Securitization and Moral Hazard

Perhaps no academic paper has done more to convince scholars and policymakers that mortgage securitization led to lax screening by lenders and fueled the subprime crisis than did the recent paper by Keys, Mukherjee, Seru, and Vig (forthcoming in the Quarterly Journal of Economics, 2010) (hereafter, KMSV, who published a post in June on the Forum, which is available here). In an innovative paper, they argue that mortgage purchasers follow a “rule of thumb” in deciding which loans to purchase: they are, for exogenous reasons, much more willing to buy mortgage loans given to borrowers with credit scores above 620 than those given to borrowers with credit scores below 620. In a dataset containing only securitized loans, they find that loans made to borrowers just above 620 (where securitization is easy) default at a higher rate than those just below and argue that this is strong evidence that securitization really did result in lax screening by lenders.

In a new paper, Securitization and Moral Hazard: Evidence from Lender Cutoff Rules, which we recently presented at the Harvard Business School / Harvard Economics Finance Lunch Seminar, we reexamine the credit score cutoff rule evidence with a better dataset and through a theoretical lens that assumes rational equilibrium behavior and reach a very different conclusion.

First, we argue that credit score cutoff rules used by lenders emerge endogenously when there are fixed costs in investigating loan applicants. Lenders must choose for which borrowers to bear these fixed costs, and under the natural assumption that the benefit to lenders of collecting additional information is greater for higher default risk applicants, lenders will only collect additional information about applicants whose credit scores are below some cutoff.

Second, we examine a large loan-level dataset that contains both portfolio loans and securitized loans and show that for several key subsamples, including the subsample used by KMSV, there is a jump in the default rate but no discontinuity in the rate of securitization at the 620 cutoff. Our findings demonstrate that the jump in defaults at 620 documented by KMSV is not evidence that securitization led to lax screening.

Third, we consider the implications of our theory for securitization and argue that the discontinuity in screening behavior by lenders induces a discontinuity in private information; lenders generate more information about borrowers below the 620 cutoff. One potential response by rational securitizers to this discontinuity is to reduce their loan purchases below the cutoff to maintain lenders’ incentives to screen loan applicants. We look to the data and find that, in markets populated by smaller private-label securitizers, securitization rates are indeed lower below the lender cutoff than above. The response of securitizers to the lender cutoff rule suggests they were aware of the threat of moral hazard and took steps to mitigate it.

Though our findings suggest that loan purchasers were more sophisticated with regards to the moral hazard problem posed by securitization than previous research has judged, the extent to which securitization contributed to the subprime mortgage crisis remains an open and pressing research question.