Monday, March 05, 2007

A Tantamentary on "Messier Mortgages"

by Bill McBride on 3/05/2007 10:56:00 AM

The following is a Tantamentary (by Tanta) on a New York Times column (pay, excerpted to put Tanta's comments into context).

NY Times columnist Gretchen Morgenson writes: Mortgages May Be Messier Than You Think

WHAT investors don’t know about why the home mortgage securities market is in distress could fill volumes. ... only after fiery markets burn out do we see the risks that buyers ignore and sellers play down.

... Unlike recent corporate disasters that have occurred at hyperspeed ... the mortgage securities boom seems to be unwinding in slow motion.
Tanta: Explain that to FMT and NEW. It looks like the tempo is picking up. the end of last week, many there were celebrating the fact that the indexes on mortgage securities ... had stabilized. Big brokerage firms have also tried to persuade investors that mortgage woes would be limited to subprime loans — those given to people with weak credit histories.

But last Thursday, the annual report from Countrywide Financial, a major lender, told a different story. While it confirmed fears about subprime loans — 19 percent of those in its portfolio were more than 30 days delinquent at the end of last year, up from 15 percent in 2005 — Countrywide also indicated that the percentage of prime borrowers encountering difficulties is rising. Delinquencies in the company’s prime home equity loan portfolio totaled 2.93 percent, almost double last year’s 1.57 percent.
Tanta: You had to have believed in the 1.57 as a reasonable number to be really shocked by this. No, I’m not in that crowd who thinks CFC has been lying in the reports. I’m in that crowd who has believed for years that the DLQ numbers were “masked” by easy lending conditions and a hot RE market, not by loss-mit games. This does beg for the “spread to prime” meme to be placed into some context, namely, that during the boom prime even performed better than prime. As it were.
And consider the disclosure last Thursday of American Home Mortgage Investment, a home mortgage originator and investor that specializes in loans to those with middle-tier, not weak, credit histories. As of year-end, the company said, 8.13 percent of its loans held for sale (not investment) were non-accruing. During the same period in 2005, that figure was just 0.43 percent.
Tanta: Having non-accrual loans in the held-for-sale pipeline is like having delinquent loans in the warehouse. They aren’t supposed to be there long enough for that to happen. What is getting me at this point is not, actually, AHM’s nonaccrual rate (although 8.13% is some major ick). It’s that anyone is allowed to continue to call them “held for sale.” Think of it like a grocery store. If you have bananas that are starting to turn black, do you leave them out in the produce section and pretend that someone might buy them, or do you trot them out to the dumpster and take a charge-off? You can ask someone like me all day long how one handles nonaccrual loans in the sale pipeline. And I’ll tell you all day long that I have no idea, because it’s never happened to me before. If I ever had loans that were starting to look that bad in the sale pipeline, I either sold them immediately to a scratch & dent dealer or, more likely, put them in the portfolio and took the write-down, hoping to cure them, season them, and then unload them later at a better price, while following the rules about where they go on the books while I am taking this risk. I never let anything get old enough to go nonaccrual in the HFS. The HFS is funded by the warehouse (a literal borrowed warehouse or an internal corporate warehouse, the latter being the way a bank that funds its own loans does it. In either case, it’s “temporary financing.”) The end of the story is that we’re in some brave new world of accounting, as far as I can tell, not just some brave new world of loan failures.
“The problems are far broader than subprime,” said Josh Rosner, a managing director at Graham-Fisher ... Mr. Rosner says he believes that ... investors will soon recognize that credit quality problems have also begun to seep into “the upper tranches” of the loan market.

Credit risk ... shows up over months and years, while market risk plays out in seconds and minutes. ... it is easy for some to say that a big decline in the mortgage securities index is an overreaction.

Perhaps it is. But the fact that loan losses have not yet shown up in mortgage holdings should calm no one. The way Wall Street packages home loans and sells them to investors adds to their complexity and makes them far tougher to value than other securities.
Tanta: True. The way Wall Street does this is pretty hard to figure. But to what extent is the Street playing in the prime mortgage loan market? I begin to sense, once again, that tranche/pool problem in play here. I mean, a Freddie Gold 30-year MBS is AAA. Some senior tranche of some whacko Alt-A or subprime security is AAA. These are both “prime” ratings of the security or small slice thereof. They don’t have the same loans backing them. This confusion is going to get worse before it gets better, I fear.
Moreover, a lag in reported defaults is almost certainly attributable to the increasingly aggressive practice of loss mitigation among lenders, in which they try to keep stretched borrowers from defaulting. But investors get little to no information about how lenders work with troubled borrowers and whether those efforts actually cure the problem — or simply postpone the inevitable. As a result, investors do not know whether the default figures they are seeing reflect reality.
Tanta: Translation: “I have no actual information about what the lenders are doing. Therefore, I am “almost certain” that they are doing the wrong thing. Of course, would I pick up the phone and call one or two mortgage servicers to get a comment on record? Nah! It’s much more fun to speculate and call it certainty.”
In 2002, the Office of Inspector General in the Department of Housing and Urban Development reviewed the department’s loss-mitigation program and found some disturbing practices. Financial institutions that were surveyed from May 1999 to April 2001 in many cases chose to delay foreclosures even when it was obvious that loan workouts would probably not succeed ...

