Three to four years removed from the time non-Agency RMBS and housing markets began to collapse, this article provides a historical performance update of the subprime loan sector for vintage years 1997 through 2008. Performance of the 2005, 2006, and 2007 vintages is far and away worse than for any other year, even prior to 2000 when subprime lending was still fairly new. In addition, the extent of negative equity in the borrower repayment pipeline for never-modified and at-least-once-modified subprime loans for the 2005, 2006 and 2007 vintages is measured. Given the extreme and broad corrections in residential property prices since approximately mid-2006, the mortgage repayment struggles of borrowers, and the intended effects of loan-modifications, negative equity density has been pushed to the performing and early-stage delinquency end of the pipeline. That is, modifications have resulted in a larger share of performing (active and current) borrowers with significantly more negative equity than otherwise would be the case had a loan modification not occurred.

Subprime Update

Across the subprime sector, CDRs are on a general downward trend and voluntary prepayments are negligible. CDRs for 2005, 2006, and 2007 loans have returned to more normal 5% to 10% levels. 60+ delinquencies are on a downward trend, especially for the 2005, 2006 and 2007 cohorts. Although 60+ delinquent loans are trending downward, a large proportion is transitioning into later-stage foreclosures.

Since 2003, performance of bankruptcy rates has been very similar. Rates prior to 2002 peaked at more than double the levels of after-2002 vintages. In the past year and a half rates on 2004 and 2005 vintage loans have risen after having leveled around 3.5%. Improved performance can partly be attributed to economic recovery and expansion following the 2001 recession and to the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005.

Without exception, FCs are currently rising. The absolute levels of 2005, 2006 and 2007 rates are a major concern. Notably, the tails of the 2000, 2001, 2002 and 2003 cohorts have also increased markedly over the past year or two. This is undoubtedly the result of long-drawn-out high unemployment rates and significant negative equity effects.

REOs are also on a general downward trend with 2005, 2006 and 2007 vintages having returned to normal levels on the seasoning scale. However, the large buildups in foreclosures will - short of major concessions (bank, servicer, policy, modifications) - eventually begin to drive REOs higher.

For 2005, 2006 and 2007, then, inter-temporal behavior of CDRs, 60+ delinquencies, FC, and REO can largely explain the behaviors of cumulative liquidations. And the broad, ongoing home price corrections no doubt account for the majority of the absolute levels and differences in levels in cumulative severities. The cumulative net loss rate series for 2005, 2006 and 2007 vintage years again reflect the magnitude of impacts from home price depreciation, unemployment and lost income, and the negative effects of the so-called affordability mortgages issued.

Approximately 85¢ of every $1 of the $375 billion (plus) outstanding subprime loan debt is from 2005, 2006 and 2007 issuance where cumulative severities and cumulative net losses are also the worst performing. The next section focuses exclusively on these vintages.

Negative Equity

This section presents the pipeline distribution of negative equity for never-modified and at least once-modified loans as of August 2011. The ABSNet Loan data set includes 1,605,873 outstanding 2005, 2006, and 2007 vintage loans, reflecting an actual outstanding balance of $292 billion. 497,128 have been modified and have a total actual outstanding balance of $105,265,160,453. 943,121 loans, 59%, are upside-down. And there is $30.8 billion of negative equity in the non-performing pipeline. This is equivalent to 10.5% of the outstanding actual ending balance.

It is evident that loan modifications have resulted in a redistribution of negative equity toward the performing "Active/Current" end of the pipeline. However, the aggregate of the 150+ buckets comprises 87% of all negative equity in the non-performing pipeline for modified loans. Thus, degrading performance of Active/Current modified loans can be expected.

In passing, our research also highlights the importance of positive equity borrowers. For the Active/Current Pristine group, Never-Modified Positive Equity of 7.2% is 1.8 times the negative equity percentage for the same class. This is testimony to the notion that having "skin in the game" as it relates to housing and mortgage purchases is vital.


Performance of sub-prime RMBS within the 2005, 2006 and 2007 vintages has been dreadful. Despite improvements in CDRs, 60+ delinquency rates and REOs, foreclosures continue to increase in all vintages. Further, liquidations, net losses, and cumulative severities continue to rise. Loan modifications have increased in importance as a means for helping distressed and non-distressed borrowers cope with quite bleak economic and housing conditions. As of August 2011, there is approximately 1.9 times the amount of negative equity in the modified borrower performing versus the non-performing pipeline. Conversely, there is approximately 0.76 times the amount of negative equity in the never-modified borrower performing versus the non-performing pipeline. Whether this redistribution of negative equity to the performing end of the borrower pipeline through modifications will have a positive outcome is yet to be decided. Re-default rates on modified loans among 2005, 2006 and 2007 borrowers is at 25%, and weighted average months from modification to re-default is about 11. A year from now, a re-evaluation of the negative equity pipeline might give a good indication of the eventual success of loan modification programs.

Subscribe Now

Access to a full range of industry content, analysis and expert commentary.

30-Day Free Trial

No credit card required. Access coverage of the securitization marketplace, including breaking news updated throughout the day.