants exiting the market. Figure 1.

Auto ABS Issuance and Outstanding ($ billion)

120

175

100

150 125

80

100 60

75

40

50

20

25

0

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Issuance(LHS)

Outstanding (RHS)

Source: Lehman Brothers

COLLATERAL TYPES Prime versus Subprime Prime issuance has declined since 2000, whereas subprime issuance has remained relatively steady

Auto ABS deals can be broadly classified as prime, near prime, and subprime (and more recently, deep subprime), based on pool characteristics. The segmentation is mainly based on the overall credit quality of the borrowers in the pool. Prime deals tend to have higher average FICO scores, lower APRs, and higher loan balances (discussed later). Prime issuance has declined since 2000, but subprime issuance has increased (Figure 2). Thus, the proportion of subprime collateral in the auto ABS market has almost doubled from 14% in 2000 to 29% in 2007. The prime sector is dominated by the financial subsidiaries of automotive manufacturers (Figure 3). The Big Three (Ford, GM, and Chrysler), along with Honda and Nissan, have dominated issuance in the prime sector. USAA and Capital One are the only large prime issuers that are not subsidiaries of an automotive firm. Historically, Chase had also been a regular issuer, but it has not issued a deal since 2006. The near-prime sector has shrunk as traditional issuers were acquired. For example, Onyx was acquired by Capital One in 2004, and its collateral is now split between Capital One’s subprime and prime shelves. The subprime collateral market is dominated by independent finance companies. AmeriCredit and Capital One are the biggest participants in the subprime sector. HSBC also has a sizeable subprime platform from its acquisition of Household Finance. 1

April 30, 2008

The focus of our primer is retail auto loan ABS, not dealer floorplan or lease transactions.

3


Figure 2.

Auto ABS Issuance Across Sectors ($ billion)

120 100 80 60 40 20 0 1993

1995

1997

1999 Prime

2001

Near Prime

2003

2005

2007

Subprime

Source: Lehman Brothers, Intex

Figure 3.

Top Issuers Across Sectors, 2002-2007 ($ billion)

Collateral Type

Issuer

Prime

Ford

33

Honda

28

GMAC

27

Chrysler

23

Nissan

19

USAA

19

Chase

12

Capital One Prime

10

Wachovia (WFS)

29

Hyundai

5

AmeriCredit

25

Capital One Subprime

24

Household

13

Drive

10

Triad

10

Near Prime Sub-Prime

Issuance ($ billion)


Captive Finance versus Independent Finance Issuers One of the key features that differentiates captive deals from finance-issuer deals is subvented loans

April 30, 2008

Issuers can also broadly be classified as captive finance and bank/independent finance issuers. Captive finance issuers are subsidiaries of automotive companies, whereas bank/independent finance issuers are just that. Captive firms tend to dominate loan origination for new cars, whereas finance companies tend to dominate used-car loan originations. A key feature that differentiates a captive finance deal from an independent financial deal is the presence of subvented loans, defined as loans with below-market interest rates. These loans are originated as part of an incentive program by an auto manufacturer to increase sales. The presence of subvented loans affects both structure (discussed later) and collateral behavior. Overall, captive finance deals tend to be prime deals (Figure 3), with higher FICOs, lower APRs, shorter loan terms, and higher concentration of new cars. The share of issuance of independent finance companies has increased as the market for subprime financing has grown (Figure 4). Independent finance companies now account for a majority of new issuance.

4


Figure 4.

Share of Issuance of Captive and Independent Issuers

100% 80% 60% 40% 20% 0% 2000

2001

2002

2003

Captive Finance

2004

2005

2006

2007

IndependentFinance


Pool Characteristics Across Sectors Average FICO scores for subprime deals are close to 600, while average FICO scores for prime deals are close to 710

The static pool characteristics of the various sectors are shown in Figure 5. Some of the key variables in which collateral differs across the credit spectrum are: 







Figure 5.

APR: Borrowers with worse credit will tend to receive higher loan rates (APR). Thus, subprime deals tend to have higher APRs. FICO: Average FICO scores for subprime deals are close to 600, while average FICO scores for prime deals are close to 710. % New/Used Cars: Prime deal issuers generally tend to be subsidiaries of auto manufacturers and tend to have a high concentration of new cars in their pools. Loan Term: Loan terms in subprime pools tend to be longer, owing to the tighter budget constraints of subprime borrowers. This trend is not as true as it was several years ago because of the lengthening of loan terms from the U.S. captive finance companies.

