> > > >
The recent crisis in the housing market has policymakers scrambling to craft programs to deal with its continuing fallout. Much of the effort has been focused on “nonprime” mortgages, which comprise subprime and Alt-A loans. The former are mortgages made to borrowers with some flaws in their credit history, while the latter are generally larger loans made to those who are more creditworthy but choose not to provide the income or asset verification necessary to attain a prime mortgage. Both types of mortgages are typically higher cost than prime loans.
While the dramatic rise in the default of nonprime mortgages has been documented, the cause of the defaults, especially early in the loan, remains a subject of investigation. Such an understanding will assist in crafting measures that address the situation in the short term and help in the structuring of long-term solutions that prevent its reoccurence. Andrew Haughwout, Richard Peach, and Joseph Tracy weigh in with a recent study that centers on two possible explanations: relaxation of underwriting standards and changes in economic forces. The following is a summary of their study.1
Marvin M. Smith, Ph.D., Community Development Research Advisor
The authors note that traditionally there are several risk factors (or underwriting criteria) relied upon to gauge the probability that an individual will default on a mortgage. Those factors include the loan-to-value ratio (LTV),2 the debt-service-to-income ratio (DTI),3 the borrower’s credit score,4 and the degree to which a borrower’s income and assets are verified independently through sources such as employers, tax returns, and bank account statements. They further point out that in an effort “to expand the potential pool of borrowers, nonprime (subprime and Alt-A) mortgages by design relaxed one or more of these underwriting criteria beyond the margins required for prime mortgage loans.” As a consequence, the authors indicate that we would expect the default experience of nonprime loans to be worse than that of prime mortgages. The authors report that “industry data confirm that the performance of the very first vintages of nonprime loans was significantly worse than that of prime loans.”5 They reveal that, starting with the 2005 vintage, the performance of nonprime mortgage loans became markedly worse than prior vintages. “By 12 months following origination, the 2005 vintage had a 90 day or more delinquency rate that was not reached by the 2003 vintage for 20 months, and the 2006 vintage at 12 months had a rate that was not reached by the 2003 vintage even by 30 months.” According to the authors, this precipitous decline in loan performance was puzzling to investors in these mortgages, since the observed risk factors failed to fully explain the trend.
The authors focus their attention on mortgages that exhibit defaults very early in the life of the loans, which they characterize as “juvenile delinquents.”6 In addition, they define an “early default” as a mortgage that is 90 days or more past due during the first year following origination. The authors investigate how much of the sharp rise in early defaults of the 2005 through 2007 vintages of nonprime mortgages can be explained by changing underwriting standards over time (i.e., “bad credit”). They further note that many housing markets experienced a peak in housing sales in late 2005, which was eventually followed by a decline in housing prices. Thus, the authors “also explore the extent to which house price dynamics over the housing cycle as well as other economic factors help explain the early default behavior of the more recent vintages of nonprime mortgages (i.e., ‘bad economy’).”
The authors draw their mortgage data from LoanPerformance, a San Francisco company. The data provide loan-level information (on a monthly basis) on roughly 7 million active, securitized subprime and Alt-A loans. They use a 1 percent random sample of firstlien nonprime loans, which yields 115,000 loans for analysis. They also add other economic data, such as measures of house price appreciation and labor market conditions.
First the authors examine tabulations on nonprime mortgages, with special attention paid to the early defaults in the mortgages as they relate to the various risk factors mentioned above. Then they use regression analysis to explore the determinants of early default.
The authors considered the distribution of early defaults by initial LTV ratio and year for subprime and Alt-A mortgages. They observed that the incidence of early defaults more than quadrupled for both types of mortgages from 2003 to 2007. Also, “for any given range of LTV the early default rate for subprime mortgages tended to be higher than for Alt-A mortgages.”
Likewise, a distribution of early defaults by the DTI ratio from the sample data over the same period showed an increase in the incidence of early defaults for both subprime and Alt-A mortgages but with a relatively common change across DTI intervals.
The distribution of early default rates by FICO scores overtime revealed that early defaults in each year typically decline as FICO scores increase. Moreover, except for subprime mortgagors in 2006 and 2007, “borrowers with a FICO score of less than 600 are at least three times more likely to experience an early default as borrowers with a FICO score of over 660.”
Finally, the authors classified the underwriting of the nonprime mortgages in the sample as full documentation, low documentation (‘limited-doc”), and no documentation (“no-doc”). The resulting distribution over time revealed three noteworthy findings. “Despite the focus in the press made on no-doc mortgages, in each year the incidence of no-doc mortgages was in single digits, and was declining over the sample period.” But the authors thought that an even more notable discovery was the shift in underwriting from fully documented to limited documented. The tabulations showed that the share of fully documented subprime mortgages fell from 77.8 percent in 2001 to 61.7 percent in 2006, while fully documented Alt-A mortgages declined from 36.8 to 18.9 percent over the same period. The distributions also revealed that “in each year for subprime mortgages, early defaults are more prevalent for limited as compared to fully-documented mortgages,” while the incidence of early defaults in each year for Alt-A mortgages “generally increases as one moves from fully-documented to limited doc mortgages, and from limited doc to no-doc mortgages.”
The regression results yielded valuable insights on the effects of key variables on the probability of an early default (ED) for nonprime mortgages. These are some of the findings:
Further statistical analysis centered on examining “the question of the relative importance of credit effects versus economy effects in explaining the sharp rise in early defaults.” Their “results suggest that while both of these factors—bad credit and bad economy—played a role in increasing early defaults starting in 2005, changes to the economy appear to have played the larger role.” However, the authors hasten to add that their estimating model “predicts at most 43 percent of the annual increase in subprime early defaults during the 2005-2007 period.” While the authors have added considerably to our understanding, much is left unexplained—something they are currently working to rectify.