by Tanta on 2/01/2008 01:33:00 PM

FDIC Chair Sheila Bair gave another speech the other day about mortgage loan modifications. Short version: the fact that this isn't all just about subprime 2/28s that can be fixed with a rate freeze has become obvious. The solution has not.

Unfortunately, some borrowers pose even more difficult issues because their debt far exceeds the value of their homes. Servicers have always had to evaluate whether the best option in these cases is foreclosure or some other process, such as a short sale, that results in the loss of the home. There may be no alternative except foreclosure for loans that were made to speculators, under fraudulent circumstances, or to borrowers who have no reasonable ability to repay (even with restructuring). However, in today's market, servicers should carefully consider whether some writedowns of part of the principal balance to the value of the home or forgiveness of arrearages of principal and interest are better options than foreclosure or even short sales in appropriate circumstances. Permitting borrowers with an ability to make reasonable payments to stay in their home would provide greater value to lenders and investors than forcing foreclosures that undercut the value of the property and harm the value of other properties in the neighborhood.

Until recently, strategies involving writing down the value of the loan did not provide a feasible alternative for most borrowers. When lenders restructured loans in this manner, borrowers faced a potential tax liability on the amount of the forgiven debt.

Last month, however, Congress addressed the issue of tax liability for mortgage debt forgiveness in a way that makes long-term workouts involving principal writedowns a reasonable alternative to foreclosure. Such an option might be considered for borrowers having financial difficulty making their payments after their loans reset and where foreclosure is a looming possibility. Congress is to be commended for enhancing the workout options available to borrowers and lenders for negotiating long-term, sustainable restructurings.

Enactment of the Mortgage Forgiveness Debt Relief Act of 2007 provides an additional option for keeping borrowers in their home. This Act recognizes that cash strapped borrowers who are already facing financial difficulty cannot afford a potential tax liability that could hinder their ability to make their modified loan payments. It also provides greater assurance to lenders and servicers that borrowers will be able to perform after their loans are modified and decreasing the principal value will decrease the loan to value ratio, thereby potentially expanding the number of homeowners who could qualify for GSE refinancing. This will allow lenders and servicers to consider forgiving a portion of the principal balance owed to a level a borrower can realistically afford to repay, as long as it produces a net present value that is greater than the anticipated net recovery that would result from a foreclosure. This would require lenders and servicers to ascertain the existence and amount of any second mortgages, and obtain releases from these obligations to the extent appropriate. While this type of modification results in the recognition of a loss by the lender or servicer, it is virtually certain that the amount of the principal write-down will be less than the amount of loss sustained from foreclosure in today's market.

Permanently forgiving part of the principal amount can provide a better financial result for investors than foreclosure by creating long-term, sustainable solutions that will allow borrowers to stay in their homes. This approach also has the added benefit of limiting the overall adverse affect of declining property values on communities.

Let's go through this carefully:

1. It excludes "loans that were made to speculators, under fraudulent circumstances, or to borrowers who have no reasonable ability to repay (even with restructuring)." This suggests that the very first thing any servicer has to do is sufficient rooting around in the file, plus looking at updated documents (a new credit report, new income verification from the borrowers, evidence of occupancy) to eliminate this group. So we aren't talking about "streamlined" mods (such as the "teaser freezer") any longer. Such a process will, necessarily, also identify any borrowers who can afford their current loan, but just don't want to pay it. (You could include them with "speculators," but I'm not entirely sure that's what Bair means by the term). That means that the servicer will have to decide what to do for those borrowers. I am simply pointing out a fact here. In order to exclude those who could not afford even the modified terms, you will identify those who can afford the current terms. I'm afraid this might work out like the old James Bond movie: Now you know. So we have to kill you. That is, now you can be sued by some pissed-off investor. If you didn't do the due diligence first, you could have claimed ignorance, but then you'd get sued for doing a workout that merely postpones (and thus increases the cost of) foreclosure for borrowers who can't afford any terms.

2. "Such an option might be considered for borrowers having financial difficulty making their payments after their loans reset and where foreclosure is a looming possibility." Except for rhetorical effect, it isn't clear why resets have any relevance here. The proposal is paring down principal. If you have to reduce the principal as well as freeze the rate, then the pre-reset payment was obviously not affordable in the first place. I do not understand why Bair is muddying this up with references to payment resets. I am going to assume that the borrower is not required to even have an ARM to be eligible for this kind of workout, let alone be required to be facing imminent rate reset. If we're willing to abandon the "it's just subprime" thing, can we go whole hog and imagine that it's not just ARMs?

