by Calculated Risk on 6/07/2009 12:24:00 AM

Note: The following is an article that Tanta wrote for our April 2008 newsletter. This has not appeared on the blog before (ht Jillayne for reminding me of this).

As CR mentions in his essay on housing supply in this issue, plenty of markets are seeing very large segments of for-sale inventory coming in the form of short sale listings. Yet completed short sales remain a small segment of actual “distressed sales” or final dispositions of “worked out” loans. It is difficult to get reliable current data on short sales completed; the MBA reported in its first major analysis of workout data back in Q3 2007 that in that quarter, there were around 384,000 foreclosure starts and about 9,000 short sales completed. Certainly we have a lot of anecdotal evidence in the mainstream news media and from real estate agent reports suggesting that while short sales were never especially easy to accomplish, they are getting less so as time goes on.

Because so many observers of the scene express endless frustration that short sales have not been embraced by lenders as a “solution” to the foreclosure crisis—and no doubt because many potential real estate investors may be thinking that buying in a short sale is a good way to find a bargain—I thought it would be worthwhile to examine some of the reasons why they just don’t happen as often as one might expect. Part of the problem is overwhelmed servicers dealing with securitization documents that give them limited negotiating room, but that isn’t all of it. There is, I think, a certain psychological issue with two of the important participants in a short sale scenario other than the servicer, namely the current owner and the potential buyer, and this issue is causing short sale proposals that just don’t pass muster with servicers.

We should start by reviewing why it is that a servicer might be inclined to accept a short sale: the idea is that the loss incurred is less than the loss that would be incurred in a foreclosure. To start, then, the servicer must have some model it can use to project likely losses in foreclosure. A good deal of the data going into that model involves not just a projection of the likely sales price of the eventual REO, but also the servicer’s estimates of the costs and expenses of foreclosure, which the short sale is supposed to be minimizing or eliminating. This modeling is dynamic: as a local RE market changes (prices fall, inventory mounts, timelines for foreclosure extend), the model has to keep updating its loss projections.

There is no particular “rule of thumb” for what loss severity in a foreclosure is in any market, but one can start with a fairly simple example of where the losses come from by looking at some aggregated national data prepared by two Freddie Mac economists, Amy Crews Cutts and William A. Merrill, in a recent paper entitled “Interventions in Mortgage Default: Policies and Practices to Prevent Home Loss and Lower Costs” (available for download at www. freddiemac.com). The authors estimate the “breakout” of foreclosure losses on Freddie Mac (conforming dollar, prime or near-prime credit quality mortgages) in the period up to the third quarter of 2007 as follows:

• 20% Principal loss (the difference between the unpaid principal balance or UPB of the loan and the REO sales price, or the amount of the borrower’s “negative equity”)

• 37% Pre-Foreclosure Expenses (24% interest expense, 4% legal fees, 8% property costs such as taxes, insurance, assessments and utilities)

• 43% Post-Foreclosure Expenses (21% sales commission/concessions, 2% legal fees, 20% property costs and maintenance/repair

To take a simple example, then, in a market which is currently experiencing a “loss severity” of 40% (of the unpaid principal balance or UPB of the loan), the breakout would look something like this on average (using a smaller than “average” loan balance just to simplify the math):

Loan Amount	100,000
REO Sales Price	92,000
UPB Loss	8,000
Pre-FC Expense	14,800
Post-FC Expense	17,200
Total Loss	40,000
Loss Severity	40%

A short sale, then, would be of interest to the servicer only if it could result in a loss severity less than 40%. Of course, we do not necessarily expect the buyer in a short sale scenario to be offering $92,000 in this example, because that is not, in the buyer’s mind, a “discount” from the price the property would fetch as REO. The idea is that the servicer can be brought to accept a lower price for the property today than it could (potentially) get for the REO, because the short sale saves on the “foreclosure expenses.” Much frustration is being expressed by certain market participants or buyers who are nonplussed by servicers who apparently don’t want to “save” this money by accepting low-ball offers.