Static Pool Characteristics Across Sectors Issuer

Deal Name

APR

GMAC

CARAT 2007-1

4.5%

33%

87/13

702

53

20,999

Ford

FORDO 2007-A

4.4%

28%

87/13

705

57

21,761

Nissan

NAROT 2007-A

5.7%

1%

90/10

750

53

19,220

USAA

USAOT 2007-1

6.7%

0%

66/34

723

58

18,951

Near-Prime

Wachovia

WALOT 2007-1

12.4%

0%

28/72

638

60

16,688

Sub-Prime

AmeriCredit

AMCAR 2007-AX

17.2%

0%

25/75

591

68

17,865

Capital One

COAFT 2007-A

13.7%

0%

33/67

607

68

16,496

Prime

APR <1% New /Used %

FICO

Term (Rem.) Avg Loan Balance

Source: Lehman Brothers, deal prospectus

April 30, 2008

5


COLLATERAL PERFORMANCE

AUTO LOAN CHARACTERISTICS We summarize a few of the key characteristics of auto loans below: Loan Term: Auto loans are generally four to five year fully amortizing fixed-rate loans. The recent trend has been toward longer term loans, and 72-month loans are not uncommon. Prepayments: Most prepayments are due to trade-ins and defaults. Rate-related refinances are rare because of the smaller loan size, declining value of the collateral, and higher borrowing rates for used cars compared with new cars. Loan-to-Value Ratio: The initial loan-to-value ratio averages between 90% and 110%, but higher amounts are not uncommon. LTVs can be higher than 100% due to the inclusion of tax, title and other fees. Originators often differ in their methods for calculating LTV. The collateral value declines faster than principal amortization during the initial years, causing the mark-to-market LTVs to increase initially. Loan Losses: Default rates tend to increase until months 12-18, when the default curve typically starts to flatten. The ease of repossession and sale in case of default makes for relatively high recovery rates (close to 50%) compared with other consumer ABS sectors.

AUTO ABS CONVENTIONS Typical prime deals are priced using a flat 1.5 ABS curve; near-prime and subprime deals price closer to 1.7 ABS

Since the collateral is amortizing, securities are priced assuming a prepayment speed. Prepayment speeds in auto ABS are often quoted in ABS terms rather than in CPR (Box A). Typical prime deals are priced using a flat 1.5 ABS curve, while near prime and subprime deals price closer to 1.7 ABS. The average life of the security is quoted based on the pricing speed and usually ranges from 0.25 to 4 years. Dealer quotes are provided as a spread to base rate (swaps or Eurodollar) at a particular speed, and prices are derived from spreads using the yield tables (Box B). For example, the quote on CARAT 07-1 C (as of April 1, 2008), would read N+700 at 1.0 ABS. This would imply a price of 87-22 and an average life of 3.6 years.

BOX A: ABS PREPAYMENT RATE

The auto sector uses a prepayment measure called the absolute prepayment speed (ABS). It measures the monthly rate of loan prepayments as a percentage of original pool balance less scheduled amortization of the original pool. This method differs from the conditional prepayment rate (CPR) used in the mortgage sector, which reports prepayments as an annualized percentage of the current pool balance. A constant ABS speed implies a monotonically increasing CPR curve. Mathematically, ABS can be related to SMM as:

ABS

=

1+

SMM SMM * ( Age

−

1)

100

For a more detailed explanation of ABS, please refer to the publication titled “Auto ABS: Is ABS an Appropriate Measure ?”.

April 30, 2008

6


BOX B: E SPREAD VERSUS N SPREAD

Two different conventions, N-spread and E-spread, are used in pricing auto ABS. Auto ABS started pricing against the swap curve following the 1998 Russian default and LTCM crises. Since the swap curve is less liquid inside two years, market participants use the Eurodollar spot curve for securities with weighted average life less than two years. In the secondary market, shorter bonds are priced using a cash flow spread over the Eurodollar spot curve (Bloomberg E-Curve), whereas longer bonds are priced using a yield spread over the swap curve (Bloomberg N-Curve). For further details, refer to “Deconstructing the N vs. E Curve Drop (Page 97)”.

BOX C: INTEX VERSUS BLOOMBERG SPEEDS

Bloomberg and Intex, the two leading providers of ABS data and analytics, adopt different methods for calculating historical prepayment speeds. Significant differences can result from switching between the two CPR conventions. On both platforms, auto ABS collateral is modeled by aggregating loans into a smaller number of representative loans. Bloomberg uses a single representative loan, commonly referred to as the “rep line,” whereas Intex employs multiple representative loans. As a result, Bloomberg speeds tend to be faster and tend to make the weighted average life of bonds look shorter. For a detailed analysis of the two methods, refer to “Measuring prepayment speeds in Auto ABS ”

VOLUNTARY PREPAYMENT CHARACTERISTICS Even though credit performance has been the focus recently, voluntary prepayment behavior has historically played a more critical role

Even though credit performance has been the focus recently, voluntary prepayment behavior has historically played a more critical role in bond valuation. Prepayment variations can have a dramatic effect on average life and, hence, on valuations. The effects are more pronounced when the securities are trading away from par. For securities higher in the capital structure, prepayment behavior typically plays the most important role in valuation; credit characteristics are more important in valuing subordinate bonds. Auto originators normally price new deals using a constant ABS prepayment curve. In real life, actual prepayments are driven by a variety of factors: •

Loan Level Factors: FICO, LTV, new/used mix, and loan term

•

Deal Level Factors: Seasoning and sector (prime versus subprime)

•

Macroeconomic Factors: Auto sales and rate incentive

We examine each of these in more detail in the following sections.

