Throughout the recent foreclosure crisis, lenders and policymakers alike have searched for effective ways to reduce mortgage foreclosure among underwater borrowers. Most of these attempts, such as the government sponsored Homeowner Affordability Modification Program (HAMP), have focused on modifying mortgages to more favorable terms for the borrower. These interventions have yet to yield a satisfactory result. One reason modifications have been ineffective stems from the difficulty in finding ways in which a modification is beneficial to both the lender and the borrower, both of whom must agree to any changes in the mortgage terms. In August 2010, OCWEN Financial Corporation began a pilot program which experimented with a new solution to this problem: Shared Appreciation Modifications (SAMs). The OCWEN SAMs are mortgage modifications that reduce the outstanding balance on a mortgage until the borrower is no longer underwater. If the home gains value in the future, the SAM entitles OCWEN to a portion of the gain once the home is sold. Given that economic models of mortgage default cite being underwater as a necessary condition for default, SAMs hold the potential to drastically reduce default rates in a way that benefits both the borrower and the lender.

This paper examines whether OCWEN SAMs offer a true pareto improvement for both parties involved. One inevitable problem that arises from evaluating a program of this sort is finding an appropriate control group. Ideally, we want to find a set of borrowers who are identical except that some randomly received SAMs and others did not. Since OCWEN did not randomly assign SAMs, we control for selection effects by restricting our dataset to a common set of borrowers. In particular, we use a unique dataset that provides loan-level data to track mortgages from origination to present. The dataset is managed by the Corporate Trust Services (CTS) of Well Fargo Bank and is matched to HAMP data. The dataset is then restricted to those borrowers who entered and then subsequently dropped out of the HAMP program. A key assumption is that borrowers who decided to enter HAMP and did not make the required monthly payments were otherwise identical aside from the owner of their mortgage. The outcomes of the mortgages owned by OCWEN are compared to those of other servicers to determine how much improvement in performance comes from the SAMs. Initial results suggest SAMs significantly reduce mortgage default rates, however, a full cost-benefit analysis of the program depends crucially on future house price expectations.