Introduction

The Federal Reserve Bank of New York (“FRBNY”) launched its Term Asset Backed Securities Loan Facility (“TALF”) on March 3, 2009 at a time when the market for automobile loan asset-backed securities (“ABS”) was moribund. With the U.S. suffering a protracted period of high unemployment, deteriorating consumer confidence and reduced disposable income together with a frozen credit market, the automobile market witnessed a decreased volume of automobile loan origination, increased loss frequency on existing loans and weakened collateral values on certain types of vehicles. These factors eroded the credit quality of existing automobile loans and constricted the availability of new loans, which ultimately weakened collateral coverage and increased the risk of loss holders of ABS would face in the event of a default. Growing concerns regarding credit risk had widened spreads considerably. With respect to subprime securitizations in particular, this concern had also led to increased levels of overcollateralization and other forms of credit enhancement, further reducing the appeal for issuers of securitizing automobile loan pools. Furthermore, by early 2008, most monolines faced or had suffered downgrades of their credit ratings. This combination resulted in the virtual elimination of note insurance as a source of credit support (only one public offering of automobile loan ABS in 2008 was a wrapped deal) and the exit from the term market of virtually all subprime issuers.  

The result of these conditions was a deep reduction of the automobile loan securitization market. The ABS market saw approximately 33 public issuances of automobile loan ABS in 2008 for a total offered balance of $36.62 billion. This was considerably below 2007, which had comparable figures of approximately 45 and $49.05 billion, respectively. The position of subprime automobile loan ABS deals was even worse: in 2007 approximately 16 deals had publicly offered notes with a total principal value of $15.21 billion; in 2008 the comparable figure was approximately $3.85 billion from 4 deals. To reverse these trends, TALF was implemented with a view to stimulating demand for ABS by making low interest loans available for the purchase of qualifying ABS. The increased demand, the thinking went, would reduce spreads on ABS and therefore reduce the cost of funds for consumer lending in various segments of the economy. The savings would presumably be passed along to the consumer.  

Prime Issuances

The market in which the TALF program would operate was one sorely in need of stimulation—there were only five public offerings of prime automobile loan ABS between Q4 2008 and the implementation of TALF in Q1 2009, and none reported a spread on the last maturing class of its AAA-rated notes of less than 350 basis points (“bps”). More positively, though, it was also a market whose participants had begun to adjust to prevailing conditions. Most originators responded to poor performance in existing pools by tightening origination standards on new loans. While the stated criteria for rating automobile loan ABS remained unchanged, rating agencies began to focus on the risk of future downgrades of, as well as the risk of default on, the rated notes. This involved conducting an analysis of whether the ratings on the notes could withstand certain multiples of expected loss levels in addition to the application of existing ratings standards. In light of expected deterioration in collateral perform-ance, it also meant a closer examination of a transac-tion’s collateral and an emphasis on higher levels of credit enhancement and collateral quality to protect against subsequent downgrades. Similarly, investors and underwriters, responding to higher than expected losses in collateral pools, also looked for higher quality collateral and demanded higher levels of enhance-ment—often at levels higher than those required by the rating agencies. Acting in tandem, TALF and these market driven adjustments soon began to influence the prime automobile loan ABS deals coming to market.  

