"Bear Stearns": Marktmanipulationen?

Tanta: Modifications, Buybacks, True Sales, and Puzzlement

Yes, friends, Tanta is even more puzzled than usual when the subject is hedge funds. May I beg our readers who have more familiarity with this part of the financial markets to help us out here?

The Financial Times and a number of other sources have reported lately that hedge funds are accusing the investment banks--Bear Stearns specifically--of "market manipulation" by modifying deliquent or soon-to-be delinquent securitized mortgages. The hedgies' interest in all this, it appears, is not that they own these bonds--although that is hardly clear to me--but that they have been on the other side of some credit default swaps which means they will lose if the bonds perform better than anticipated.

Reuters reports that the WSJ reports that the issue is "purchases":

NEW YORK (Reuters) - Hedge fund managers are accusing Bear Stearns Cos. of trying to manipulate the market in securities based on subprime mortgages, the Wall Street Journal reported in its online edition.

The confrontation provides a rare look into the complex trading in the mammoth U.S. mortgage market, which played a critical role in financing the housing boom, and the complicated relationships between hedge funds and investment banks, the paper said.

Hedge funds that had sold short such securities made profits when an index tied to a basket of subprime bonds was falling. But the index has recovered in recent weeks, leading to howls of protest from hedge funds, according to the report.

The chief critic, John Paulson of Paulson & Co., a $12 billion fund, says Bear Stearns wanted to prop up faltering mortgages-backed securities by purchasing individual mortgages that were rapidly losing value to avoid doling out billions in swap payments, the Journal reported.

I am not a subscriber to the Wall Street Journal; perhaps someone who read the original article can help us understand this.

Whatever do they mean by "had sold short such securities"? Are we really talking about a CDS trade? No doubt the "index" in question is the ABX, but can anyone help me decipher the actual trade here?

Furthermore, what could "purchasing individual mortgages that were rapidly losing value" possibly mean in this context? Are we talking about repurchases (EPDs and other rep/warranty failures)? Bear would presumably be the "purchaser" in this case if it had originated the loan (or at least sold it to the security trustee), which would make Bear responsible for taking it back (and, in turn, forcing it back onto whatever hapless originator sold it to Bear in the first place).

Much ink has been spilled on the subject of Bear (and other IBs, particularly Merrill) forcing loans back to originators to the extent of forcing originators into bankruptcy. The implication of the Reuters piece is that Bear, specifically, is accused of doing this not to clean up the security (remove the defective loans at a full payoff of principal and accrued interest to the bondholder) but to avoid having to pay out to the hedge counterparties.

If that's what we're talking about--and feel free to correct me if I'm wrong--then we have a nice can of worms here. Contractual buyback provisions are supposed to protect bondholders from defective loan collateral that can sneak into those securities precisely because the loans are sold on a rep and warranty basis. You can do limited (define that any way you choose) due diligence on the loan collateral itself, relying on the data tapes (the specific "representations" supplied by the loan originator) plus a custodian's report on the presence and acceptability of the note and mortgage documents, because the contract allows you to put back anything that violates those reps.

In addition, these contracts almost always specify that a loan is repurchased at par (the price is the unpaid principal balance plus any interest adjustment for the sale timing). That can easily mean that the original buyer bought the loan for 102 and is putting it back at 100. It almost always means that the originator is buying it back at 100 and going to have to sell it for a lot less than that. The idea is that nobody wins, particularly, in this situation.

Nobody is supposed to win. The traditional "par repurchase" is supposed to mitigate certain kinds of moral hazard. In any case, a par repurchase has exactly the same effect on the bondholder as a refinance (full return of principal). In terms of the reported credit quality of the security, it removes a delinquent loan or a loan that looks like it's going to become delinquent or (in the case of certain kinds of fraud) a loan that looks like it will be hard to recover anything from in foreclosure. Some investors are willing to give the benefit of the doubt on some kinds of loan defects; no one (sane) does anything with a loan that has title or legal mortgage problems except put it back. We have been reading about lenders who suddenly find they cannot foreclose or take possession because of sloppy loan closing or mortgage assignment practices. This threat is real.

The problem appears to be the identity of interest issue: the security sponsor owes the investors the duty of selling defective loans back to the originator at par. This is supposed to be expensive for all parties (think of what you "saved" by not doing your due diligence), but less so than not doing it (the same logic applies to modifications). However, the hedgies seem to be alleging that Bear will also either profit or avoid losses on its CDS trades by forcing these buybacks. I don't know about you all, but it sounds to me like time to clarify how many hats Bear is wearing here: does it own any classes of any of these securities? Is it the sponsor, the servicer, and/or the originator of these loans? Is it buying or selling credit protection? Whose position is it hedging?

Whether we are talking about modifications or repurchases or both, all of this raises some ugly accounting issues, it seems. Reuters, again:

FAS 140 governs whether a bank can treat assets held in various asset-backed securities as sales or secured financing. If the asset is treated as a sale, it allows banks to keep it off their balance sheets.

Banks say the standard prevents them from helping borrowers modify loans easily. But market experts have said the complications are also a legal issue related to the terms set out at the time a mortgage-backed security is created.

The board has revisited FAS 140 several times since it was approved in 2000, but Seidman said on Wednesday it has not gotten any easier.

"What has become clear to me is that, when we look at the way investors and analysts treat securities transactions, if there hasn't been a free and clear sale, they are unwinding the accounting and putting assets and liabilities back on the books," Seidman said. "I've come to the conclusion that ... we're going in the wrong direction -- trying to maintain a standard that's taking assets off the books when investors view it as economically still associated with the seller."

Some more clarity on this issue would certainly be helpful. If I am reading this correctly, the issue is whether the sale of the underlying mortgage loans to the security trust is a "true sale" or not. If it is, the securitization is truly "off balance sheet." If it isn't, the securitization is on-balance sheet financing of the originator's mortgage loans.

That, in turn, has something to do with how the securitizer books gain (or loss) on the securitization transaction, but for our present purposes it seems the issue is how much a "true sale" a transaction treated as off-balance sheet can be if the securitizer can, in essence, "control" the collateral (by modifying it or by forcing the trustee to sell loans back to the issuer). Traditionally, when there is "recourse" in a sale, it generally has to be treated as a financing rather than a true sale, since the seller retains not just an obligation for performance of the collateral but presumably then some rights to make good on that obligation in ways (substitute loans, repurchases, mods) that may benefit the seller as much as the buyer. My guess here--I'm just reading the news, folks, so it's a guess--is that the Financial Accounting Standards Board is fixin' to possibly decide that some of these "nonrecourse" transactions are, actually, recourse transactions, and not true sales, and therefore not on the right balance sheet, and ugly ugly ugly.

So, if I'm following all this, a torrent of buybacks which has already forced a lot of originators into bankruptcy and evaporated a lot of market cap at the same time it has weakened some servicers to the point that nobody's even sure about getting payments collected on the remaining performing loans, not to mention accomplishing those foreclosures for the ones the security couldn't get out from under, has now called into question the accounting basis of the whole deal, which means potentially evil ugly nasty restatements for anyone still in a position (this side of BK) to restate, and at the same time it has dragged the curtain away from this "all risk has been hedged" mantra to show that behind that curtain, the transactions supposed to protect investors' interests are in fact so riddled with conflicts of interest that the liability those who thought they were purchasing "credit protection" may have to the unhappy campers who were providing that credit protection could more than overwhelm the benefit of the insurance.