naked capitalism

Dear readers, I had wanted to go through the geeky American CoreLogic Mortgage Resets study, which gives mind-numbing detail on what type of mortgages reset when to come up with a more refined guesstimate of how many borrowers might be eligible for the formerly-Paulson, now Administration subprime program.

The Administration claims its program could help as many as 1.2 million borrowers, or two-thirds of the subprime cohort. That number sounds like pure PR. Non-profit housing experts put the numbers much lower. The Greenlining Institute pegs the percentage of subprime borrowers who would qualify at only 12% (240,000); the Center for Responsible Lending estimates only 145,000 would benefit.

But further analysis of the numbers will have to wait. I wanted to address one topic now, which is whether the plan could be successfully challenged by unhappy investors. My reading is yes. Why do I think so? Beyond the fact that lawyers are very clever and there are doubtless some flies in the ointment, there is at least one obvious basis for attack. The American Securitization Forum, one of the key members of the so-called New Hope Alliance, has, in endorsing this plan, done a 180 degree turn its recommendations fo industry standards for loan mod procedures, and in a short period of time to boot.

Tanta at Calculated Risk’s initial (and detailed) discussion argued that the plan was awfully convoluted and accomplished comparatively little because it was designed not to run afoul of current agreements (note that there is no “safe harbor” or other legislation proposed by the Administration to shield mortgage servicers, the ones who would be making the loan modifications). But the American Securitization Forum may be hoist on its own petard.

Please note very carefully what I am saying here. I am not predicting whether investors will sue; that depends upon whether they think the plan will have an adverse impact on them, and whether it is big enough to merit the cost of filing a suit. They may have very good legal grounds but may feel it’s not worth the bother.

The other notion to understand clearly is what I mean by “successfully challenge.” I am not predicting that a plaintiff will win in court; 95% of lawsuits are settled and never reach trial. But one can prevail either by stopping or limiting behavior one doesn’t like, or by achieving a reasonable negotiated settlement.

How do those things happen? In all candor, it does not depend entirely on the merits of one’s case. It depends on being able to mount an argument that is strong enough to survive a motion for summary judgment (that is a merits of the case issue). Then the outcome can be significantly influenced by how costly, unpleasant, and embarrassing one can make the discovery process. Success in litigation often depends on the pain one can inflict on one’s opponents. For example, if you have legitimate reasons to depose senior executives of your opponent’s important clients, and the line of questioning would make them uncomfortable or better yet, damage the relationship with your opponent, that alone might lead to settlement negotiations.

A quick reading of the documents at the American Securitization website on the subprime plan suggest some clear avenues for a suit. And I am sure clever attorneys who had access to the servicing agreements could come up with much better and more specific ideas. Any lawyers or investors who have litigation experience are particularly encouraged to weigh in.

The plan goes to some effort to camouflage the elephant in the room, namely, the American Securitization Forum’s repudiation of its former position. Let’s first review the basic parameters. Michael Shedock has posted the key terms at his site, which is a bit faster than registering for free at American Banker). I’ve simplified this a bit but it’s pretty faithful:

Borrowers who are deemed to be “eligible to refinance into other products,” aka Segment 1, are out for the most part, although the servicers can include them on a case-by-case basis. Having a FICO of over 660 or a LTV or CLTV better than 97% puts you in this group. Borrowers also must be current, which means presently not more than 30 days past due and at most over 60 days past due only once in the last 12 months (under the Mortgage Bankers Association definition, not more than 90 days past due more than once). The homeowner must not be eligible for FHA Fast Secure and must occupy the house (determination based on borrower’s representation).

Segment 2, the so-called “fast track,” will have FICOs of less than 660, LTV higher than 97% . Borrowers whose FICOs have improved 10% fail the FICO test (as do ones whose FICOs are greater than 660) but can be evaluated on a case-by-case basis.

Segment 3 are the bottom tier, not eligible for fast track, but still may be able to get a conventional mod (in theory, of course).

Now to the fun part. In June 2007, the American Securitization Forum, which was depicted in the press as a representative of investors in the New Hope Alliance (in fact, its members include issuers, rating agencies, financial intermediaries, guarantors, law firms, accounting firms, so it would hardly seem able to represent any one group) published a “Statement of Principles, Recommendations and Guidelines for the Modification of Securitized Subprime Residential Mortgage Loans” (starts on page 18 of this pdf, boldface mine).