Loan-Level Factors Unlike mortgages, loan-level data are not available for auto ABS

In a standard deal-level analysis, it is difficult to isolate the effects of different loan-level characteristics such as borrower credit score (FICO), loan-to-value ratio, loan term, and the percentage of used cars. Unlike mortgages, loan level data are not available for auto ABS. In this analysis, we use Lehman Brothers’ proprietary loan-level database to identify the effect of these individual loan-level characteristics on collateral credit performance. Since our loan-level database encompasses performance from 1998 to 2003 for a specific set of loans, the results may not be fully applicable today. Nonetheless, we believe that this provides a good benchmark. Based on our analysis, we make the following observations: •

April 30, 2008

Borrower Credit: FICO has a significant effect on voluntary prepayments. Better quality borrowers with higher FICO scores tend to prepay faster (Figure 6). A prime new car borrower with a 750 FICO prepays 50% faster than a subprime new car borrower with a 600 FICO.

7


Used versus New: Holding other variables constant, there is only a limited difference between used and new car prepayment rates (Figure 6). Used car borrowers seem to prepay marginally faster than new car borrowers, potentially because of shorter used car lives.



Loan to Value: LTV seems to have a moderate effect on prepayment behavior. Lower LTV loans prepay 50% faster than higher LTV loans over the first two years of the loan (Figure 7a).



Loan Term: Loan term has a significant effect on prepayment rates. Voluntary prepayment rates on shorter term loans are significantly higher than prepayment rates on longer term loans (Figure 7b). This is likely because borrowers with longer loan terms are often more budget constrained.



Figure 6.

Voluntary Prepayments Across FICO (CRR, %)

40

30

20

10

0 6

12

18

24

600 (new )

30

600 (used)

36

750 (new )

42

48

750 (used)


Figure 7a. Prepay Speed Across LTVs (CRR, %)

Figure 7b. Prepay Speed Across Loan Terms (CRR, %)

40

40

30

30

20

20

10

10

0

0 6

12

18 100


April 30, 2008

24 120

30

36 140

42 160

48

6

12

18 36

24 48

30 60

36 72

42

48

84


8


Deal-Level Factors Voluntary prepays on subprime deals tend to decline after the higher quality borrowers prepay

Sector and Seasoning: Voluntary prepay profiles (as defined by CRR) for prime and subprime deals tend to be similar during the first couple of years. Subprime prepayment speeds tends to decline after that (Figure 8). This can be attributed to higher credit quality borrowers prepaying in the initial period, leaving the pool with lower quality borrowers who tend to prepay slower. Voluntary prepays account for approximately 90% of total prime prepay speeds and 60% of subprime speeds. Figure 8. Voluntary Prepayment by WALA, 2004 Vintage (CRR, %) 30 25 20 15 10 5 0 4

14

24 Prime

34

44

Subprime


Macroeconomic Factors Prime borrowers with a higher incentive tend to prepay faster than borrowers with no incentive

April 30, 2008

Interest Rates

Although interest rates have a limited effect on overall voluntary prepayments rates, they tend to have a more significant effect on higher credit quality borrowers. Using a rate incentive based on the change in 2-year swap rate since origination, we observe that prime borrowers with a higher incentive tend to prepay faster than borrowers with no incentive (Figure 9). Borrowers with a 750 FICO score prepay a new car loan at 16% CRR when 200 bp out of money, while the CRR goes up to 26% when they are 400 bp in the money. In practice, rates do not often exhibit such large swings, and the option cost for auto ABS is usually less than 1 bp. Please note that these conclusions are based on relatively old data and may be less applicable today because of the growth in subvented loans.

9


Figure 9.

Voluntary Prepayment Speed vs. Rate Incentive (CRR, %)

35 30 25

) % ( 20 R R 15 C

10 5 0 -200

-80

40

160

280

400

Incentive (bp)

600

650

700

750

800


Auto Sales

Prime auto ABS prepayment speeds are correlated with the level of auto sales

Auto ABS prepayment speeds are somewhat correlated with the level of auto sales. Figure 10 shows normalized auto sales, defined as actual sales divided by prior 12-month average, plotted against the level of prepayments. The linkage is strongest for prime deals, for which 90% of the prepayments are voluntary in nature. The relationship also tends to be stronger for Detroit manufacturers. Figure 10.

Normalized Auto Sales vs. Prepayment Speed (ABS)

1.5

1.2

1.4 1

1.3 1.2

0.8

1.1 1 Jan-00

0.6 Jan-01

Jan-02 36 Wala A BS

Jan-03

Jan-04

Jan-05

GM A uto Sales (RHS)


April 30, 2008

10


CREDIT CHARACTERISTICS Defaults account for 40% of total prepayments in subprime deals and 10% of prepayments in prime deals

With the continued deterioration in consumer credit, there has been increased attention on collateral credit and the drivers of credit performance. Most AAA securities are well protected, but this does not mean that credit is not an issue. Higher defaults also lead to higher prepays and, hence, affect the weighted average life and valuation of senior securities. Defaults can account for 40% of total prepayments in subprime deal, but are often just 10% of prepayments in prime deals. Similar to voluntary prepayments, collateral credit performance is driven by a host of collateral and macroeconomic factors.

