30 Pages Posted: 12 Jan 2009 Last revised: 7 Aug 2009
Date Written: January 9, 2009
The subprime foreclosure crisis has resulted in residential mortgage debt burdens far beyond what borrowers can repay. Many economists have recognized the need to deleverage the American homeowner. The excess mortgage debt is depressing home prices and consumer spending, and acting as a drag on the broader economy. Empirical evidence from mortgage servicer reports to investors show that for the most part, the necessary deleveraging of homeowners is not happening.
In a prior paper I compiled mortgage foreclosure and modification data from a sample of securitized subprime mortgage pools. This paper reports on an expanded study of data on more than 3.5 million subprime and alt-A mortgages, including about one-sixth of all foreclosures pending, and about 20% of the monthly total modifications in November 2008.
My findings from both the smaller sample and the larger sample are consistent, and are as follows:
1. Modifications are not reducing principal debt, they are increasing it. Almost no modifications include significant cancellation of either past due interest or principal, and many modifications involve capitalizing unpaid interest and fees and reamortizing the loan. This occurred in 68% of loan modifications. The average capitalized amount added to loans was $10,800, on average mortgage debt of about $210,000. Some principal was canceled, and reported as a partial loss, for about 10% of modifications. However, two servicers (out of 43), Litton Loan Servicing and Ocwen Loan Servicing, accounted for nearly all of these principal reductions. Only 8% of loans reflected some write-off of unpaid interest.
2. Servicers are incurring huge losses for investors by foreclosing. The average foreclosure loss on a first mortgage in November 2008 was $145,000 or about 55% of the average amount due. Loss severities increased steadily throughout 2007 and 2008 and are expected to worsen in 2009. In these circumstances, rational investors should accept mortgage principal reductions corresponding to home value declines of 20% or so, were it not for the various obstacles to servicers' restructuring of mortgage loans.
3. Voluntary mortgage modifications are not consistently reducing monthly payment burdens. Only 49% of modifications in the November 2008 report reduced monthly payments below the initial payment, while 17% left the payment the same and 34% increased the monthly payment.
4. The variations among servicers in the number and quality of modifications are enormous. This variation suggests that not every servicer is doing the maximum possible to reach and work out terms with every defaulted borrower.
5. Fewer than half of modifications made in January 2008 were current in payments on November 25, 2008. This is not surprising, given the onerous terms of the 2008 modifications.
6. Many modifications are temporary. For example, some adjusted interest rate and amortization term only for five years, with rate and payment increases after five years. Servicers also use balloon payments and other forms of deferrals in order to reduce payments without reducing total debt. Thus, the totals reported by the industry include many loans that are being modified to include deferred payment shocks, negative amortization or other non-amortizing features of the sort that caused the foreclosure crisis.
7. Significant numbers of mortgage loans are seriously delinquent, but not in a modification program or in foreclosure. The foreclosure crisis is overwhelming the ability of servicers to either restructure or foreclose on all the delinquent loans.
The paper discusses the many reasons that necessary mortgage restructuring is not happening and proposes several policy responses.
Keywords: subprime, foreclosure, financial crisis, mortgage modification
JEL Classification: K20
Suggested Citation: Suggested Citation
White, Alan M., Deleveraging the American Homeowner: The Failure of 2008 Voluntary Mortgage Contract Modifications (January 9, 2009). Connecticut Law Review, Vol. 41, p. 1107, 2009. Available at SSRN: https://ssrn.com/abstract=1325534