Behaviorism and Mortgage Workouts

By Alan White

Some useful applications of behavioral economics have emerged from efforts to deleverage American homeowners through mortgage renegotiations.  Servicers have complained that they have great difficulty persuading borrowers to respond when they offer help and solicit information, like current income verifications, in order to rewrite loan terms.  The FDIC/Indymac program took the simple additional step of calculating a new (lower) monthly payment that could result from a successful application for modification, and included that vital bit of information in its mailing to homeowners.  It's the same behavioral device as a pre-approved credit card.  Not surprisingly, borrower response rates improved.  According to a story in Friday's L.A. Times, Indymac got a 73% response rate from borrowers who had previously called, and to whom it sent a proposed modification with a prepaid, return FedEx envelope.  73% is an unheard-of response rate in this context.

The state regulators' September data release reports that there remains a large gap between homeowner applications for workouts and completed workout deals, on the order of two or three to one.  Why would a homeowners start the process and not see it through?  Or on the other hand, why would servicers spend time and effort beginning a workout application but not see it through?  From the standpoint of rational choice theory, both parties would have much to gain and little to lose, assuming that the servicer has done some initial screening (foreclosure value prospects not good, homeowner still in the home, homeowner believes they have income to service modified loan.)  Remember, we are talking about applications that remain undecided, not applications that are rejected. 

The large drop-off in foreclosure workout applications may in part be due to inadequate staffing and resources on the servicer side.  But it could also be due to channel factors, i.e. exactly what affirmative steps the borrower must take, and how and when they receive some positive encouragement about the process.

There are, of course, some possible rational choice explanations for low borrower response rates:  perhaps many borrowers who don't call or follow through realize their income is just too low.  They may have found a better, cheaper housing alternative, or have been involved in a questionable or fraudulent mortgage application, as is the case with investors who misrepresented themselves as owner-occupants.  Nevertheless, the large difference in response rates produced by very simple servicer strategies reinforces my view that more foreclosures can be prevented.

Comments

[Mike McCarthy]

These macro discussions of survey responses, like this is social science research, seem to miss the point. So far, little connection and comparison is being made, systematically, between the market value of an MBS, or for a lender-held mortgage the final dispositional value of the property after foreclosure net of all lender costs, and the specific circumstances of the borrower.

The servicer system works for the industry in good times because financial errors made in disallowing justified work-outs do not so substantially affect the lender's bottom line as to invite analysis of the business model.

When bad times occur, servicers quickly become disfunctional and the system spins out of control. Way past the point when the initial approach of "forebearance" becomes counter-productive, by setting borrowers up for later default or by constructively encouraging walk-aways ("These people are insane, I have nothing to lose...").

Further the servicing contracts on bundled mortgages seem to also be modeled on modest levels of defaults in good times and are too inflexible to permit the kind of options that can prevent foreclosures in bad times.

Also many servicers are not set up to do efficient loss mitigation that fully servies the interests of lenders. They make you talk to a different person each time (after excessive waits on hold), bar direct discussions with loss-mit staff, fail to respond to mail, lose faxes, refuse to allow use of email, and seem to have no sense of urgency (e.g., the insistence of a minimum period of default before considering loss-mit, even when the borrower has hit an iceberg).

As an example, I represented a subprime borrower, who had serious self-employed business setbacks, before one of the largest servicers in the nation re a mortgage in an MBS (trustee USBank). Borrower was insolvent, >90 days in default, with a big reset to occur in a yr. Without a substantial mod, this would soon be a foreclosure in a very bad Fla housing market. Relying on deep-pocket borrower relatives (810 FICO), I offered to save 87% of the asset (mainly principal red. to $150K), pay $10K cash to deal with cash-flow & closing issues, freeze the current 7.5% 30-yr interest rate & convert it to 7.0 15-yr to accelerate principal paydown & reduce future lender risk, and have the mortgage converted from sub-prime to prime by having the deep-pocket relatives co-sign on the re-worked mortgage. This was my INITIAL offer, and the borrower was willing to go up to a 90% asset preservation as a counter. The servicer offered only a standard forebearance settlement with higher payments for 4 mos. to the point of "reinstatement". Two mos. later, with no negotiation, foreclosure action was filed. This house, appraised for $215K 2 yrs earlier, will now sell as a REO for about $75-80K (after expenses USBank will net far less).

After intensively following the housing crisis since 2004, and the industry's response in 2007-08, I am convinced that at least two things are needed to prevent a serious price crash & continued neighboorhood destruction - -

1. The bankruptcy code needs to cover primary residences the same as other properties, so judges can, as a bottom-up counter-weight to industry resistance to change, make justified loan mods to accomodate current market values of properties and borrowers circumstances. Because of settled rules, most simple bankrupcies can be resolved with little court involvement. And the mere existance of this counter-weight can assist borrowers' attys in securing justified loan mods without haveing to resort to formal bankruptcy proceedings.

2. An intermediary agency need to be empowered to essentially do "class-action" loan mods for both lender-held and MBS bundled mortgages. The connection need to be made systematically between the current real values of primary homes at risk (considering financial factors, not moral ones) and the circumstances of borrowers. Current "social-science" type business models like H4H and the FDIC IndyMac experiment (which excludes MBS mortgages entirely), cast too small a net on the first cut and weed out too many potentially eligible borrowers in successive cuts. An intermediary is needed to promptly force proper valuations, esp. MBSs, as a baseline for affirmative dispositions with borrowers headed for foreclosure.

[Mortgage]

This article was very interesting. Thanks for your insights!

[Robert]

[The response rate is higher -- 73% -- for borrowers who had already talked with IndyMac and provided their current income.]

A 73% response rate would be unheard of if the mailings had been sent cold.

The fact is, the "Bair approach" has been tested before and it typically yields a response rate in the 30's when used by mortgage servicers.

I find it both amusing and discouraging that the Indymac program is being hailed as either "innovative" or "wildly successful."

[There are, of course, some possible rational choice explanations for low borrower response rates]

There may be, however this overlooks the fundamental principle: MANY borrowers in default do not act rationally. Any "program" predicated on rational behavior by the borrowers is based on false assumptions.

OTOH, you have borrowers who are acting VERY rationally: why get a "workout" on a loan when they're underwater by 10-20% or more?