Rolling the Dice on Foreclosure Prevention: Differences Across Mortgage Servicers in Loan Modifications and Loan Cure Rates

Abstract

Mortgage servicing has garnered increased attention since the foreclosure crisis. As the interface between borrowers and investors, servicers make the decision to either grant a loan modification or to foreclose. This study examines servicer loan modification practices for a national sample of delinquent subprime loans, and assesses the extent to which those practices are associated with foreclosures. The research reveals significant differences across servicers in loan cure rates, which are related to servicers’ propensity to offer loan modifications and to the level of relief offered to borrowers. The observed differences across servicers and the implications of this heterogeneity for foreclosure prevention underscore the importance of additional data, research, and policies that can increase the uniformity and transparency of servicing practices.

KEYWORDS:

• Mortgage servicing
• loan modifications
• foreclosures

Acknowledgments

This research was supported in part by the Center for Community Capital at the University of North Carolina, Chapel Hill. In addition, Carolina K. Reid gratefully acknowledges support from the Fisher Center for Real Estate and Urban Economics for her work on this project. The authors would like to thank Mark Zandi of Moody’s for providing data on servicer characteristics. We would also like to thank Laurie Goodman, Mark Calabria, Janneke Ratcliffe, Anthony Yezer, and two anonymous reviewers for helpful comments on previous drafts.

Notes

1. Dual tracking occurs when a mortgage servicer continues to foreclose on a homeowner's home while simultaneously considering the homeowner's application for a loan modification.

2. Servicer fees are not explicitly negotiated; instead, the fees are related to the yield on the MBS, which is negotiated between the seller of the MBS and the investor. The required yield on MBS at any given time is generally lower than the interest rate paid by the borrower. The difference between the two is the servicer fee (Cochran, Coffman, & Harless, 2007).

3. Alan White, for example, showed that the majority of voluntary modifications at the start of the crisis typically increased a borrower’s monthly payment, as well as the principal owed on the loan (White, 2009a, 2009b).

4. Borrowers are eligible for a HAMP modification on first-lien loans for owner-occupied properties with an unpaid principal balance of less than $729,750, originated on or before January 1, 2009.

5. HAMP also provides a bonus incentive of $1,500 to lender/investors and $500 to servicers for modifications made while a borrower is still current on mortgage payments but at imminent risk of default.

6. HAMP includes outreach and solicitation requirements; participating servicers must solicit all borrowers who become 60 or more days delinquent for a HAMP modification, and they are required to evaluate eligible loans using the standardized modification terms and the standardized NPV test (Holden et al., 2012).

7. The five banks signing on to the settlement were Ally/GMAC, Bank of America, Citi, JP Morgan Chase, and Wells Fargo. In addition, Bank of America, JP Morgan Chase, and Wells Fargo signed a separate settlement with the California attorney general to provide an additional $12 billion in relief to California homeowners.

8. See (Federal Reserve Board of Governors, 2014) for more details on the Independent Foreclosure Review.

9. The 2009 amendments to the Truth in Lending Act provide a safe harbor for servicers that modify a distressed loan, as long as that modification maximizes the loan’s NPV. In addition, the amendments specify that the duty to maximize the NPV of the mortgage is a duty owed to all investors, rather than to any one investor in particular, protecting servicers from competing obligations to different tranches of MBS investors (Levitin & Twomey, 2010).

10. As Levitin and Twomey (2010) point out, servicers’ compensation structures create significant principal–agent conflicts between them and the investors, to the detriment of delinquent borrowers who need a loan modification to prevent foreclosure.

11. Despite Ocwen’s modification efforts, it too has faced regulatory scrutiny for its servicing practices (CFPB, 2013).

12. "It’s more art than science," said Guy Cecala, publisher of Inside Mortgage Finance. "Who knows whether the borrower will default, what the value of the property is, what will happen to home values," he said. "I’m skeptical of all of it” (Andrews & Witt 2009).

13. The Community Advantage Program is a community reinvestment loan secondary market program initiated in 1998 through a partnership among the Ford Foundation, Fannie Mae, and Self-Help, which is a nonprofit lender with headquarters in Durham, North Carolina. Self-Help purchases qualifying loans from originating lenders and then resells them to Fannie Mae while retaining recourse.

14. The Community Advantage Program loans were primarily low down payment, 30-year, fixed-rate, purchase-money mortgages made at near-prime mortgage interest rates to borrowers whose income did not exceed certain thresholds.

15. In early 2011, Treasury released the first loan-level data on the HAMP program. However, 79% of active permanent modification records and 82% of trial modification records in the data file lack information identifying the race or ethnicity of the borrower. The handful of studies that do exist on loan modifications by borrower type have generally found no differences in the number or nature of loan modifications by race or ethnicity (Ambrose & Capone, 1996; Been et al., 2013; Chan, Sharygin, Been, & Haughwout, 2014; Collins et al., 2015; Mayer & Piven, 2012).

16. The Wells Fargo investor report files containing data on the loan pools are available at Wells Fargo’s CTSLink, www.ctslink.com.

17. The data do not allow us to see whether the decline in the balance is related to principal forbearance or forgiveness.

18. Collins and Urban (2014) found that state policies can influence servicer behavior; in Maryland, state-level reporting requirements for state-chartered servicers led to both more modifications and foreclosure filings than those not subject to the state rules. For this reason, we were hesitant to include servicers operating only in one or two states.

19. However, as we discuss in the Conclusion, some of the servicers in our sample were acquired or transferred to other institutions over the observed time period, suggesting that the consolidation of the servicer industry is an important area of future research.

20. The matching procedure was completed while one of the authors was working at the Federal Reserve Bank of San Francisco, providing access to the nonpublic HMDA data, which include origination date. CTS loans were matched to HMDA on site, then all identifying HMDA variables (including loan number) were deleted from the matched record, resulting in a CTS data file with race/ethnicity and income attached to each loan record, but no ability to regenerate the origination date or link to the public HMDA file.

21. Other studies that have used matching to merge HMDA data with loan performance records employ a probability matching technique so data on loans with multiple matches are not lost (Bocian, Li, Reid, & Quercia, 2011). We are not aware of any research that has compared and contrasted these methods and the strengths and weaknesses of the different approaches.

22. This is less than 0.5% of observations.

23. CFPB-proposed rule amending Regulation C to implement amendments to the Home Mortgage Disclosure Act (HMDA) made by Section 1094 of the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank), Section (1003.4(a)(1)(i)).

24. By definition, when a servicer modifies a loan, it returns to current and is considered cured. In a later analysis, we look at the likelihood that a modification results in a loan remaining cured or current for an extended period of time.

25. By definition, the modification rate is merely the overall cure rate minus the self-cure rate in Figure 1.

26. This category also includes a small percentage of Native American and other races.

27. The one exception to this trend is Servicer 18, which appears to have higher cure rates for loans held by borrowers who are Hispanic, and lower cure rates for loans held by borrowers who are Black, compared with Servicer 20. However, it is not more likely to see loans by cured by borrowers who are White.

28. We need to estimate a linear probability model instead of a hazard because there are several servicers with too few modifications to identify the effect on each servicer.

29. Unfortunately, because of data limitations, we are unable to effectively observe loan performance over a longer time period after modification. Additional research is needed to understand the impact of modifications for loans beyond the 1 year that we can observe, especially because under the HAMP program, interest rate reductions are only required to be in place for 5 years.

30. Effective January 2014, the CFPB’s MSR, 12 Code of Federal Regulations 1024 and 1026, is a collection of nine separate rules.