Mortgage servicers “are approving borrowers for loss mitigation when, based on the servicers’ expertise and past experience with delinquent borrowers, the workout is unlikely to succeed,” the report said. “These actions are delaying the foreclosure process, increasing the cost of foreclosure, and subsidizing borrowers who do not pay their mortgage for extended periods of time.”
Loss-mitigation practices grew out of the significant losses incurred by big lenders in the real estate collapse of the late 1980s. Because lenders were experiencing losses of 30 percent to 60 percent of the outstanding home loans in foreclosures, they started identifying ways to keep borrowers in their homes.
Tanta: Nice selective paraphrasing from that IG report. How about the full context?
“In our opinion, servicers should be allowed to determine whether the borrower had a hardship when exercising good business judgment. Otherwise the Department is only encouraging irresponsible behavior. For example, we reviewed a case where the borrower’s hardship letter explained that their increase in expenses was due to gambling losses. Since the borrower had an increase in expenses, the servicer felt obligated to approve a partial claim. The borrower subsequently filed for bankruptcy protection. In another case, a servicer granted loss mitigation to a borrower who was current on their Mercedes payments but was not making payments on the less expensive home mortgage.

A first vice president of another servicer told us that HUD criticized it for not doing enough for borrowers whose expenses increased and income decreased. The HUD National Servicing Center told the servicer to make it work. HUD instructed the servicer that if a borrower meets all of the guidelines for a specific loss mitigation tool, the servicer could not deny the loss mitigation action, even though past experience shows that a borrower under the same circumstances will default again, resulting in foreclosure. The first vice president said that the servicer has to do what HUD says, because the servicer has not been given any authority or charter to make decisions. The serivcer would like to see the initiation of more discretion by HUD to the loan servicer on whether a borrower should be given additional chances to correct their delinquency. If the loss mitigation tool or tools do not work based on circumstances, then the servicer should end the process.

Finally, a senior vice president of a third large servicer said there are borrowers who learn the system and take advantage of it, as well as borrowers who do not want to deal with the fact that they are defaulting on their mortgage and may lose their home. The servicer believes they should be able to look at the behavior patterns of the borrower and decide on no more workouts when the behavior indicates that the borrower is going to default again. The servicer believes, in these instances, that providing the loss mitigation tool is just delaying the inevitable.

These servicers stated to us that they interpreted instructions from the National Servicing Center that any borrower who has had either an increase in expenses or decrease in income must be approved regardless of the borrower’s hardship or circumstances. Such action only delays the foreclosure process.

Using this criterion, virtually every interested delinquent borrower qualifies for loss mitigation. The servicers are reluctant to use good business judgment because HUD may require them to indemnify the loan if loss mitigation is denied. Consequently, servicers are delaying the foreclosure process, which increases the cost of foreclosure, and subsidizes borrowers who do not pay their mortgage for extended periods of time.

A HUD official with the National Servicing Center told us that trainers tell servicers that they must consider all borrowers for loss mitigation but are to use their best judgment when determining whether a loss mitigation tool is beneficial and warranted. If HUD expects the program to succeed, HUD needs to clarify instructions to servicers to eliminate the misunderstanding or perception that is preventing servicers from using good business judgment when working with delinquent borrowers.

A HUD official advised us that one of the biggest problems with the assignment program, which the loss mitigation program replaced, was the different ways the field offices applied it. For example, some field offices were very diligent in permitting borrowers into the program while other offices had very liberal policies and let almost anyone participate. Indications are that differing interpretations of HUD requirements by servicers, as previously discussed, may be creating a similar problem with the loss mitigation program.

We also reviewed 105 cases where the borrower continued to miss mortgage payments after receiving home retention loss mitigation assistance for the first time. The cases were reviewed to determine whether servicers are performing due diligence when approving borrowers for loss mitigation and are taking appropriate actions to apply disposition options or initiate foreclosure when it is clear that the borrower does not have the ability to repay their mortgage. We found that servicers were performing due diligence when approving the borrowers for loss mitigation. The servicers based their decisions on the information provided by the borrower. If the borrower does not have a history of problems, the servicers are inclined to take the borrower’s assertions at face value.

HUD commissioned a study by Abt Associates on the loss mitigation program. The Abt Associates report, issued in November 2000, stated that in order to evaluate program effectiveness, it is important to have information on current borrower circumstances, including their credit score, income, and expenses, as well as the property’s condition and the current estimated loan-to-value ratio. However, we found during our review that the biggest factor in determining whether loss mitigation will be effective relates to social conditions, such as desire to maintain the home, family problems, drug or alcohol abuse, and gambling.