Collateral

Confronted with higher than anticipated levels of default and delinquency in existing asset pools, and recognizing the negative impact of a sustained economic slowdown on asset pools’ future performance, issuers began improving their collateral quality. The TALF requirement that substantially all of the underlying automobile loans included in a collateral pool be originated no earlier than October 1, 2007 accelerated this trend, particu-larly where originators had moved to more rigorous underwriting standards in the wake of the credit crunch. While lingering unemployment and depressed income continue to adversely affect pool performance relative to pre-2008 levels, most platforms have seen lower rates of increase in delinquency and loss figures in recent periods; some have even seen these figures decline in recent periods. For example, CarMax Auto Owner Trust transactions experienced increases in both total delinquencies as a percentage of its total receivables portfolio and net losses as a percentage of average outstanding principal amounts in calendar 2008 and 2009. However, these increases in 2009 were signifi-cantly lower than the comparable increases in both 2008 and 2007. Similarly, Honda Auto Receivables Owner Trust (“HAROT”) transactions experienced increases in net losses as a percentage of average outstanding principal amounts in fiscal years 2009 and 2008, which were significantly lower relative to those increases experienced in fiscal years 2008 and 2007, respectively. With respect to the delinquencies as a percentage of the total receivables portfolio, even though increases were reported in fiscal year 2008 relative to 2007, decreases were actually reported in fiscal year 2009 relative to 2008. As collateral performance continues to normalize, credit risk associated with the related transactions has begun to decline.  

Credit Enhancement

Increases in actual and expected losses also led to adjustments in the levels of credit enhancement employed in prime automobile loan ABS. Credit support on issuances occurring during and after Q1 2009 was primarily in the form of subordination and overcollater-alization. About 50% of the automobile loan ABS issuances occurring during this time had a senior-subordinate note structure, where credit enhancement on the AAA-rated notes is provided by a combination of overcollateralization and subordinated note classes. The majority of issuances with this structure occurred in Q4 2009 after spreads had tightened enough to allow for the placement of the subordinated classes. The remaining 50% relied simply on overcollateralization. In the vast majority of the transactions, the most junior note or certificate was initially retained by the depositor or an affiliate of the depositor.  

There is considerable variation among securitization platforms with respect to levels of overcollateralization.1 However, with a high degree of regularity, prime automobile loan ABS issuers incorporated significantly more overcollateralization in deals completed during 2009 than in prior years. In some cases this even surpassed the levels that were required by ratings agencies during this time. This trend, however, has already begun to reverse: transactions completed in Q4 2009 and Q1 2010 have generally involved overcollater-alization levels measurably lower than their peak levels from 2009. For example, Ford Credit Auto Owner Trust (“FORDO”) transactions increased from an overcollater-alization level of 15.20% during Q1 2009 to a peak overcollateralization level of 17.16% in Q3 2009 before dropping to 9.15% in Q2 2010. Similarly, the HAROT transactions increased from an overcollateralization level of 3.00% during Q1 2009 to a peak overcollaterali-zation level of 4.25% in Q4 2009 before dropping to 3.00% in Q2 2010. Comparable patterns are observable across other prime automobile loan securitization platforms.  

Interest Spreads

Since the implementation of the TALF program, spreads on prime automobile loan ABS have progressively tightened. For example, the HAROT transaction pricing in January 2009 had a spread of 375 bps on the last maturing class of its AAA-rated notes. Spreads on its next two offerings declined to 250 bps and 135 bps, respectively, on the last maturing class of AAA-rated notes in each of those Q3 2009 transactions. Its Q1 and Q2 2010 offerings boasted spreads of 20 bps and 25 bps, respectively, on their last maturing AAA-rated classes—levels comparable to those seen in late 2007. Similarly, the last maturing class of AAA-rated notes in the FORDO platform had a spread of 425 bps in its Q1 2009 offering, a spread of 235 bps in its Q2 2009 transaction, spreads of 235 bps and 105 bps in its Q3 2009 transactions, a spread of 70 bps in its Q4 2009 transaction and a spread of only 25 bps in its Q2 2010 transaction. At differing levels, these patterns are observable across prime automobile loan securitization platforms.  

Improved collateral quality and higher levels of overcollateralization have begun to address investor concerns relating to credit risk on recently issued automobile loan ABS. Coupled with the availability of TALF funds, these developments have witnessed a return of investor demand for prime automobile loan ABS—since the implementation of TALF, there have been 41 public offerings of prime automobile loan ABS, a marked increase over 2008. It is significant that most issuances during Q1 2010 were not TALF eligible, indicating that the increased demand had now become independent of the ability to borrow low cost funds through the TALF program. This heightened demand has brought down spreads and has begun, together with improved origination standards, to permit lower levels of overcollateralization on the most recent offerings. If this combination continues to reduce issuers’ cost of funds, 2010 should see more transactions come to market.  