The overall purpose of this Statement is to provide guidance for servicers modifying subprime residential mortgage loans that are included in a securitization….

Loan modifications should be considered and made on a loan-by-loan basis, taking into account the unique combination of circumstances for each loan and borrower, including the borrower’s current ability to pay. The ASF is opposed to any across-the-board approach to loan modifications, and to any approach that would have all modifications structured in a particular manner. The ASF is also opposed to any proposals that would provide an across-the-board moratorium or delay period on foreclosures.

Generally, the ASF believes that loan modifications should only be made….

f. Where there is a reasonable basis for the servicer concluding that the borrower will be able to make the scheduled payments as modified; and

g. In a manner that is designed to provide sustainable and long-term solutions, but does not reduce the required payments beyond the magnitude required to return the loan to performing status, or beyond the anticipated period of borrower need…..

In evaluating whether a proposed loan modification will maximize recoveries to the investors, the servicer should compare the anticipated recovery under the loan modification to the anticipated recovery through foreclosure on a net present value basis. Whichever action is determined by the servicer to maximize recovery should be deemed to be in the best interests of the investors.

Now a lawyer is going to have a field day with this. It’s a given that considerable thought and effort went into this document. Yet a mere six months later, the same organization, due to pressure from the Administration, repudiates some of the key precepts it outlined. In particular, the June document argued for a case by case approach, no “one size fits all” mods, that any mods were (in essence) to get as much from the borrower as was feasible, any reduction was to be limited to the period of borrower need, and the servicer was to compare the attractiveness of the deal versus the recovery in a foreclosure (remember, these borrowers won’t default overnight. It might take a year or two and in some locations, the real estate market might have rebounded) . All these steps, recommended as proper practice, have now been thrown to the wayside.

The critical point is without taking these steps, there is no way of being certain a mod is the best outcome for the investor. That’s precisely why these measures were in place. I would assume that there is an argument to be made that this program is strictly for servicer convenience (yes, mods are often the best answer, but the servicers should staff up, do it right, and eat the costs, and not maintain their profit margins at investor expense).

The December American Securitization document “Streamlined Foreclosure and Loss Avoidance Framework for Securitized Subprime Adjustable Rate Mortgage Loans”, predictably, tries to finesse its earlier position:

As to Segment 2 loans eligible for a fast track loan modification, the servicer may make the following presumptions:

o The borrower is able to pay under the loan modification based on his or her current payment history prior to the reset date.

o The borrower is willing to pay under the loan modification, as evidenced by a) an agreement to the modification after being contacted or b) in the event that the affirmative agreement of the borrower cannot be obtained, the borrower’s
payment of two payments under the loan as modified after receiving notice of the modified terms.

o The borrower is unable to pay (and default is reasonably foreseeable) after the upcoming reset under the original loan terms, based on the size of the payment increase that would otherwise apply.

o The modification maximizes the net present value of recoveries to the securitization trust and is in the best interests of investors in the aggregate, because refinancing opportunities are likely not available and the borrower is able and willing to pay under the modified terms.

“May make the following presumptions” is an awfully strong assertion for a document that does not have the force of law. The narrative might even sound reasonable until you compare it with industry practice (and as Tanta has pointed out repeatedly, industry practice ain’t what it used to be).

In case you were wondering, the second dot point deals with the scenario in which the servicer is unable to contact the borrower but decides to offer the mod.

Note dot point thred. The borrower is presumed to be unable to pay. The outlines (both American Banker and American Securitization Forum) look to ONLY the size of the reset, the FICO, the LTV, and the payment history to make that determination. Payment history and FICO alone can probably be successfully attacked as an inadequate basis for making a determination (expect duelling experts here).

Similarly, dot point four, the mod is “presumed” to be the best solution; no reference to foreclosure, either currently (remember, these people are not in default, or even delinquency, so if and when they will default is subject to debate) or prospectively. This is the equivalent of a big wave of a magic wand in place of analysis.

