Saturday, September 13, 2008

Loan Modifications: Anecdotes and Data

by Tanta on 9/13/2008 10:56:00 AM

From an email I got yesterday from a mortgage broker (with 28 years in the business), on the subject of loan modifications:

It really seems out of control with no one even able to get a straight answer. All the reports I see say that borrowers in trouble but who are willing to try to make a go of it are not getting help. Many are told, "Unless you are 90 days in arrears we can't help you, and BTW, we won't commit now to what we might or might not do if you do decide to miss a few payments."
I certainly see a lot of "reports" of this as well, which is why I think it's worthwhile to tackle the assumptions here (again).

First, what are "borrowers in trouble"? If they are current and have never been late on a payment, what kind of "trouble" are they in? Servicers do "loss mitigation" modifications when default on the loan is imminent or reasonably foreseeable. That is because there must be the probability of "loss" before anyone tries to "mitigate" it. How is default imminent or reasonably foreseeable if the borrower is paying as agreed?

Possibly we mean that the borrower is only one or two payments behind. The problem here is that most servicers do not consider default imminent after one missed payment. 30-day delinquencies are very likely to cure themselves: the borrower makes up the missed payment in the following month quite frequently, pays the late charge, and goes on. Obviously some do "roll" to 60 days down or worse; you can't have a 60-day delinquent loan that wasn't 30 days delinquent in the prior month. But even 60-day delinquencies have a fairly high "cure" rate.

It is very difficult to make the call at 60 days. As a matter of "traditional" loss mit practices (which may of course be undergoing some changes very recently), what the servicer does depends on what the borrower has told the collections department about the reason for delinquency. (Everyone has already had several collections calls by the 60th day of delinquency. If you refused to take those calls or refused to give a reason for your delinquency, you should not be surprised to find your "willingness to make a go of it" under some question.) If it appears to be a temporary matter (business was slow last month because of the hurricane, there were non-recurring unanticipated expenses to deal with), the servicer is likely to offer a forbearance or repayment plan rather than a modification. If it appears to be a longer-term problem (ARM reset, job loss, divorce), a servicer might consider a modification at this point. The two choices are not exactly mutually exclusive; a time-tested servicing strategy is to put the borrower on a repayment plan to verify his or her willingness to continue to make payments at the reduced level before going ahead with a permanent modification.

Those of you who are still convinced that it is somehow "unfair" that a loan really has to hit the 90-day down category before the servicer concludes that default is imminent should remind yourselves that it is extremely rare for a servicer to begin foreclosure proceedings until a loan is 90 days or more delinquent. Why is that? Because most loans with a 30- or 60-day delinquency cure themselves. If you want servicers to conclude that default is imminent the moment a borrower becomes one payment behind, you are really expecting servicers to start foreclosure proceedings that early in the game as a matter of course.

I don't happen to believe that borrowers are being told across the board that "unless you are 90 days in arrears we can't help you." I do believe that borrowers are regularly being told that unless they're 90 days in arrears they are not going to be considered for a permanent modification. The problem here is that we don't know what these borrowers asked for when they got that answer, what reason they gave for their "trouble," or what other options the servicer might have offered that they turned down. I have my suspicions about that, given that my broker correspondent indicates that the servicers are also saying "we won't commit now to what we might or might not do if you do decide to miss a few payments."

No sane negotiator would ever commit to modifying a loan as long as it was 90 days delinquent. Severe delinquency is a necessary, but not a sufficient condition. For one thing, if you are "deciding" to miss a few payments, there's some question about whether you are suffering a bona-fide hardship or not. It honestly really shocks me that people would expect a servicer to commit to giving you a deal prior to delinquency: that is simply inviting people to withhold payments in order to get their interest rates reduced. Why would anyone expect a servicer to do that?

Perhaps we all have widely-varying points of view about the wisdom of negotiating with hostage-takers, but I suspect that most of us believe that the police would be in error by declaring in advance that anyone who takes a hostage will have his demands met. It might, you know, encourage hostage-taking. What I suspect we're hearing from here are those unhappy campers who are basically telling their servicers they want a rate- or principal-reduction modification or else they'll "walk." The servicer tells them that nothing is on the table unless they're 90 days down, because that, at minimum, tests the seriousness of this threat to walk away. They come back with "So if I miss three payments, you'll modify my loan?" and the servicer says "Maybe. Maybe not." Of course that's what the servicers say. The rules of the game of chicken require this.