Loan-Level Factors We again use Lehman Brothers’ proprietary loan-level database to identify the effect of these loan-level characteristics on collateral credit performance. We analyze the effect of credit quality (FICO), LTV, used/new car mix, and loan term on collateral credit performance and make the following observations: •

Borrower Credit: FICO has a significant effect on cumulative collateral losses. Cumulative defaults at month 36 for the 600 FICO bucket (which is the average for subprime deals) are 2.5 times the cumulative defaults in the 700 FICO bucket (average for prime deals; Figure 11a).

•

Loan to Value: Front-end LTVs have a significant effect on credit performance. Cumulative defaults at 36 months in the 140 LTV bucket are twice those of the 100 LTV bucket (Figure 11b).

•

Loan Term: Loan term has a significant effect on default rates. At 36 months of seasoning, cumulative defaults for 72-month loans are more than twice those of the 36-month loans (Figure 11c).

•

Used versus New: Holding borrower credit constant, there is only a limited difference between used and new car cumulative defaults (Figure 11d).

Figure 11a. Cumulative Defaults by FICO (%, by WALA)

Figure 11b. Cumulative Defaults by LTV (%, by WALA) 12

20

10

15 8

10

6 4

5

2

0 1

13

24

36

0 0


April 30, 2008

550

600

650

700

750

800

12

24

95

100

160

165

120

36 140


11


Figure 11c.

Cumulative Defaults by Loan Term (%, by WALA)

Figure 11d.

12

Cumulative Defaults by % Used (%, by WALA)

12

10 8

8 6

4 4 2

0 0

0 0

12 36

24 48

60

12

24

36

36 72

84


600 (Used)

600 (new )

750 (Used)

750 (new )


Deal Level Factors Subprime default rates are 4x-6x prime default rates; subprime cumulative losses tend to be 6x-8x higher than prime

Sector: The default rate expressed in percent CDR terms tends to be 4x-6x higher for subprime deals than for prime deals. Prime CDRs average close to 2%, w hereas seasoned subprime CDRs average close to 12% (Figure 12a). The cumulative losses in subprime deals tend to be 6x-8x higher than cumulative losses experienced in prime deals. This ratio is higher than the ratio of default rates because of the lower recovery rates in the subprime sector. Prime recovery rates average 50%, compared with close to 40% in subprime (Figure 12b). Seasoning: Default rates in both the prime and subprime sectors tend to increase with seasoning. The default rate in subprime tends to increase sharply in the initial months and flattens after that; prime, on the other hand, sees a steady increase in default rates over the life of the deal. For example, in the 2002 vintage, prime defaults increased by 50% from 18 to 36 WALA, while subprime defaults increased by only 20% between the same WALA.

Figure 12a.

CDR (%), Across WALA

Figure 12b.

14

Recovery Rate by WALA (%, 2002 Vintage)

50

12

45

10 40

8 6

35

4 30

2 0 6

12

18

24

30

36

25 6

12

18

24

30

36

WALA Prime


April 30, 2008

Sub Prime

Prime

Sub-Prime


12


Macroeconomic Factors Proportionally, changes in the labor market have similar effects on prime and subprime default rates

Figure 13a.

Historically, the level of unemployment has had a significant effect on the credit performance of auto ABS collateral (Figure 13). Subprime defaults more than tripled from trough to peak during the 2001 cycle. Prime defaults are also very sensitive to the level of unemployment, but the peak default levels in the 2001 cycle were still below 2%. Although the absolute level is different, proportionally the effect is similar. Non-farm payroll addition is also well correlated with the level of defaults (Figure 14). Some of this correlation can be attributed to change in underwriting standards over time. During periods of strong employment and economic conditions, issuers tend to lower underwriting standards, and they tighten them as conditions start to deteriorate. This cyclical change in underwriting standards can amplify the effect of macroeconomic conditions on collateral credit performance as the economic cycle turns.

Prime CDR

Figure 13b.

2

7 )

6 % ( t n e m 5 y o l p m e 4 n U

) % ( 1 R D C

0 Jan-98 Jul- 99 Jan-01 Jul- 02 Jan-04 Jul-05 Jan-07 12-24 Wala CDR

Subprime CDR

16

7

12

6 % (

)

t n e m 5 y o l p m e 4 n U

) % ( R 8 D C

4

3

0 Jan-98 Jul- 99 Jan-01 Jul- 02 Jan-04 Jul-05 Jan-07 12-24 Wala CDR

Unemployment (RHS)

Source: Lehman Brothers, Intex, BLS

3

Unemployment (RHS)

Source: Lehman Brothers, Intex, BLS

Figure 14.

