Skin in the Game-True Sale Implications

There's a proposed joint federal rulemaking on the Dodd-Frank Act asset securitization risk retention requirements that would mandate some form of "skin-in-the-game" in most asset securitizations.  The proposed rule making provides a number of options for acceptable risk retention, and some exemptions from the requirement, but I think it raises a more fundamental problem for asset securitization, namely whether it will be possible to get clean true sale opinions with federally mandated risk retention.

The economics of many securitization deals hinge on the transfer of assets to the securitization vehicle being a "true sale," meaning that they are property of the securitization vehicle and not the assets' originator (or aggregator) and thus bankruptcy remote in the sense that they cannot be claimed as part of the originator or aggregator's bankruptcy estate.  (What constitutes a sale is arguably slightly different for bankruptcy, tax, accounting, and bank regulatory purposes, but that's not really germane here.)  Investors like securitized assets precisely because the assets are (in theory) impervious to the fate of the originator or aggregator.  (Obviously that isn't true if the originator/aggregator services or provides maintenance or warranties for the securitized asset.) 

Perhaps the most fundamental aspect of a sale is the transfer of risk and reward associated with an asset. If I transfer a mortgage on Blackacre to Prince William, but bear risk on the performance of the mortgage, it's questionable whether I have sold the mortgage to him or merely loaned or leased it to him. The federal risk retention requirements make this issue hairier.  If I am on the hook for the first 5% of the losses (or 5% of the total losses), could a bankruptcy trustee come after that mortgage as an asset of the estate? The fact that the 5% stake (vertical or first loss) is mandated by federal securities law strikes me as irrelevant to the true sale question, which is not a securities law issue, and on which Dodd-Frank takes no stance. If the transaction involves too much risk retention for whatever reason, regulatory requirement or voluntary deal design, there might be problems with the sale treatment.  I'd be curious if anyone's been thinking about this potential problem. 

Comments

[Ebenezer Scrooge]

There ain't no such thing as clean true sale opinion. They have more hedges than a British estate.
Ken Kettering did a very nice job with this a few years ago, in an article on legal structures that are too big to fail in court--or at least its proponents hope.

[Scott]

Seriously, what "clean true sale opinions"? There's no such thing. True-sale opinions are "reasoned" opinions. Anyone who has ever written a true-sale opinion (or read one, for that matter) knows that it's always been impossible for a true-sale opinion to be "clean."

If by clean you mean accepted by the rating agencies, then that's another issue entirely. None of the rating agencies have particularly strict or demanding criteria for acceptable true-sale opinions. In reality, all they require is a reasoned opinion that ultimately concludes that a court would likely find a true sale has taken place. (I've penned a few true-sale opinions in my day.) Since the legal analysis in a true-sale opinion has never really had much bearing on the ultimate conclusion (which is always a foregone conclusion), I don't see why it would start to matter now. A 5% risk-retention requirement certainly isn't enough to make every major US law firm refuse to provide true-sale opinions.

And just intuitively, it's hard to imagine that a mere 5% risk-retention would mandate recharacterization, especially when you consider the myriad forms the risk-retention can take under the proposed rule.

[Adam Levitin]

Fair comments both on the clean point, but there's still a fundamental conflict between risk retention and sales. Perhaps its just the cost of a true sale opinion that will go up.

[Kathy tiss]

Another outstanding post Adam! Your the inspiration to so many hardworking AMERICAN'S! Don't give up we haven't! As I said in my post on the Hamlet

--here's a good piece to start your day w/if you haven't already read. Hugs and many Hugs to this man's hard work for the AMERICAN PEOPLE!

[David Johnson]

"Investors like securitized assets precisely because the assets are (in theory) impervious to the fate of the originator or aggregator. (Obviously that isn't true if the originator/aggregator services or provides maintenance or warranties for the securitized asset.)" Why does it matter whether the originator/aggregator services the asset to make it bankruptcy remote? If it was a true sale, it doesn't belong to the servicer. I get my car serviced by the dealer. If the dealer goes bankrupt while my car is on their lot being serviced, I can get it back by showing my registration. What's the difference here?

[Adam Levitin]

David Johnson--this isn't like the auto dealer going bankrupt while your car is being serviced. An ABS investor doesn't own anything except a security. The assets belong to the securitization vehicle. The concern is that the assets of the securitization vehicle will be substantively consolidated with those of the originator/aggregator. If so, the ABS investor would just have a claim in the originator/aggregator's bankruptcy.

Retained servicing causes further difficulties. The better analogy is not the dealer going bankrupt, but the OEM going bankrupt and liquidating. If Chrysler had liquidated, the value of a Chrysler would have plummeted precipitously, even though Chrysler doesn't own the car, because of servicing, parts, and resale issues.

So consider what happens if there is retained servicing and the originator's finances' slip. First, the originator's credit rating is downgraded. When that happens, the originator's servicing rating is downgraded. And when that happens, the rating of the MBS/ABS is downgraded, so the value of the MBS/ABS declines. At some point, the originator/servicer's financial problems could interfere with its ability to service the loans. PSAs provide for servicing transfers in such cases, but those can be messy (payments might not go to the right place or might not be made at all) and ultimately cost investors.

Joseph Mason and Joshua Rosner have a nice discussion of this problem in a 2007 article they wrote on the ratings of CDOs.

[David Johnson]

Thanks Adam for the further clarification. I really enjoy your postings. I'm concerned here because of the WAMU bankruptcy and the Chase takeover of their servicing facility. I have a case where Chase is claiming ownership and then assignment to Bank America of a mortgage/note which was clearly part of a securitization. Chase claims to have purchased the mortgage from the FDIC under the terms of the FDIC liquidating WAMU. WAMU was the initial servicer and initial custodian of the mortgage pool.

[polaco]

Guys, just relax and enjoy. That is nothing comparing with is coming after july.

[Bob Lawless]

"True sale," "title," "lease," "pink rock candy mountain" . . . it doesn't matter what labels get thrown around. The consequence is what matters -- namely inclusion in the bankruptcy estate. If we think the risk retention rules create a problem in regard to the treatment of the sale in a bankruptcy--and I think they might--then isn't the answer also to change the consequence. There is no reason we could not have a rule requiring risk retention and have a rule that says the asset remains outside the bankruptcy estate.

[Knute Rife]

Seems to me that any time you have a repo or a guarantee that is effectively a repo, you don't have a sale, you have an executory contract for sale, and it ought to get sorted out in the estate under 365.

[how 'bout]

Adam,
Apologies in advance for any failure to comprehend. I understand your concern, but it seems it could be addressed faiurly easily. Why not draft the risk retention rules such that they could be met by direct holding of some of the resulting assets? If the proposed rules says "these rules may be met by originator holding 5% of each resulting securitized tranche," doesn't this problem go away?