Part of the difficulty here is coming to terms with what the components of a servicer’s expenses are, and also what a difference it makes whether the current owner is delinquent (and cash-strapped) or not. It simply isn’t clear to me that everyone who is listing a property as a short sale these days is, actually, delinquent yet. That certainly makes a difference to a servicer’s willingness to negotiate, but it also makes a big difference in terms of what the cost comparisons are.

We also need to consider questions of timing. We looked in last month’s newsletter at some comparative foreclosure timeline numbers, making the point that they are dependent on the state’s laws as much if not more so that the servicer’s capacity or efficiency, although the latter is certainly significant. A short sale, more or less by definition, needs to be a “quick sale.” Delinquent borrowers do not have time to “expose” the property to “typically motivated” buyers for long enough to get the “optimum” price for the property; that’s why they’re facing foreclosure. Non-delinquent borrowers, presumably, can. However, they don’t want to deal with the extraordinarily extended “exposure” time required in a thorough-going RE bust, which moves from a matter of months to, apparently, a matter of years. Short sales offered by non-delinquent sellers fall into this odd sort of time-continuum: not as “quick” as a distressed borrower, but much “quicker” than a borrower seeking at least his loan amount. This no doubt accounts for a lot of the “failure to communicate” between non-delinquent borrower and servicer that we’re hearing about these days.

Let’s look at a theoretical set of alternative transactions based on the average numbers I laid out above. Please note that what I’m trying to capture here is the “mathematics of psychology,” not precisely the mathematics of “reality.” That will, I hope, be clearer after we look at some numbers:

	Foreclosure With 40% Loss Severity	Short Sale with Delinquent Owner		Short Sale with Non-Delinquent Owner
		Deal Fails	Deal Passes	"Typical" Exposure	"Quick Sale"
Loan Amount	100,000	100,000	100,000	100,000	100,000
"Gross" Sales Price	92,000	80,000	84,000	80,000	80,000
Sales Expense to Buyer	-	8,400	8,400	-	21,200
"Net" Sales Price	92,000	71,600	75,600	80,000	58,800
UPB Loss	8,000	28,400	24,400	20,000	41,200
Pre-FC Legal & Property	5,200	5,200	5,200	2,000	2,000
Pre-FC Interest Expense	9,600	9,600	9,600	-	-
Post FC Legal & Property	8,800	-	-	-	-
Post-FC Sales Expense	8,400	-	-	-	-
Total Loss:	40,000	43,200	39,200	22,000	43,200
"Discount" From REO Price	-	22%	18%	13%	36%
Post-FC Sales Expense to Owner	-	-	-	8,400	-
Pre-FC Interest Expense to Owner	-	-	-	9,600	-
Pre-FC Property Expense to Owner	-	-	-	3,200	-

In the foreclosure scenario, all costs to maintain the property, as well as cover interest payments on the outstanding loan balance due to the investor, are carried by the servicer and reimbursed from liquidation proceeds. This includes the “post-FC” costs of sales commissions and typical sales concessions (such as paying all or part of a buyer’s closing costs).

In our first example of a short sale with a delinquent owner, the buyer offers $80,000 for the property. However, we are going to assume a cash-strapped seller who is unable to cover the $8,400 in commissions/ concessions that would, in the foreclosure, have been paid by the servicer, and so we will assume that these costs are “passed on” to the buyer of the home, resulting in a net sales price of $71,600, or a 22% “discount” from the assumed REO price of $92,000. I don’t think it’s unreasonable to believe there might be buyers out there telling themselves that surely a servicer would rather lose 22% than 40%.

As we can see, however, it is really only the “post-FC” expenses that the servicer is “saving” in this deal, because the property seller has not, we assume, been making mortgage payments in a time-frame we will, for example purposes, consider to be equivalent to an “average” pre-FC period. You can, of course, and you would reduce the servicer’s expenses if the short sale were very fast, fast enough to make a meaningful difference in the servicer’s property expenses and interest costs. We’re just trying to keep the math constant to show, exactly, that you do have to sell not just “short” but “quick” to “pass” the servicer’s test here.