Default Rate versus Non-Farm Payrolls (12 to 36 WALA Prime, 2003 to 2007)

3.0 2.5

R2 = 73%

) 2.0 % ( R 1.5 D C

1.0 0.5 0.0 -200

-100

0

100

200

300

400

Payroll (1000s) Source: Lehman Brothers, BLS, Intex

April 30, 2008

13


SECURITIZATION STRUCTURE

KEY CHARACTERISTICS In this section, we summarize the features associated with auto ABS structures, focusing on the various credit enhancement methods and how they are used in prime and subprime deals.

Sequential Prepay Structure Auto ABS deals generally have sequential-pay structures. Figure 15 shows the structure of the Ford 2007-A trust. The chart shows both assets and liabilities, along with the credit rating, weighted average life, and size of each tranche. All principal amortizations and prepayments are allocated to the shortest maturity class until it is fully repaid and then directed to the next shortest class. Losses beyond the reserve account and overcollateralization are first written down from the lowest rated tranche available. The AAA notes are pari passu, and if there are any writedowns, they are taken pro rata. Money Market Class

In most prime auto ABS deals, the class A-1 notes are structured as a money market– eligible security

Figure 15.

In most prime auto ABS deals, the class A-1 notes are structured as a money market– eligible security. According to rule 2a-7 of the Investment Company Act of 1940, money market–eligible securities cannot have a maturity longer than 13 months. The size of the money market class is determined using stress assumptions with low prepayments and high delinquencies. The total principal that will amortize under these assumptions, by the final date, is the face amount that is 2a-7 eligible.

FORDO 2007-A Trust Structure TRUST ASSETS (Total = $2,225m)

TRUST LIABILITIES ( Total = $2,032m)

A1 (AAA, 0.3yr, 0.3yr, 23%) $466m, 23%) A1 (AAA, A2 (AAA, 1yr, $588m, 29%)

e c n a l a B l o o P

Adjusted Principal Balance ($1,992m, 89%)

P r i n c i p a l P r i o r i t y

A3 (AAA, 2yr, $549m, 27%) A4 (AAA, 3.1yr, $289m, 14%) BB(A,(A, 3.8yr, $60m, 3.8yr, 3%) 3%)

n o i t a c o l l A s s o L

C (BBB, 3.9yr, $40m, 2%) YSOC ($233m, 10.5%) Reserve ($11m, 0.5%)

D (BB, $40m, 2%) D (BB, 2%) Equity ($193m)

Source: Ford, Lehman Brothers

April 30, 2008

14


Floating-Rate Class

The majority of auto ABS issuance is in fixed-rate securities because of the fixed-rate nature of the underlying collateral. However, floating-rate securities can be created using interest rate swaps. These classes follow principal payment rules like any other fixed-rate class. In a few cases, the floating rates are capped, but most are not. Clean-Up Call

Historically, the clean-up call was set at 10% of the initial pool balance, but it is now often lower

Most deals provide the servicer with the option to buy the securities at par when the balance is below a certain threshold. Historically, the clean-up call was set at 10% of the initial pool balance, but more recently, the threshold has been reduced to 5% and even as low as 2%. The clean-up call offers protection against bond extension. Since the bonds must be called at par plus accrued interest, a deal in which the clean-up call is exercised would not suffer any principal writedowns.

CREDIT ENHANCEMENT For subprime auto ABS deals, excess spread can be as high as 800 bp, while it is close to zero for some prime deals

The key forms of credit enhancement used in auto ABS are excess spread, senior/subordinated structure, over-collateralization, bond insurance, and reserve funds. In this section, we detail the basics of these credit enhancement methods and analyze how they are used in prime and subprime deals Excess Spread: The first line of credit protection in an auto ABS deal is the excess spread. Excess spread is defined as the difference between the interest cash flow from underlying loans and the combined investor coupon, servicing, and trust costs. The excess spread is available to cure defaults. For subprime auto ABS deals, excess spread can be as high as 800 bp, while it is close to zero for some prime deals. The reason for this difference is the presence of subvented loans in prime deals. Subprime deals will use excess spread as a significant source of credit enhancement, while prime deals do not. Targeted overcollateralization reserve accounts are generally used to capture excess spread.

Because of the presence of subvented deals, prime deals have a designated yield supplement overcollateralization account

Senior-Subordinate Structure: The senior subordinate structure is widely used in auto ABS. This structure designates one portion of the transaction as subordinated to the remaining portion. The lowest rated class still outstanding incurs writedowns before the bonds senior to it. Prime deals rely more on the senior subordinate structure for credit enhancement, while subprime deals are often wrapped by a monoline insurer (explained later). Overcollateralization: In overcollateralization, the amount of securities issued against a pool is smaller than the amount of collateral in the pool. Overcollaterization is usually used in conjunction with a senior-subordinate structure to provide protection to the senior classes. •

April 30, 2008

Prime: Because of the presence of subvented loans, prime deals have a designated yield supplement over-collateralization account (YSOC). This is an overcollateralization account created to make up for the interest shortfall from subvented loans when yields do not cover the trust’s cost of capital. The size of the YSOC account is determined by summing the present value of interest shortfalls from all subvented loans in a pool. The primary purpose of YSOC is to support interest payments on subvented loans, where loan rates are lower than the trust’s cost of capital. However, conservative assumptions are usually made in calculating YSOC, so in some deals it also acts as a form of credit over-collateralization.