3. "It also provides greater assurance to lenders and servicers that borrowers will be able to perform after their loans are modified and decreasing the principal value will decrease the loan to value ratio, thereby potentially expanding the number of homeowners who could qualify for GSE refinancing." I understand the first part of that sentence: if borrowers don't have to pay a big tax bill, they have more money to apply to mortgage payments. I don't understand the second part. Are we trying to modify a loan (retain it at a lower balance) or refinance it (move it onto another lender's books, having accepted less than 100% of the balance in the payoff)? If we are trying to forgive enough principal to make the loan refinance-eligible, then we are practically speaking waiving more than the difference between the loan amount and the current market value of the property (assuming that there aren't many 100% LTV refis out there in declining markets). You are certainly getting very close there to exceeding the cost to the investor of foreclosure. In terms of borrower motivations, I should think that in any state where purchase-money loans are non-recourse (and "troubled debt restructurings" are not considered refinances), your truly ruthless borrower doesn't want a refi. That might kind of adversely select your modified pool.

4. "This will allow lenders and servicers to consider forgiving a portion of the principal balance owed to a level a borrower can realistically afford to repay, as long as it produces a net present value that is greater than the anticipated net recovery that would result from a foreclosure." Actually, I understand the rule to be that the modification has to produce a net present value greater than any other option, including but not limited to foreclosure. We just went through this: if short sale or deed-in-lieu produces a greater recovery to the investor, then the servicer is obligated to pursue it. You can talk all you want about what FCs do to neighborhoods or short sales do to families, but there seems to be some agreement that servicers are obligated to do only what maximizes recoveries to the investor on this loan.

The "rate freeze" thing is actually pretty straightforward compared to a principal reduction. In a rate freeze, you are assuming that you end up getting the full principal balance back, at the lower interest rate, over the remaining term of the loan. You calculate what you lose in reduced interest income over the term of the freeze, compare that to the present value of a foreclosure recovery, and the former is almost always a better deal than the latter. The FC usually wins only if the house is not very far underwater on the one hand and the borrower's repayment ability is so shaky on the other that you simply can't assume the modified loan is fully collectable. But if the modified loan is so shaky that it has to be discounted to less than FC recovery, you really don't have the choice of modifying in the first place.

Doing a principal balance reduction is (traditionally) on the table in two scenarios: one, the delinquent borrower can resume payments (perhaps at a modified interest rate), but can't catch up, so you forgive the past-due amount. There's no magic here to calculating the amount to forgive. You discount the value of the loan by the amount of the principal reduction, compare that to FC recovery, and once again take your lowest loss. As this situation generally involves no more than 3-6 months of past due interest, escrow balance, and fees, the write-off amount is practically quite close to foreclosure expenses, less the RE broker commission and actual court costs. Once again, with any reasonable estimate of collectability (performance of the modified loan), this almost always beats FC unless current LTV is quite low, which it probably isn't.

The second scenario involves going at the problem not from interest rate or LTV but payment: you have a borrower who cannot afford the contractual payment, but you have agreed to some lower payment that the borrower can (verifiably) afford (and has indicated a willingness to pay). In this case you are, in essence, just backing into a new balance (and interest rate and term to maturity, in some combination) that produces the payment you want. How that stacks up against FC recovery is a mere question of how much lower that payment is than the contractual payment, of course. In earlier bouts of economic distress, borrowers were qualified at much more reasonable payments (and higher rates and shorter terms) to start with. In the current environment, we're having much more trouble with these calculations.

Now we're introducing a third scenario: simply reducing the principal balance of the loan to the market value of the property (or somewhat less than that) in order to keep the borrower right-side up or treading water, and thus to prevent default rather than cure it. If you are no longer trying to arrive at the highest payment the borrower can practically make, you are no longer producing calculations of loss that are mostly likely to be a better deal than FC. This is where Bair is implicitly--but not explicitly--asking servicers to use something other than net present value of FC recovery based on current market prices and marketing time estimates of the subject loan. She is asking the servicer to use possible future FC recoveries based on the assumption that all servicers will fail to forgive loan balances (i.e., that FC recoveries on other loans will be worse than they would have been if you foreclose this loan), and collectability estimates on the pared-down loan based on the assumption that all servicers will forgive loan balances (which will reduce FC inventory and stabilize prices, therefore leading to fewer future defaults). Moral hazard, meet tragedy of the commons.