Here is the section that specifically tries to get around the earlier statement (boldface mine):

The June 2007 Statement provides further interpretive guidance that in evaluating a loan modification, the servicer should compare the anticipated recovery under the loan modification with the anticipated recovery through foreclosure on a net present value basis, that whichever action maximizes recovery should be deemed in the best interests of investors, and also that the standard “in the best interests of” the investors should be interpreted by reference to the investors in that securitization in the aggregate.

For any securitized pool and set of fast track loan modifications, as to each loan in the group, the servicer will have determined individually that the borrower is not able to refinance, that the borrower is able to pay at the current rate, and that there is a reasonable basis for believing that the borrower will not be able to make payments under the loan as originally required after the upcoming reset date.

In light of current market conditions including home value trends, it appears that key elements of any net present value determination (such as default rates with or without a modification, and loss severities) cannot be accurately predicted based on historic data. Nevertheless, we believe that a servicer can appropriately take the view that a group of loans modified under the fast track procedures for Segment 2 loans will in the aggregate result in a better recovery on a net present value basis, when comparing the reduction in interest payments that may result from the modifications with potential losses upon foreclosure that might have resulted absent the modifications. Accordingly, we believe that the methodology for making fast track loan modifications under Segment 2 complies with this guidance, and will result in action that is in the best interests of investors.

For the first bold section, the claim that the servicer “determined individually” based on crude and unreliable data is a stretch and subject to attack.

The second bold section is not only questionable but could be used to inflict pain on the people in New Hope Alliance involved in the negotiations. It’s an open invitation for an attorney to depose the people involved in the development of the plan and ascertain to what degree, if any, they did financial analysis (and remember, senior people hate being deposed). Given how quickly this plan was put together, and the language used, it’s a near certainty that they didn’t. Therefore, they are making assertions as to outcomes with no factual underpinning. The lawyers will be able to extract, in fact, that outcomes aren’t completely indeterminate (which is what that language asserts), that one could develop floors and ceilings nationally, and by local market, of what outcomes might be (houses are not going to double in price nor are they going to zero, and the estimates can be bounded more tightly than that. I suspect most people involved would agree that 20% below current recovery levels is probably a floor in most markets.)

So then the interrogation will continue: the New Hope Alliance could have set some default parameters and done a little financial modeling and added some screens related to recovery. And the attorneys will have great fun waving sheaves of output from the very detailed databases on loan resets and pointing out that the coalition spent a lot of time on this part of the equation, why not on the other? Believe me, it will be very easy to show these guys punted on this question. I am fairly confident that given Paulson’s eagerness to announce a plan in a week, they did not do the CYA measures to forestall this line of attack.

I have seen some statements that said the services would rely on borrower representations as to income. This opens up another line of legal challenge. Why no verification? Borrowers have an incentive to understate income so as not to be deemed to be too well off and be subject to a reset. The experience with no-doc loans has demonstrated beyond doubt that borrowers (many of these very same borrowers) lied. We’ve seen estimates that 40% to 60% of the no-doc borrowers misrepresented their income to a significant degree; the subprime plan is in effect no doc refis.

So we have another line of attack: borrower representations, particularly in this segment of the market, have been shown beyond a doubt to be rife with fraud, Imagine the attorneys making the various originators and servicers go over their reports to the FBI. It’s undeniable that the members of the New Hope Alliance were well aware of this problem. Yet they are recommending the use of the same certain-to-be-rife-with-fraud process to force payment reductions on to investors. It’s one thing when they started this no-doc mess; no one anticipated how much fraud would result (I personally don’t buy that, but it is a defensible argument). Now they know better. How can they justify enabling fraud. particularly at the further expense of investors, when their own guidelines call for them to combat it?

Now having said that, there are some interesting wrinkles in the mod plan:

For borrowers that are eligible for a fast track modification, the fast track option is non-exclusive and does not preclude a servicer from using an alternate analysis to determine if a borrower is eligible for a loan modification, as well as the terms of the modification.

This raises the possibility that the servicers might not offer “one size fits all” mods after all. As the Financial Times said:

Sheila Bair, chairman of the Federal Deposit Insurance Corporation, said in response to questions about the use of credit scores as a screening device that the plan would be undergoing “further refinements”.

It will be interesting to see whether that turns out to mean they will flesh out the many yet unclear procedural details, or will wind up beating a retreat on some of the initial parameters.