Today the Commission announced a new rulemaking on ABS representations and warranties. The Headless Bouncer begins to wonder if maybe ABS is going to be just another liquidity facility, after all.
You can’t make a silk purse out of a sow’s ear. Unless, that is, you’re a team of talented chemists working on the boss’s pet project.
What are the sow’s ears that the wizards of analytics had to work with back in the day? The Headless Bouncer proposes to pick on a Lehman deal, SASCO LMT 2007-1, since they aren’t around to get their feelings hurt. What, precisely, did the modelers have to work from?
Would it have been the Braindead Pro Supp filed on January 4? Hmm, THB doesn’t think that
Interest will accrue on each class of offered certificates (other than the Class AP Certificates or any other principal-only class of certificates) at specified annual rates.
provides quite enough to go on. Of course the related registration statement didn’t fill in the blanks either.
How about theTerm Sheet filed on January 31? Hard to believe that there’s enough in the 12 pages to fully limn out all the paydown rules, but THB guesses it’s possible. What to do about all the blanks and TBDs, however, is another problem.
Good news comes with the collateral term sheet filed also on January 31. Here’s some actual data. Unfortunately, we don’t know which loan group any particular loan belongs to.
A final term sheet with a bunch of new classes and punting some of the calculations to the pro supp comes next.
Next the book itself with gobs more detail. For example,
Mortgage Loans in Pool 1 with Net Mortgage Rates Greater than or Equal to 7.00% and Less than or Equal to 7.50%. Principal received or advanced on each Mortgage Loan in Pool 1 having a Net Mortgage Rate greater than or equal to 7.00% but less than or equal to 7.50% will be allocated between Collateral Group 2A and Collateral Group 3A on the basis of the Applicable Fractions of the related Mortgage Loans. The Applicable Fraction for the portion of each such Mortgage Loan allocated to Collateral Group 2A, is (7.50% minus Net Mortgage Rate) / 0.50%, and the Applicable Fraction for the portion of each such Mortgage Loan allocated to Collateral Group 3A, is 1 minus ((7.50% minus Net Mortgage Rate) / 0.50%).
That’s kind of interesting. The book is dated January 26.
“Quick,” THB shouts to his imaginary admin, “get me Howard Baker!”
“Howard? Senator? How are you, Sir? Good. Yes, thanks for asking. No, I don’t have migraines anymore. I hate to bother you on a simple question like this, but what was it you kept saying during the Watergate hearings. Yes? What did they know … ? Yes, … and when did they know it? Yes, yes, that’s it. Thank you so much. I can’t tell you how much … Off the record? I can’t quote you on that? Absolutely, sir. Scout’s honor. No, sir, I wasn’t actually a scout. They don’t? If I reveal my source you’ll rip my head off? Well, in that case, goodbye and thanks, again.”
Now THB strongly suspects that the done! in this deal happened well before the book was available, didn’t it? And a deal like this was built up in pieces, wasn’t it? Maybe right up until the last piece, eh? (Those goofy Canadians, gotta love ‘em.) So, puzzle me this: what did the commercial model outputs actually represent?
The Headless Bouncer picks up his air bass and strums a few licks, talking ’bout my generation. Yes, THB is a Boomer, and proud of it. (Of course, he would be proud if he were a Greatest, or an Xer or any of the other younger generations.)
Now, you may say that the Boomers have produced some outstanding figures in the arts and sciences and even political life but as a generation we just haven’t lived up to our potential. You can’t say we weren’t warned. Although we made our rep on railing against the dead hand of the past, the timidness born of the Great Depression and the refusal of our parents to believe in limitless futures, we are now mostly over 50.
None of us grew up with PCs. Few of us in the securitization biz studied computer science, math beyond high school slide rule or ever wrote a program for fun. We think that this stuff is supposed to be hard.
Our children and grandchildren mostly have a different view. All except the oldest grew up with PCs and learned to use them before they knew it was supposed to be hard. (THB was, for that reason, vastly amused by the idea of a parental control chip for TVs; who, do you think, would program it?)
Let’s face it, boys and girls, the old folk ain’t gonna learn to read Python or any other programming language. And they ain’t gonna be the ones staying up late nights checking that the logic of the program matches the book and the PSA. Are they, now? No. Young people are going to be doing that. People still educable, at least about this and for a while before they, too, ossify. They’re the ones who have to be able to deal with it. It does not signify that THB’s peers can’t.
The power of not knowing that something is supposed to be impossible is encapsulated in the old saw “give it to a grad student who doesn’t know enough about the problem to be deterred” or as George Danztig related
During my first year at Berkeley I arrived late one day to one of Neyman’s classes. On the blackboard were two problems which I assumed had been assigned for homework. I copied them down. A few days later I apologized to Neyman for taking so long to do the homework – the problems seemed to be a little harder to do than usual. I asked him if he still wanted the work. He told me to throw it on his desk. I did so reluctantly because his desk was covered with such a heap of papers that I feared my homework would be lost there forever.