In the first case, you're asking MBS investors to "take one for the team." I have no problem with that idea in theory, but as a practical matter I'd rather ask a cat to quit developing hairballs. The mathematics of "net present value to the trust" means this trust, not somebody else's trust and not some other loan. You can present this as self-interest: the less you FC, the more you are contributing to stable RE prices, but you will run into someone who realizes that if every other servicer of every other pool is forbearing, and therefore reducing REO inventory, I should go ahead and FC now while I can. Hence Bair brings in the perfectly respectable but, in the MBS context, rather alien concern about protecting communities. Of course some clever economist could calculate that "externality" and add it to the NPV calculation for us, but there isn't anything in these PSAs about that. And if you don't think we can't buy our own economists who will come up with a different number, you don't know us very well. (Bair, I think, is used to dealing with federally chartered depositories and the GSEs, who can to a certain extent be bullied into "social responsibility" because of their charters. But if all the lenders in question were owners of a federal charter, we probably wouldn't be having this conversation.)

Without the "externality" adjustment, and without the assumption that we are aiming for the highest payment the borrower can afford short of the contractual payment, you simply have a situation in which loans are going to fail unless and until the servicer forgives sufficient principal to match the liquidation value of the property. (You can try to simply match the "fair market value" of the property instead of the "REO liquidation value," but that assumes that FMV isn't converging on liquidation value and that failure to match REO prices will stop defaults sufficiently.) There is therefore no obvious loss mitigation here, absent the assumption that this will arrest the slide in RE values. If you happen to believe that REO inventory isn't the only downward pressure on prices, you risk having to pare these loans down every quarter until the bubble blows off. That builds enough "re-defaults" into your modification analysis model that any given set of individual loan inputs come up "FC" again.

4. "This would require lenders and servicers to ascertain the existence and amount of any second mortgages, and obtain releases from these obligations to the extent appropriate." Or you could just assume a can-opener. It is really hard to believe that Bair can gallop over this one in one short sentence. All you gotta do is make sure the junior liens are willing to extinguish themselves first, and you're golden.

The reality is that the "magic number" for the loans that are most likely to be facing trouble right now is 80% LTV on the first lien, and something up to 100% or beyond for one or more junior liens. So it is simply a fact that home value declines of 1-20% mean all the hit goes to the junior lien. If you are trying to "arrest" price declines before they get worse than 20%, you are in a situation where you aren't asking the first lien lender to forgive any principal to start with; you're just asking the second lien lender to release its lien for a reduced or no payoff. Given how many of those junior liens are in bank portfolios, it doesn't surprise me that Bair avoids coming right out and saying that banks will have to write off their junior liens before this plan gets any traction. But that's what this means.

Of course a junior lienholder may have no alternative to simply forgiving the entire debt and releasing its lien (since it would probably collect nothing in a foreclosure). But if the value decline isn't "eating into" the first lien yet, why is this a negotiation with the first lien lender? Borrowers should just be calling their junior lienholders asking to have their junior liens forgiven. If the first lien lender has to forgive principal too, then realistically price declines are greater than 20%. (A likely scenario is that the junior lienholder wants to be paid to go away, so the first lien lender is trying to "advance" a couple thousand dollars to the second lien lender in order to get the second lien cleared. That "advance" then gets "forgiven" on the first lien loan, so the effect on a 80/20 deal is that the second lienholder wrote off 18% and the first lien holder wrote off 2% or something like that. If you can get the second lienholder to bother taking a couple grand to go away.)

But at that point, we aren't "staving off" major RE market failure, we're in it. We are certainly in it if we have to forgive all the junior liens and some portion of the senior liens. Either we're far enough down that the end is possibly in sight, or we aren't. At the end of it, Bair's proposal is just a recognition that we have blown through all the "credit enhancement" that first lien lenders thought they had. In order for there to be any "risk-based pricing" on the resulting loans--which are all subprime, now--the Fed will have to keep cutting rates to zero, as far as I can tell. The best case scenario for first-lien lenders (banks or securities) is then a long lean period of years in which you have low-yielding fixed rate loans outstanding that won't go anywhere until amortization (rather than appreciation) builds up some equity. Great. We're all GSEs now.

If you care to know what I think about this as policy, the answer is that I doubt it matters at this point. Really all Bair is saying is that servicers should modify down to the point where it's no longer possible to go further without violating servicing contracts. Who actually disagrees with that? It is quite possible that it isn't any better than doing nothing, but it's possible that there are borrowers for whom it is, and I don't see how it could be worse than doing nothing. So go ahead, everybody. As long as we all go first, there shouldn't be any problems.

However, if Bair thinks this plan will reduce stress on the FDIC, she's crazier than I am by a long shot. When there is no longer "credit enhancement" on these loans, Ms. Bair, you're the credit enhancement.