Several servicers have done extensive analysis of their data to identify statistical factors that may dictate whether a workout will be successful. One servicer told us that they found no correlations in its statistical information to identify workouts that may or may not succeed. Another servicer had some similar results. For example, the servicer did not identify any significant trends when analyzing success rates by the credit quality, age of the loan, loan-to-value ratio, interest rates, or surplus/deficit amounts. Although, the servicer found that when borrowers surplus cash was higher than $1,000 the success rate started to drop. From this unusual trend, the servicer learned that they were not getting an accurate picture of the borrower’s actual expenses and instituted procedures to improve the gathering of expense related information.”
Long quote there, but I’m fascinated by the way an Inspector General report that found perverse incentives by well-meant HUD regulations and field office interpretations thereof is now being used to “prove” that servicers do not exercise good business judgment, or that servicers use loss-mit to “hide” delinquencies. There was, of course, never any time at which HUD did not consider these loans as under-performing. Furthermore, one could use this IG report as a useful way to consider the wisdom of stated income/stated asset/no doc loans rather than an argument against loss mit. But that would interfere with today’s story line.
While it is a splendid idea to do whatever is necessary to try to cure a sick borrower, it is worth remembering that the rollover of nonperforming loans was central to what made the savings and loan mess of the early 1990s so disastrous. And it is well worth asking: are loss-mitigation practices predatory since they give lenders an opportunity to squeeze the last ounce of blood out of a terminally ill patient?
Tanta: It might be worth talking to the 99% of students of the S&L crisis who thought that there were a whole lot of other candidates for the “central” problem. Holy Selective Memory, Batman. (Hint: can you spell “commercial loans”?) And it is well worth asking: are NYT business columnists so bird-witted that they really think that tossing people out onto the street in a foreclosure could not, on the “ounce of blood” theory, be considered just as predatory? Can anyone imagine the next story in the NYT, showing the heart-breaking pictures of the laid-off uninsured family with a sick breadwinner standing on the curb in the snow while the sheriff evicts them from the foreclosed property? What, we’re now making financial services policy based on who can print the saddest story, the pro-loss-mit or the pro-FC crowd? We just looked at an IG report that indicates that the big problem was servicers doing loss mit for borrowers WHO ASKED FOR IT, but should probably not have been given it. Um, Gretchen, borrowers who scam the HUD system by getting forbearance while driving a Mercedes are probably not the terminally ill prey of the servicers. Just sayin’. Let’s face the real issue: why were these people put in these loans in the first place by lenders? What were their financial incentives for doing so? Who provided those financial incentives? Can you spell “investors”?
Aggressive approaches to loss mitigation, Mr. Rosner said, mean that less than half of mortgages that in prior years would have been foreclosed no longer are. Unfortunately, how many of these workouts actually succeed is a question for which investors and even regulators have no answer.
Tanta: So since we have no answer, let’s assume that it’s a bad thing. Absence of evidence means never having to say you’re full of shit.
Academic studies suggest that within the first two years of a workout, re-defaults can approach 25 percent. But there are no publicly available data to analyze the success rates of loss mitigation. And this is something investors sorely need.
Tanta: Betcha there’s some privately-available data. How about placing a call to, oh, TIAA, and asking if someone there is responsible for analyzing the performance metrics of the mortgage-backed securities and/or whole loans they’re buying? Did you do that, and find that TIAA doesn’t have any information on its own investments? Can you quote them on that? Or do you mean that you and I don’t know what TIAA knows? I’m getting confused about whom I’m supposed to be disgusted with here. I mean, I’d absolutely love to see publically-available data on the performance of these mortgages. I wonder why it is that publically-traded companies and rating agencies and outfits like Loan Performance don’t give it away to us. Sheesh—and people think I’m some kind of socialist.
“There is nothing particularly wrong with loss mitigation,” Mr. Rosner said. “What is wrong is a lack of transparency in whether it is effective or predatory. How can we be surprised when volatility rises? The buyer is recognizing that he does not have appropriate information.”

... But experienced investors know that a reliance on fantasy will only prolong the pain that is racking the huge and important mortgage market.”
Tanta: “Experienced investors,” ahem, are presumably those who read the prospectus and the servicing agreement. They might even read the delinquency reports. They might even be investing in REMICs where loss mit options are extremely limited by REMIC law. They might even be buying No Doc subprime high-LTV loans, which are, ahem, pretty much guaranteed to generate excessive numbers of losers. But is that a fantasy? Nooooo. The “fantasy” is, apparently, that some servicers are probably trying to make loss-mit lemonade instead of flooding the market with REO. I’ll agree that that’s probably, in some of these cases, just prolonging the pain. But this business of crying for those innocent investors is getting on every nerve I have left. Next thing you know, someone is going to tell me that shareholders in weren’t told that it didn’t have a business plan, and I’m going to be so shocked I’ll fall off my Aeron chair.