Servicer and Spatial Heterogeneity of Loss Mitigation Practices in Soft Housing Markets

Abstract

Although loan modifications are being widely used as a way to stabilize housing markets by preventing avoidable foreclosures, not much is known about the ways in which specific servicer-related factors affect the likelihood of modifications. Using a large sample of nonprime loans, this study examines recent loan modification practices adopted by different servicers in two types of soft markets, in neighborhoods differently affected by the foreclosure crisis, and among borrowers in different racial and ethnic groups. The results demonstrate striking variations in the incidence of loan modifications by servicers and significant differences between the servicers more likely to modify troubled loans and those who are less likely to do so. Loan modifications are less frequent where they are needed the most: among savable borrowers in the neighborhoods hardest hit by the crisis. This considerable variation in modification practices across servicers and neighborhoods likely reflects both structural obstacles to modifications and the absence of a uniform approach to loss mitigation.

Keywords:

• mortgage
• housing
• housing crisis
• loss mitigation

Acknowledgments

The author thanks Roberto G. Quercia, Avis Vidal, Carolina Reid, Padmasini S. Raman and anonymous referees, as well participants at the 2011 APPAM conference and the 2011 ACSP conference for their helpful comments and support. The author also wants to thank Sangida Mazumder for her research support. This research was supported by the University Research Grant Award at Wayne State University.

Notes

^1. On January 17, 2013, the Consumer Financial Protection Bureau (2013) issued two mortgage servicing rules. These new rules will be in effect on January 10, 2014. They implement provisions of the Dodd–Frank Act regarding mortgage loan servicing and intend to ensure that troubled borrowers get a fair treatment by servicers. In addition, the Treasury has started to make servicers accountable by publishing quarterly servicer assessments and by withholding incentives from a few servicers with poor performance since June 2011 (U.S. Department of Treasury, 2012).

^2. Options other than foreclosure for borrowers in delinquency include (1) curing the delinquency, (2) modifying the mortgage, (3) refinancing or paying off the mortgage, (4) conducting a short sale, (5) getting lender's short-term forbearance, and (6) filing bankruptcy to delay foreclosure temporarily. In addition to these six options, additional options for borrowers already in the foreclosure process include (1) curing default by paying in full the late payments and fees, (2) obtaining a deed in lieu of foreclosure, and (3) losing the property to the lender/servicer (real estate owned property) or to a third party in a foreclosure auction (Been, Weselcouch, Voicu, & Murff, 2011).

^3. In an initiative to help troubled borrowers served by the failed thrift Indymac the Federal Deposit Insurance Corporation adopted an approach in 2008 that extends assistance to troubled borrowers by establishing affordable and sustainable payments (a debt-to-income ratio of 31%; Federal Deposit Insurance Corporation, 2008).

^4. While the HAMP generally refers to the first-lien mortgage modification program, it also includes a set of subprograms designed to modify second and higher liens and to assist vulnerable homeowners in the hardest-hit neighborhoods like the Home Affordable Unemployment Program, the Second Lien Modification Program (2MP), and other subprograms. In addition, the HAMP's Tier 2 became effective on June 1, 2012, to increase participation and the number of homeowners who are eligible in the hopes of preventing foreclosures.

^5. The HAMP is available for loans owned by Fannie Mae or Freddie Mac, or certain non-GSE loans served by participating servicers but most privately securitized mortgage modifications (21,319 of 38,313 in the third quarter of 2011) were non-HAMP modifications (U.S. Office of Currency Control, 2011). The Federal Housing Administration, secondary market players like Fannie Mae and Freddie Mac, states, and individual lenders have developed their own independent loan modification programs.

^6. The expiration date for the HAMP has been extended to December 31, 2013, and may be further extended afterward.

^7. See http://www.americansecuritization.com/uploadedFiles/ASFStreamlinedFrameworkQA121707.pdf for details.

^8. According to Amherst Securities, which is cited in a testimony by Calhoun (2011), with a typical servicing fee of 25 basis points per year ($625/year on a $250,000 loan), servicers are overpaid for traditional servicing (which costs servicers only about $48 per year) and underpaid for loss mitigation on nonperforming loans (which costs about $900/year).

