Recently I attended a two-day symposium on mortgages and the future of housing finance on on October 25 and 26, 2010, hosted by the Federal Deposit Insurance Corporation (FDIC) and the Federal Reserve System. (Here’s the link to the agenda: www.fdic.gov/bank/analytical/cfr/oct2010-conf.html.)
There were three sessions that I personally found particularly interesting that I’ll be summarize over the next few days; (1) loss mitigation and loan modification practices and performance, (2) adverse selection in mortgage securitization, and (3) mortgage default decisions including strategic default. In this first post, I’ll focus on loss mitigation.
The sessions on loss mitigation and loan modification highlighted several papers that tried to determine why we are not seeing more loan modifications (as opposed to foreclosures). Gene Amromin gave us the lay of the land based on a paper he wrote with Sumit Agarwal, Itzhak Ben-David, Souphala Chomsisengphet and Douglas Evanoff that can be downloaded here.
Basically, they document that foreclosure is the dominant form of loss mitigation, and that securitized loans are less likely to receive modifications. Borrowers with the highest ex-ante likelihood of getting back on track (“self-curing”) are less likely to be offered modifications (e.g., high credit scores and low loan-to-values). They also find that affordability, as opposed to negative equity, is the primary driver of redefaults.
Paul Willen discussed the challenges in determining which borrowers are “deserving” of concessional modifications. Servicers don’t want to offer modifications to delinquent borrowers who will self cure, or who will redefault. This assessment is very difficult, making foreclosure the path of least resistance. On the other hand, Sanjiv Das presented a theoretical loan modification model that accounted for these unknowns to show that smart modifications improve the lender’s loan value (the paper can be downloaded here). The optimal modification involves a principal writedown coupled with shared-appreciation going forward.
However, as we all know, principal writedowns are seldom being employed, and the rest of the session’s discussion focused on the why-nots. A popular scapegoat in this regard is securitization. However, Andra Ghent presented a paper that showed that concessionary residential mortgage modifications were rare even during the Great Depression (download here). Furthermore, there were no instances of principal forgiveness in her dataset (loan level data from the New York City metropolitan area). Interestingly, she posits that the Home Owners’ Loan Corporation (HOLC) may have reduced renegotiation incentives:
Our results also suggest a possible downside to government programs such as the HOLC. To the extent that lenders benefi t from government programs that remove non-performing mortgages from their balance sheets, such programs may adversely a¤ect lenders incentives to preserve the values of their mortgages through private renegotiation.
Adam Levitin brought us back to the present with a discussion of the role of servicers in loss mitigation. (He didn’t present a paper, but has written extensively around this topic, as you will see if you visit his website.) He said the the paucity of loan modifications isn’t the fault of securitization per se, although the unbundling of servicing from lending seems to me to be a symptom of securitization.
Anyways, Adam ran through the factors that distort servicer incentives. For example:
- Servicing and trust fees are fixed creating strong incentives just to minimize costs, and not maximize value. For example, there is no incentive to maintain costly loss mitigation systems through the cycle.
- Servicing fees are based on outstanding principal, which reduces any incentives to offer principal reductions.
- The big servicers are affiliated with banks that own 2nd liens on the properties that underlie the 1st lien mortgages they service. These 2nds would have to be written off in a typical principal writedown.
- Servicers often hold “residual” interests in the transactions that they service, so they’re incentivized just to keep the cash flows coming.
Adam presented a couple of potential policy prescriptions, including intervening in trust operations (e.g., impose an ombudsman funded by servicer) and allowing residential mortgage cramdowns in bankruptcy. However, during one of the Q&A sessions, he admitted that that ultimately it will be a political decision as to who will bear the cost of principal writedowns. Also, there is the problem of what to do about 2nd lien writeoffs that push some big banks to the cliff edge?
That’s all for the loss mitigation sessions, and in the next post I’ll cover the securitization sessions.
——————————————————–
* John Kiff is a Senior Financial Sector Expert at the IMF. These are his personal views, and should not be attributed to the IMF, its Executive Board, or its management.