I think it is disingenuous to characterize this as borrowers "not getting any help." I think it is disingenuous to characterize this as not getting a "straight answer." It is a perfectly straight answer; it's probably too straight, if anything. While servicers who tell borrowers that modifications aren't on the table until the 90th day of delinquency are at least being honest, it appears that you can't really be that honest with a certain class of borrowers. That's because they interpret this as meaning that they "qualify" for a mod once they're 90 days down, as if that were the only issue.

The OCC's second Mortgage Metrics Report came out yesterday, and it shows that not only are repayment plans and especially modifications increasing, they are increasing at a faster rate than new foreclosure starts. I think one of the most useful metrics in the report is the number of new loss mitigation actions as a percentage of all seriously delinquent loans:



This is not a "cumulative" metric in any sense: it measures new actions taken in the month, not a cumulative total of actions taken, and the denominator is not all loans that have ever been seriously delinquent, but only those loans that are seriously delinquent in the report month. But given that most loans have to be seriously delinquent before loss mitigation is an option, this gives you a sense of how many eligible loans are given repayment plans or modifications (those are the only loss mit activities measured by this report).

What it shows is that "prime" serious delinquencies are consistently the least likely to receive a repayment plan or modification. (It doesn't matter that prime loans are least likely to be seriously delinquent; we are only counting serious delinquencies here.) There are a number of possible reasons for this:

1. Prime borrowers may be less likely to experience serious delinquency for "temporary" or curable reasons. That is, prime borrowers may have more stable employment, more savings, and better financial management habits, such that they are less likely to become seriously delinquent because of work slow-downs or unanticipated expenses, which are the sorts of temporary hardships most likely to be cured with a repayment plan or modification that capitalizes arrearages. That would imply that when prime borrowers do become seriously delinquent, the underlying problem is more severe--less "curable"--than with other loans. Another possibility is that prime loans are more likely to become seriously delinquent for reasons such as need to relocate; in those cases, a short sale or deed-in-lieu may be a better option for both borrower and servicer than a repayment plan or modification.

2. Most of the prime loans in the OCC/OTS database (serviced by large depositories) are conforming loans in GSE securities. That suggests that they are less likely to be concentrated in the former "bubble markets" than the Alt-A and subprime loans. Insofar as a substantial fraction of these loans have much lower loss severities in foreclosure than the Alt-A and subprime loans, they are less likely to pass the "least loss to the investor" test.

3. Prime borrowers are possibly trying to play chicken with their servicers and are not winning. I have no idea whether such a thing is true or not, but these anecdotes we keep hearing like the one above have to be accounted for, if true. If anyone is "deciding" (rather than being forced by simple inability to pay) to become delinquent, you'd think it would most likely be the prime borrowers.

The question remains whether an overall monthly new loss mit rate of 8.83% of all seriously delinquent loans is much too low or much too high or about right. We really aren't given enough data in this report to judge that. (Just for one thing, we don't know how many of the seriously delinquent loans are speculative purchases or abandoned properties; those would be highly unlikely candidates for a repayment plan or modification.) Measuring new loss mit actions against all seriously delinquent loans, rather than against newly seriously delinquent loans, also leads to a question about the denominator. Since this category will include loans in the foreclosure process that have been and will be there for many months, and that may well be past the point where a workout is even appropriate, the metric will be skewed because of the sheer number of months loans in foreclosure spend in the process. But even if we conclude that this number is "too low," it certainly isn't zero.

Eventually, the OCC is really going to have to refine these metrics if we are going to have reliable data with which to answer the question of how much is enough, loss mit-wise. Certainly, proponents of this silly 90-day foreclosure moratorium for the GSEs need to explain why, exactly, they think that the number of successful workouts per delinquent loans will increase simply by eliminating foreclosure starts or delaying foreclosure completions. My fear is that we're basing public policy recommendations here on the kind of anecdote I started with, not on good data; if Congress and the regulators have better data than what's in this report, I don't know why they aren't sharing it with us. And this report still doesn't tell us how many seriously delinquent loans can practically or effectively be worked out short of foreclosure.