15


•

Subprime: Initial over-collateralization levels in subprime deals range from 5% to 10%. Over-collateralization is usually allowed to build up to a pre-set target as a way of capturing excess spread. Excess cash flows are passed to residual only after over-collateralization targets are met. These pre-set target over-collateralization limits are sometimes relaxed if collateral performance is better than certain targets.

Reserve or Spread Account: This is a cash account used to address cash shortfalls in the trust. The account usually ranges from 0.5% to 2.0% of the pool. Withdrawals from the account are replenished in future periods up to pre-set target levels. In some deals, reserve accounts step up if collateral performance deteriorates past certain triggers. Insurance: Monoline insurance wraps are often used in subprime structures. Insurance guarantee usually includes full and timely payment of interest and ultimate repayment of principal.

Credit Enhancement across Deals The total level of over-collateralization tends to be large in prime captive deals, but after adjusting for yield supplement for subvented loans, the level of over-collateralization is much smaller. Figure 16.

Credit Enhancement Across Trusts (at Origination) Prime

Near Prime

Subprime

FORDO 07-A

CARAT 07-1

USAOT 07-1

AAA Hard Support

15.5%

15.5%

3.6%

14.5%

9.0%

9.0%

41.0%

Bond Subordination

6.3%

4.7%

2.8%

14.3%

0.0%

0.0%

34.0%

Over Collateralization

8.7%

10.3%

0.0%

0.0%

7.0%

7.5%

5.5%

Reserve Account

0.5%

0.5%

0.9%

0.3%

2.0%

1.5%

1.5%

YSOC

10.5%

10.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Adjusted OC

-1.8%

0.3%

0.0%

0.0%

7.0%

7.5%

5.5%

Adjusted AAA Support

5.0%

5.5%

3.6%

14.5%

9.0%

9.0%

41.0%

No

No

No

No

Yes

Yes

No

Insurance

WALOT 07-1 AMCAR 07-AX* COAFT 07-A*

AMCAR 06-1

Source: Prospectus, calculated as of the pricing date * Are wrapped deals – underlying securities are shadow-rated to a rating lower than AAA

April 30, 2008

16


SECURITY VALUATION In this section, we discuss techniques to analyze auto ABS securities. For consistency, we use the same Ford 2007-A deal explained in the previous sections as an example.

DEAL LEVEL ANALYSIS—FORD 2007-A Credit Enhancement Yield Supplement Over-collateralization (YSOC): The deal has an initial YSOC of 11.5%. The YSOC amount is the difference between the receivables’ future cash flows discounted at their APR and discounted at the trust’s weighted average cost of capital plus a spread. This is essentially the present value of the cost (in dollars) of subvented loans in the pool, under conservative prepay assumptions. The adjusted principal balance (APB), defined as the pool balance minus the YSOC, is 98% of the note balance. Thus, the note balance is under-collateralized by 2%. Parity is reached by turboing down the class A notes. Reserve Account: The reserve account is funded on the closing date at 0.5% of the initial pool balance. Subordination: Subordinate classes do not receive any principal until classes senior to them are fully repaid. This deal has a AAA subordination level of 7%. Target Over-collateralization: Target over-collateralization is defined as the sum of YSOC and a pre-defined percentage of the pool balance. For this deal, the target overcollateralization on a specific date is defined as the sum of YSOC and the excess of 1% of current pool balance over 0.5% of initial pool balance. As long as the actual overcollateralization is below this target, excess cash is used to pay down senior note principal and not released to the residual class. This is one of the most common credit enhancement techniques for using excess cash in auto ABS deals. As the deal seasons, the target overcollateralization decreases until it hits a floor. Figure 17.

Cash Flow Priority—Ford 2007-A

Priority Tranche Payment

1

Class A Interest st

Payment Rule

Purpose

Pro rata based on Class A principal

Class A Interest has the highest priority. Makes sure that Class A is not under collateralized

2

1 Priority Principal

Pay down Class A if Note A >APB

3

Class B Interest

Interest Due to Class B

nd

4


Pay down Class A & B if Note A + B >APB

5

Class C Interest

Interest Due to Class C

rd

Ensures that class A&B are not under collateralized

6


Pay down Class A, B & C if Note A+B+C >APB Ensures that class A,B &C are not under collateralized

7

Class D Interest

Interest Due to Class D

8

Reserve Account

Replenish Reserve Account for Withdrawals

9

Regular Principal –A1 Repay A1 principal until fully paid down

Accelerates repayment of Money Market Class

10

Regular Principal

If Actual OC < Target OC, pay down notes

Captures excess spread to pay down notes

11

Residual

Pay residual if there is still excess cash

Replenishes reserve account

Source: Prospectus Adjusted Principal Balance (APB) = Pool Balance – YSOC Target OC = YSOC + Maximum (1% of Current Balance, 0.5% of Initial Pool Balance)

April 30, 2008

17


Historical Performance We start with a look at historical performance to provide a baseline for expectations of future collateral performance (Figure 18). The 2001 vintage is the worst performing Ford vintage in recent history. Historical cumulative loss for Ford deals has averaged 1.7%, and the historical CDR has averaged 2.8%. The poorer performing 2001 vintage saw a cumulative loss of 2.2% and an average CDR of 3.8%. Figure 18.

