Some market participants seem to be overreacting to the recent rating activity in the subprime mortgage ABS sector. Together, Moody's Investors Service and Standard & Poor's took more than a thousand negative rating actions on subprime mortgage ABS tranches on July 10. S&P's actions covered roughly $7.4 billion of subprime mortgage ABS and Moody's actions covered around $5.2 billion.
At first blush, the volume of securities affected by the recent rating actions seems very large. However, the volume is actually quite small compared to the $480 billion of home equity ABS issuance in 2006 and the $508 billion of issuance in 2005. Additionally, the rating actions were reasonably foreseeable and strongly concentrated in the lower-tier tranches of the deals - tranches carrying ratings mostly in the triple-B range. Indeed, watchlistings and downgrades on triple-B-rated tranches should not be particularly surprising during periods of stress.
However, the subprime rating actions caused a spillover to the CDO sector. The rating agencies took hundreds of negative rating actions on CDOs that have exposure to subprime mortgage ABS. Interestingly, because they use derivatives (i.e., credit default swaps), CDOs have a much larger exposure to the triple-B tier subprime mortgage risk than was created in the actual ABS transactions.
Uncertainty of Projected Losses
The recent rating actions do not alleviate the challenging uncertainty about the total level of losses that recent (i.e., 2006 and late 2005) vintages of subprime mortgage loans will ultimately suffer. Available performance data suggests that the recent vintages will suffer higher cumulative losses than any other vintage on record. However, it is not clear how much higher the losses might be. The available data can reasonably support estimates ranging from 7% to more than 15%. At the low end of the range (e.g., 7% to 9% losses on the pools), credit enhancement in the deals could be more than enough to protect most triple-B-rated tranches. However, at the high end of the range (e.g., 12% to 15% losses), many triple-B-rated tranches would likely suffer 100% principal loss.
Predicting the ultimate level of losses with precision is not yet possible because most of the loans will not reach their "resets" until late this year or next year. The reset on a loan is when it converts to having an adjustable interest rate after an initial fixed-rate period of two (or sometimes three) years. The reset usually causes the amount of the borrower's monthly payment to increase significantly. Many borrowers handle the reset by refinancing their loans. Some may be unable to refinance immediately, but are able to endure the higher monthly payments until they eventually qualify for refinancing. Others spiral into default, either immediately after the reset or after a period of struggling to keep up with the higher payments (and while neglecting necessary repairs and maintenance on their homes).
Credit Availability - Home Prices - Interest Rates (C-P-R)
The availability of credit, home prices, and the level of interest rates will likely be the key factors that determine what proportion of subprime borrowers is able to refinance successfully when their loans reset. Reliably predicting the "tightness" or "easiness" of the credit environment is a daunting challenge. Policymakers on both sides of the aisle are calling for tighter subprime mortgage lending standards. The guidelines recently adopted by the federal banking regulators will likely cause a measure of tightening. In addition, proposed new laws at both the state and federal level could tighten subprime credit even further. The greater the degree to which credit tightens, the larger the proportion of subprime borrowers who will be unable to qualify for refinancing loans. Likewise, projections for home price appreciation differ widely among market experts. Opinions run the range from sharply negative to slightly positive, and there seem to be as many different opinions as there are experts. Naturally, falling home prices would make refinancing more difficult for many borrowers because they would not have sufficient equity in their homes to qualify for refinancing loans. Third, the level of interest rates at the time a loan reaches it reset also affects the borrower's ability to refinance. If interest rates rise, a larger proportion of borrower's may be unable to qualify for refinancing loans because they cannot meet the required debt-to-income ratios.
The loans of the late 2005 and early 2006 vintages will not start to reach their reset dates until late this year. Moreover, even after a loan reaches its reset date, it can take several more months before the final outcome (refinance versus default) becomes apparent. Therefore, a clear picture of the fate of the first wave of loans (i.e., the late 2005 vintage) will not begin to emerge until sometime in 2008. Accordingly, home price appreciation, the tightness of credit, and the level of interest rates in late 2007 and in 2008 will likely decide the fate of the late 2005 and 2006 subprime vintages.
In the meantime, uncertainty makes for a very difficult market. Valuations of 2005 to 2006 vintage subprime mortgage ABS initially rated in at the triple-B level are highly sensitive to the projected level of losses. Because the tranches tend to be very thin, they can resemble "all or nothing" bets: "all" if losses or low and "nothing" if losses are high. That creates a somewhat unusual (and vexing) performance profile for a fixed-income security. Moreover, because reliable loss projections are not possible, analytic valuations become uncertain. At best, they represent an averaging of extreme outcomes. Fixed-income professionals become frustrated that their typical analytic tools cannot produce reliable valuations based on fundamental analysis. Instead, supply-and-demand and market sentiment become the main drivers of pricing. Investors who want to trade out of a position find mostly bids that correspond to projected losses at the pessimistic end of the reasonable range.
On the other hand, the picture is starting to become clearer for some of the tranches originally rated in the triple-B range. Those are the tranches for which the delinquencies on the underlying pools are already so high that the outcome of the loan resets may not matter. In other words, some tranches likely will be killed just from the reasonably expected losses from the present pipeline of delinquencies. For now, however, most tranches must still wait for the resets to learn their ultimate fate.
 The "home equity" ABS category includes ABS backed by (i) subprime mortgage loans, (ii) closed-end second-lien loans, (iii) home improvement loans, (iv) home equity lines of credit, (v) so-called "high LTV" mortgage loans, (vi) reperforming mortgage loans, and (vii) non-performing mortgage loans. In recent years, deals backed by subprime mortgage loans have composed roughly 85% to 90% of the total activity in the home equity ABS category.
 In contrast, triple-A-rated tranches of subprime mortgage ABS are often protected from losses up to 25% to 30%, or more.
 The "thickness" of a tranche in a securitization transaction is often described as the ratio of the size of the tranche to the size of the underlying pool. In subprime mortgage ABS, each of the three tranches in the triple-B range typically has a thickness in the range of 1.2% to 1.5%.
 The reason for such high delinquencies so early in the life of the loans remains somewhat unclear. Fraud by both borrowers and mortgage brokers is likely one of the causes, but it may not be the only one.
Mark Adelson is an independent consultant on securitizations. He previously worked at Nomura Securities as head of structured finance research, where he reported to David Jacob.
David Jacob provided essential insights and comments during the preparation of this article.
(First of a two-part series)
(c) 2007 Asset Securitization Report and SourceMedia, Inc. All Rights Reserved.