^9. Of course, the HAMP requires servicers to consider whether modifications would be appropriate for all loans owned by GSEs or certain non-GSE loans served by participating servicers.

^10. The 2MP focuses on eliminating or modifying junior mortgages. When a borrower's first-lien mortgage is modified under the HAMP, servicers participating in the 2MP are required to offer modification of the borrower's junior mortgage. However, as of April 2012, only 80,000 2MP transactions occurred despite the large percentage of junior mortgages present in the mortgage market during the subprime boom (U.S. Department of Treasury, 2012).

^11. The first servicer assessment conducted by the Treasury in June 2011 showed that four out of the 10 servicers needed substantial improvement, and the remaining six needed moderate improvement (U.S. Department of Treasury, 2011a). Four out of the top servicers had an unacceptably high number of wrongly denied loan modification cases where the Treasury was unable to determine why the servicers did not offer a permanent modification based on the documentation provided by the servicer. Of course, since the Treasury started to publish the servicer assessment report in June 2011, the performance of major servicers has been improving, and the Treasury has started to withhold incentives from a few servicers (U.S. Department of Treasury, 2012).

^12. The most important criterion is a standard net present value (NPV) test that compares the NPV to the investor of a modification to the status quo of no modification. If the NPV test shows that a modification will result in net benefits to the investor, the servicer is required to give the application further consideration. Other eligibility criteria include proof of financial hardship, documentation of sufficient income for the modified payment, and some restriction on the origination date and the loan amount (see http://www.makinghomeaffordable.gov for details).

^13. For example, Fannie Mae adopted a policy to restrict the maximum LTV ratio and CLTV ratio for properties located within a declining market to 5 percentage points less than the maximum normally permitted for the selected mortgage product. Incorporated in its Desktop Underwriting system, Fannie Mae defines a declining market as one in which home prices are currently declining as determined by their tracking of home prices in a certain geographic location. See https://www.efanniemae.com/sf/guides/ssg/annltrs/pdf/2007/0722.pdf for details.

^14. These programs include the Home Price Decline Protection and the Home Affordable Unemployment Program under the HAMP, the Hardest Hit Fund (HHF), the Emergent Homeowners Loan Program, and the NSP. Similar to a block grant, the HHF provides funds to state housing finance agencies, and those agencies then design and implement detailed foreclosure prevention programs in their states, subject to Treasury Department approval. The fund was allocated to states with sharper housing price declines and/or higher unemployment rates (Immergluck, 2011). The Emergent Homeowners Loan Program, launched in the summer of 2011, was designed to help unemployed homeowners in states where HHF funds were not available to pay their mortgages. In addition to these efforts to prevent foreclosures, another set of federal programs, most notably the NSP, has focused on addressing the negative effects that foreclosed and vacant properties have in foreclosure-concentrated neighborhoods.

^15. Using the estimated rate of loans in foreclosure or delinquent, the HUD identified the 20% of neighborhoods likely to be most distressed. This equates to an estimated serious delinquency rate (90+day delinquent or in foreclosure) of greater than 17.8%. See http://www.huduser.org/portal/datasets/NSP3%20Methodology.pdf for details.

^16. Although loans in the early stages of foreclosure could still be eligible for major modification programs, like the HAMP, this study chose to focus on recent delinquencies only for several considerations: (1) loans already in foreclosure prior to 2010 were less likely to be the target population of early intervention efforts; (2) an inclusion of loans in foreclosure would inevitably introduce the complicated issues related to the stock of foreclosure and dual track modifications, which is out of the scope of this study; and (3) a replication of the regression models using the sample further including loans in foreclosure gets quite consistent results, except the expected differences due to the inclusion of higher risk loans (results available upon request).

^17. For example, if the odds ratio for the NSP17* servicer 1 variable is significant and equal to 2, it can be interpreted as for troubled loans in a high-risk neighborhood, the odds of modification being 100% higher for those served by servicer 1 than those served by small servicers.

^18. The data do not allow us to identify loans with higher liens directly. This analysis treats that a loan has piggyback loans if its CLTV is at least 3% higher than the original loan-to-value ratio. While this is not a precise method, it should be able to capture most loans with higher liens.