Historical Performance of Ford Deals (12 to 36 Wala)

Vintage

Average ABS

Average CPR

Average CDR

Average Loss Severity

Cumulative Losses

2000

1.30

19.6

2.36

59%

1.54

2001

1.32

20.9

3.78

58%

2.17

2002

1.33

20.1

3.13

52%

1.71

2003

1.27

18.9

2.70

47%

1.71

2004

1.21

19.0

1.86

43%

1.44

Average

1.29

19.7

2.77

52%

1.71

Source: Intex

AAA PREPAYMENT ANALYSIS As stated earlier, the valuation of bonds up the capital structure is driven primarily by voluntary prepayments. Credit plays a less significant role in the analysis of AAA bonds. Valuation of bonds selling close to par is less sensitive to speed assumptions, so we focus on premiums and discounts. Please note that the following analysis, unless otherwise stated, was performed as of the issuance date and not on seasoned collateral.

Premium Bonds Historically, estimating prepayment speeds was crucial for outperformance in AAAs

We use the last cash flow AAA A4 bond in the Ford 2007-A deal to discuss the valuation techniques for premium securities. In our stylized example, 2 if the Ford 2007-A A4 is trading at a spread of N+30 bp and one assumes a prepay speed of 0.8 ABS, the price is 104-22. If the prepay speed increases to 1.4 AB S 3 (equivalent of going from 10% CPR to 20% CPR at the deal’s current age of 16 months), the life of the bond shortens by six months and the yield on the security drops by 30 bp (Figure 19). Since the yield curve is upward sloping, the spreads themselves are less sensitive to speed changes than the yields. An increase in prepayment speeds from 0.8 to 1.4 ABS causes the spread to drop by 20 bp. This may not be much compared with current spread levels, but given that AAA spreads have historically been in the single digits, investors have focused on prepayment analysis as a way of generating relative value.

2

Spreads and prepayment speeds are f or illustrative purposes only and do not necessarily correspond to current market prices or observed prepayment and/or default behavior. 3 We prefer to price securities using CPR, but we recognize that the market convention is ABS, so we use ABS in our examples.

April 30, 2008

18


Figure 19.

Prepayment Sensitivity of Premium Bonds—FORDO 07-A A4

3.9

3.2

3.8 3.0 3.7 2.8 3.6 3.5 0.80

1.00 1.20 Prepaym ent Speed (ABS) Y ield (LHS,% )

2.6 1.40

WA L (RHS, Y ears )

Source: Lehman Brothers. Analysis assumes no default

Discount Bonds At faster prepay speeds, the yield on the discount bond increases; in addition, in an upward sloping yield curve environment, the spread increases even faster. For example, assume the CARAT 2008-1 A4 bond is trading at a spread of 130 bp with an expected prepay speed of 0.8 ABS (price equals 98). If the prepayment speed increases from 0.8 to 1.4 ABS, the yield increases by 10 bp, while the spread increases by 25 bp (Figure 20). Figure 20.

Prepayment Sensitivity of Senior Discount Bonds—CARAT 08-1 A4

5.00

270

4.95

260

4.90

250

4.85 0.80

1.00

1.20

240 1.40

Prepayment Speed (ABS)

Y ield (%)

Spread (bp, RHS)

Source: Lehman Brothers. Analysis assumes no default

Prepayments and Subordinates Voluntary prepayments will also play an important role in the analysis of credit-sensitive bonds. For example, say the Ford 2007-A class D bond is trading at a $70 dollar price at a 0.8 ABS prepayment speed. As the speed increases from 0.8 to 1.4 (equivalent of CPR going from 10% to 20%), the WAL decreases by seven months, the yield increases by 112 bp, and the spread increases by an even larger 128 bp (Figure 21). Furthermore, as prepay speeds increase for a given default rate, cumulative defaults decrease.

April 30, 2008

19


Figure 21.

Prepayment Sensitivity of Subordinate Discount Bonds— FORDO 07-1 Class D

5.0

18.0

4.8 17.5

4.6 4.4

17.0

4.2 4.0 0.80

1.00

16.5 1.40

1.20

Prepaym ent Spee d(ABS) WA L(LHS, Y ears )

Y ield(RHS, % )

Source: Lehman Brothers. Analysis performed at a constant 4% CDR and 50% loss severity

SUBORDINATE SECURITY CREDIT ANALYSIS Constant CDRs Collateral credit performance becomes much more critical for analyzing securities down the capital structure. We again use the Ford 2007-A deal as the example. Figure 22 shows the constant default rate needed to cause the first dollar principal loss in the Class D BB-rated bond, at various severities. For example, at 50% loss severity, the Class D bond breaks at a constant CDR of 6.8%. Historically, Ford deals have averaged a 2.8% CDR and a 52% loss severity. Although faster prepay speeds generally improve credit performance, the effect is much smaller than changing the recovery rates (Figure 22a).