Subprime Issuances

While the volume of prime automobile loan ABS deals began to pick up, and the spreads on these deals began to tighten, relatively soon following the implementation of the TALF, it was some time before subprime issuers utilized the program. Subprime platforms had traditionally relied on credit support provided by note insurance policies as a relatively low-cost means of obtaining AAA ratings. Following the collapse of the credit market and the subsequent downgrade of most note insurers, going to market with a monoline wrap ceased to be an option. Subprime issuers were therefore confronted with either relying solely on overcollateraliza-tion for credit enhancement or on some combination of overcollateralization and subordinated classes of notes. As was the case before the downturn, the level of overcollateralization required for AAA ratings on notes backed by subprime automobile loans would have been too high to be an option. While turning to a senior-subordinated structure was a possibility, it could only be viable if the interest paid on the subordinated notes were low enough to make the overall cost of funds palatable. In 2008 and much of 2009, however, interest rates on subordinated classes were prohibitively high. It was, therefore, not until Q4 2009 and Q1 2010, after the effects of the TALF on prime transactions had begun to spill over into the automobile loan market at large, that subprime issuers returned to the market in numbers. In those two quarters five separate issuers (Credit Acceptance Corp., Prestige Financial Services, Drive Time, Tidewater Finance and AmeriCredit Financial Services (“AmeriCredit”) offered notes backed by subprime automobile loans. The subprime offerings during this time began to reflect the general tightening of spreads seen in the prime automobile deals. Prior to this period, there had been only one public offering of subprime automobile loan ABS in 2009: AmeriCredit’s AMCAR 2009-1 transaction. And in contrast to the later cluster of TALF-eligible subprime deals, the elevated spread and overcollateralization requirements suggested that the market remained apprehensive about subprime automobile loan ABS.  

Indeed, the experience of AmeriCredit’s subprime platform, AMCAR, is illustrative of the challenges faced by subprime issuers following the collapse of the market and during its recovery under the TALF. In 2007, AmeriCredit brought four AMCAR offerings to market, for a total of $5.2 billion in offered notes. Spreads on the last maturing class of AAA-rated notes in these deals ranged from 4 bps at the beginning of the year to 95 bps towards the end of the third quarter, when the impact of the credit crunch had begun to affect spreads. All of these deals were wrapped and all relied on a target level of overcollateralization of 11% in addition to a 2% cash reserve account. These tight spreads (at least at the beginning of the year) and low levels of overcollateralization were consistent with the experience of other issuers in early to mid 2007 in a market flush with subprime offerings.  

Although there were three AMCAR offerings in 2008, the aggregate of the notes was only approximately $1.75 billion and spreads on the last maturing class of AAA-rated notes in these deals ranged from 365 to 500 bps. Furthermore, only one of these deals was wrapped. The targeted overcollateralization level for the wrapped deal had more than doubled relative to 2007 levels; for the senior/subordinated deals it had roughly tripled. On top of the high levels of overcollateralization, the senior/sub deals faced interest rate yields as high as 18% on the most junior class of subordinated notes. Moreover, it was impossible to speak of a market for subprime automobile loan ABS in 2008—aside from the AMCAR transactions, there was only one publicly offered subprime automobile loan securitization during that year.  

