With the housing bust and foreclosure crisis of 2008, much attention has been focused on the relationship between mortgage servicers and delinquent mortgage borrowers. Specifically, the focus has centered on negotiations, or the lack-thereof, between these two parties in the period between delinquency and the initiation of foreclosure proceedings. During this period, the servicer has the option to forego filing foreclosure proceedings, and to instead adjust the terms of the mortgage contract in a manner that increases the probability of future repayment. This process is called modification, and a mortgage servicer, acting on behalf of a profit-maximizing mortgage lender will make this decision by weighing the expected costs, which include the costs of foreclosure (such astaking possession of the house, fixing any property damage, and re-selling the property) against the expected benefits, which include an increased likelihood of repayment by the borrower.
To-date there have been very few modifications, relative to the number of foreclosures performed. As a result, there has been an outcry on the part of policy makers, consumer advocacy groups, as well as some prominent academic economists, who believe that there are not enough loan modifications being performed by the private market. They point to the presence of institutional frictions in the mortgage securitization process as one explanation for this discrepancy. The perceived frictions relate to the incentives that servicers face when passing mortgage payments through to the ultimate investors. As a general rule, servicers are paid a regular fee for every mortgage they service in a given month. Absent any restriction on their activities, servicers would thus have an incentive to keep troubled loans active no matter what, by making large and indiscriminate modifications that would favor borrowers who faced little probability of default. Such modifications, of course, would also reduce the income to the investors, so the legal agreement between the servicer and lender, called the pooling and servicing agreement (PSA), prohibits this type of behavior. As a general rule, PSAs allow servicers to make modifications, but only in cases where default is likely and where this can be shown with a net-present-value calculation.
Thus, one of the popular explanations for the low incidence of modifications among mortgage servicers is the legal restrictions embedded in the PSAs. It is important to note here, that there are two potential principal-agent agent problems that could be relevant. One exists between the servicers and the bondholders as a whole, and was described above. Another concerns the different tranches of investors in a given MBS. Because the different tranches receive payments from mortgages with different priorities, a loan modification could advantage one tranche more than another. For example, the holder of a AAA tranche might prefer to foreclose on a borrower in default, because this investor is paid first out of the proceeds of the foreclosure sale. Investors in the BBB or equity tranches could be much less likely to get anything from a foreclosure sale, so these investors might be more willing to gamble on a modification even if the modification reduces the net present value of the loan. Thus, the incentives of the different tranches with respect to modifications may not be aligned (“tranche warfare”), and the servicer may forgo modifications for fear of being sued by at least one of these parties.
Either explicitly or implicitly, those that take this viewpoint, are arguing that modifications are in the interest of both the borrower and the mortgage holder, and absent any frictions involved in the process, they would be a viable and indeed, preferable alternative to foreclosure. But, in the absence of systematic and conclusive data, it is still an open question as to whether there really are profit-maximizing loan modifications that are not being made because mortgage servicers are afraid of being sued by the investors they work for. To ultimately answer this question, one would need detailed and comprehensive, loan level data on both mortgage originations, and monetary losses from foreclosure proceedings. Unfortunately, we do not have the latter piece of information, and so we cannot address the question of whether modifications would yield higher profits to the investor compared to foreclosure in a frictionless world. However, we do have detailed, loan-level data with which we can distinguish between securitized loans and mortgages held in portfolio. With this data we are able to first document that there are significant differences in the probability of transition of delinquent loans into foreclosure when we compare loans held in banks’ portfolios with privately securitized ones (as was previously shown in work by Piskorski, Seru, and Vig 2009).
These differences persist even after accounting for all available observable characteristics (including the full loan-to-value, which includes second-liens on a property). We then compare the relative frequency of modification between securitized and non-securitized mortgages, and thus can shed some light on the question of whether institutional frictions in the secondary mortgage market are playing an important role in generating differences in performance of delinquent loans held by different investors and in fact inhibiting the modification process from taking place.
Download
PDF Ebook Renegotiating Home Mortgages: Evidence from the Subprime Crisis