The second delinquent-owner column shows that the proposed deal doesn’t “pass” the servicer’s less-loss test until the sales price gets up to $84,000. That is still a “discount” from $92,000, if you choose to think of it that way, but not anywhere near a “lowball” offer.

The last two columns show how “psychological math” changes when we have a non-delinquent seller. In the first column, we assume a “rational” seller who understands that he needs to keep paying his mortgage until the sale takes place, meaning that he carries that “interest cost” and “pre-FC property expense” that the servicer would carry in the foreclosure scenario. Also, he will have to pony up at settlement for commission and typical concessions. If he is not particularly pressed for time here—he isn’t, after all, facing imminent foreclosure—then we’ll call his situation something like a “typical” marketing time for the property, which we are for simplicity’s sake making the equivalent of the foreclosure timeline. That scenario results in what should surely be an attractive deal to the servicer: a loss of only $22,000. With the property seller “paying” most of the costs that would otherwise have been the servicer’s. (I assume for example purposes that the legal costs of pre-foreclosure ($2,000) are equivalent to the legal/administrative costs to the servicer of processing a short sale.)

The only problem with that scenario, besides the fundamental question of the servicer’s motivation to accept loss when there is no delinquency to signal probable eventual foreclosure, is that this seller is displaying a set of behaviors I suggest might not be very common: wanting to sell quickly (compared to the years you might have to wait out a bust) but not too quickly (willing to pay out of pocket for sales commissions and market-driven concessions in order to get the best possible price in listing time comparable to the average 6-9 months of a foreclosure timeline). All while considering the monthly mortgage interest, taxes, and insurance as simply an inevitable expense that he needs to carry until he is approved by the servicer and relieved of ownership of the property.

My sense is that what we are actually seeing on the ground with non-delinquent short sale offers is probably closer to the second column: the one in which the seller in essence “decides” to “pass on” the sales costs and interest expense to the new buyer, to “get out without a loss.” (Again, I’m talking a certain kind of psychology here, not what you might call realistic math.) In this case, the seller presents the servicer with a “quick” low-ball price which will, he hopes, get him out before he incurs any more expense—expense here including the mortgage payment. The result of that? A loss to the servicer that is basically identical to the loss in the “failed” delinquent owner short sale.

In reality, of course, these costs that a “buyer” is taking are really costs that the servicer/investor is taking: it’s really the investor who ends up covering the transaction costs in most of these scenarios (because of the reduced sales price). Yet I don’t think enough prospective short-sellers have that in mind. Many, many people weren’t very good negotiators—or rational cost analysts—when they first bought these homes or took out these mortgages. It’s rather bizarre to think they’ll suddenly take a different view of things now, isn’t it? Our examples suggest that the deal most attractive to the servicer—the non-delinquent “typical” exposure or adequately-marketed property—is the one least likely to be offered. The “math of reality” works on these deals, but the “math of psychology” doesn’t.

My view is that, realistically, only the short sales with delinquent owners stand much of a chance of getting “passed” by servicers, and those few are being overtaken by events, namely, such a rapidly falling price environment that it’s hard to get that $84,000 bid and make it stick long enough to close the deal. This part is where servicer inability in so many cases to “just pull the trigger and close the deal” is hampering things, that being mostly a result of securitization rules requiring servicers to get investor/trustee approval, plus the endlessly updating numbers on those “models” that the servicers are using to find the “trigger point” where a short sale makes sense.

There are no doubt some bargains to be had for real estate investors looking into short sales, but for nearly anyone except an expert in certain markets, I think I’d suggest not wasting your time. Unless, that is, the listed “short” price has already been “pre-approved” by the owner’s servicer. That does sometimes happen; the seller will in that case have a letter from the servicer indicating the minimum allowable sales price/maximum concessions. It will also help to be working with a real estate agent who knows more than a little about doing short sales. Otherwise I suspect you’ll end up waiting for Godot. I hate to spoil the plot, if you’ve never seen that play, but in the end he never arrives.

Psychology of a Short Pig