In July 2009, AmeriCredit’s AMCAR 2009-1 transaction was the first publicly offered, TALF-eligible subprime automobile loan ABS. In terms of deal size, this offering’s $725 million of offered notes roughly split the difference between the AMCAR deals immediately prior to the credit crisis and those following it. Although this transaction saw an increase in credit enhancement levels as compared to 2008, with a targeted overcollat-eralization level of 38.00%, its pricing had improved. And while the spread of 175 bps on the last maturing class of AAA-rated notes was nowhere near 2007 levels, it did represent a considerable improvement over the position in 2008. The impact of the TALF was evident as spreads on this deal had begun to benefit from the general tightening of spreads already witnessed in prime automobile loan ABS deals.  

Much, however, had changed by the time AmeriCredit completed its second TALF-eligible subprime automo-bile loan securitization in February 2010 (AMCAR 2010-1). The economic terms of the transaction painted a much brighter picture. The spread on the last maturing class of AAA-rated notes was a modest 65 bps. The targeted overcollateralization level of approximately 23.25% was substantially reduced from comparable structures in 2008 and 2009. Most importantly from a structuring position, the yield on the most junior class of subordinated notes was a manageable 6.75%.  

In the two months following the expiration of TALF-eligibility for automobile loan ABS, AmeriCredit has brought two additional subprime deals to market. AmeriCredit’s recently priced AMCAR 2010-2 transac-tion reflected a continuation of the generally positive trends seen in AMCAR 2010-1. While AMCAR 2010-2’s deal size ($585.6 million of publicly offered notes plus $14.4 million of BB-rated Rule 144A notes) remained below pre-credit crisis levels, the spread on its last maturing class of AAA-rated notes (55 bps), the interest on its most junior class of publicly offered subordinated notes (6.24%) and particularly its targeted level of overcollateralization (15.25%) represented improve-ments over the comparable figures for AMCAR 2010-1. AmeriCredit’s AMCAR 2010-A transaction was its first wrapped deal in nearly two years. With its relatively small deal size ($200 million of offered notes) and reliance on a note insurance policy, this transaction deviated somewhat from the pattern of recent AMCAR deals and may therefore be of limited value for purposes of comparison. That said, while the spread of 175 bps on its last maturing class of AAA-rated notes was higher than either of the other 2010 AMCAR deals, it was still considerably lower than any post-crisis, pre-TALF AMCAR transaction.  

What developments in AmeriCredit’s subprime transactions can be observed in this progression towards a normalized subprime market? One obvious change has been the move from note insurance to senior/subordinated structuring as the main form of credit support. This was necessitated by the elimination of virtually all monolines as viable wrap providers. Counter to this trend, the AMCAR 2010-A transaction relied on a note insurance policy. The pricing on this transaction would seem to suggest a degree of caution in the market about the return of wrapped deals; it remains to be seen whether this transaction presages a real return of the wrapped structure. Another less obvious development has been the general improvement in collateral quality. Following the market collapse, AmeriCredit reduced its targeted origination volume through tightening credit by raising the cut-off on its custom credit score. These have led to collateral pools with higher average obligor credit scores, lower average LTV ratios and shorter weighted average terms to maturity. Performance of the pools has also benefited from the gradual stabilization of the overall economy.

Together, these trends have helped to alleviate some of the credit concerns weighing on automobile deals generally and the subprime market in particular.  

In combination, these developments bode well for the return of an active market for subprime automobile loan ABS. With spreads returning to levels not seen since late 2007, issuers long absent from the subprime space have begun to test the viability of securitized borrowing. For instance, Santander Consumer USA recently consummated a public offering of $1 billion of subprime auto ABS—its first public offering since October 2007. While it is still too early to tell, the ability of AmeriCredit to come to market with its wrapped AMCAR 2010-A transaction could portend the return of an important element of the subprime market, or at least the availability of an additional structuring option for prospective issuers. More importantly, with private placements of subprime automobile loans by Credit Acceptance Corp., Prestige Financial Services and Drive Time in Q4 2009, and a private placement by Tidewater Finance and public offerings by AmeriCredit and Santander Consumer in 2010, it may finally be possible to once again speak of a market for subprime automo-bile loan deals.