The constant CDR that breaks the bond provides a good first-order approximation of credit protection

In deals with limited excess spread, such as those with substantial amounts of subvented collateral, faster prepayment of higher APR loans during the initial periods may negatively affect the bonds’ credit by lowering breakeven loss multiples. Figure 22a. First Loss CDR vs. Severities (Class D)

Figure 22b. First Loss CDRs vs. Severities (1.0 ABS) 70

20

60

R D C

16

50

12

R 40 D C 30

8

20 10

4

0 20

0 20

40

60

80

Severity (%) 0.75 A BS


April 30, 2008

1.5 A BS

40

60

80

100

Severity (%)

100 A 4 (A A A )

B (A )

C (BBB)

D (BB)


20


We perform the same analysis across the capital structure in Figure 22b. For example, the CDR needed to break the A4 bond in this deal is three times that of the class D bond. This analysis ignores the upward sloping nature of the default curve but provides a good first-order approximation of how protected a given bond may be.

Figure 23a. 2001 Vintage CDR Seasoning Curve

Figure 23b. First Loss CDR Multiple, Severities—Class D

16

3.5

12

3.0 e l p i 2.5 t l u M R 2.0 D C

R D8 C

4

1.5 0 10

22

34

46

WALA

Base

2X

1.0 50

60

80

90

100

Severity ( %)

3X


70


Using Loss Multpiles A more accurate approach is to construct a theoretical loss curve based on pool-level data and find the multiple of the curve that causes a loss in the bond. This technique takes into account the upward sloping nature of the default curve. Alternatively, one could use a historically poor performing vintage as the base case. We use the default experience from the poorly performing 2001 vintage as the base case. We then generate CDR curves at a multiple of these base-case CDR levels and find the one that causes the bond to take a loss (Figures 23a). For example, we estimate that at 50% severity, the CDRs have to be three times the 2001 levels for the Class D bond in this deal to break. We also show the sensitivity of the loss multiple to loss severities (Figure 23b).

Deleveraging Effect Auto ABS securities benefit from a deleveraging structure

April 30, 2008

Deals tend to deleverage as they season. During the initial period of the deal, defaults tend to be relatively low (upward sloping CDR curve), so the excess cash flows are used to pay down the senior securities and build overcollateralization (as long as actual overcollateralization is lower than target overcollateralization), causing the deal to deleverage. So long as the realized CDR is lower than the break-point CDR, the bond will build credit enhancement as it seasons. Figure 24 shows the losses needed to break the Ford 2007-A bonds, as of the pricing date and as of the current date. At issuance, the AAA A4 bond broke at a 12.5% CDR. After seasoning, as of current date, the bond is protected up to a CDR of 18%. The securities up the coupon structure tend to deleverage faster than lower-rated securities.

21


Figure 24.

Auto ABS Deleveraging Effect, First Loss Breaking Points Pricing Date, Pool Factor = 1.00

Class

First Loss CDR

Seasoned, Pool Factor = 0.76

Cumulative Losses Cumulative Defaults

First Loss CDR

Forward Cum. Losses

Forward Cum. Defaults

A4

12.5

13.1

21.9

18.0

16.2

27.0

B

9.6

10.3

17.2

13.5

12.6

20.9

C

7.7

8.4

14.0

10.6

10.1

16.8

D

5.6

6.2

10.4

7.6

7.4

12.3

Source: Lehman Brothers, Intex, Assumes 60% loss severity and a voluntary prepayment speed of 1.0 ABS

April 30, 2008

22

Analyst Certification The views expressed in this report accurately reflect the personal views of Brian Zola and Saravanakumar Velayudham, t he primary analysts responsible for this report, about the subject securities or issuers referred to herein, and no part of such analysts' 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 ([email protected] ; 212-526-3173) or clients may go to https://live.lehman.com/. 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 dis tributed in Australia by Lehman Brothers Australia Limited, and in Singapore by Lehman Brothers Singapore Pte Ltd. Where this material is distributed by Lehman Brothers Singapore Pte Ltd, 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, and in Taiwan by Lehman Brothers Securities Taiwan Limited. Where this material is distribut ed by Lehman Brothers Securities Taiwan Limited, please note that recommendations expressed herein are for reference only. Investors should carefully evaluate the investment risks and are reminded that they are solely responsible for their investment decisions. 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. © 2008 Lehman Brothers. All rights reserved. Additional information is availab le on request. Please contact a Lehman Brothers' entity i n your home jurisdiction.

Lehman - Auto ABS Primer

Recommend Documents