About six weeks later, one Sunday morning about eight o’clock, Anne and I were awakened by someone banging on our front door. It was Neyman. He rushed in with papers in hand, all excited: “I’ve just written an introduction to one of your papers. Read it so I can send it out right away for publication.” For a minute I had no idea what he was talking about. To make a long story short, the problems on the blackboard which I had solved thinking they were homework were in fact two famous unsolved problems in statistics. That was the first inkling I had that there was anything special about them.
Listen up, Clubbers, The Headless Bouncer is going to cut through the κακα muddles being made of the waterfall computer program.
THB is going to set aside the whole question of how the program is to be implemented. Let it be written in Pangloss, the best of all possible computer languages. The real confusion is in just what it is that the program is supposed to do.
Further, THB will accept the SEC’s conclusion that the existing state of the art for ABS disclosure is unsatisfactory in a way that requires a different disclosure regime, based on open computer models. The reason that current disclosure is unsatisfactory is that it is abstract.
The first cause of the abstraction is that disclosure revolves about a contract between a scarecrow and a cowardly lion. The scarecrow is the trustee and the cowardly lion is the issuer. The scarecrow has no responsibility for the cashflow structure, and could care less, other than following any express direction it is given in the transaction documents. The cowardly lion’s only interest in the cashflow structure is making sure that however unwieldy it may be it has been exhaustively described, qualified, circumscribed and reified to minimize disclosure liability. Neither party has any stake in or curiosity concerning the economic substance behind the bargain struck between the dealer and the buyer that the disclosure purports to describe.
Because the real parties-in-interest, the dealer and buyer, barely exist for purposes of the transaction documents and their purposes and motivations in structuring the transaction are hidden from view, it is unsurprising that the disclosure is more abstract than the language in which the dealer and buyer bargain for the design of the structure. That language differs from the disclosure in two important ways: it involves specifics based on pool asset characteristics and it is expressed primarily as a model with numeric inputs and cash outputs.
That dialog does not proceed along the lines of an algebraic description of the structure. The investor is not buying a formula, he is buying a projected cash flow and wants the actual cash received to depend on as few assumptions as possible. He is buying a result based on stated conditions. The basis of his bargain is “I give you money and I get this cash flow from this asset pool if a, b and c happen and d, e and f don’t happen.”
To discover that bargain in the disclosure or operative documents is not easy. Their abstraction provides both too little and too much. Missing is any way of connecting the facts of the collateral to the logic of the paydown rules. Present are the rules the apply to every other class of security in the deal, some of which directly affect a security of interest, some of which may indirectly affect it and others of which operate purely independently, depending on the structure. An investor seeking to confirm the bargain that he thought was struck has not only to understand the instrument that he bought, he has to understand every other instrument as well, at least well enough to assess its influence on his holding.
The defects of a disclosure that is based on a fictionalized meeting of the minds over a rarified version of a structure that is actually hammered out by others out of hand crafted material are obvious in reading the bloodless result. Never has so much steak been sold with no sizzle included. The prospectus as we have known it is nothing better than a glorified tombstone erected to memorialize a deal done elsewhere by others who didn’t depend on it to understand what went down.
That the ABS market worked at all is a tribute to the providers of independent analytic services who offer to do the heavy lifting of melding the collateral characteristics with the structure and producing alternative futures limited only by the number of dials that an investor cares to tweak. Bless ‘em all, THB says, but industry standard proprietary solutions are not the answer, for two reasons.
First, they are secret. Not only is there no way of knowing if their output is numerically correct, given the inputs, there is also no way of knowing if they are even wrong consistently on the same deal for different clients. Now, you can make a good argument that given their role as necessary enablers the market has validated their results by virtue of trading upon them. But that, as THB’s litigator buddies like to say, assumes facts not in evidence. Let’s just hang the “lack of transparency” label on this part of the problem.
Second, the issuer is not accountable. You say you bought $400 million, bunkie, on the strength of that model run? Didn’t turn out well for you? Can’t put ‘em back because of that damned fine print in the book? Vendor blew it? Your lawyer tells you that you signed up for a no consequential damages clause? They’re only offering you a $50K credit on next year’s subscription as their way of saying “sorry, dude”? Don’t know how you’re going to explain this one to the old committee?
What to do?
The disclosure must integrate the collateral description with the structural cash distribution rules in a way that makes explicit their relationship in a concrete form. If you start with these loans under these rules, your security gets this much in this period assuming these facts that cannot be predicted with certainty.
The joke to this punchline is that this is something that we already do and have done for a long time. It’s called a decrement table. It uses rep lines, a bunch of unrealistic assumptions and yields up a rough measure of duration. The only thing missing is the lifting of the kilt to show the construction.
To buff up the dec table to what the SEC is calling for requires no more than increasing the number of rep lines to equal the number of pool assets (again, THB is speaking from his RMBS bias) and showing the homework. You take into account everything that is knowable from the face of the note (balance, interest rate, duration, etc.), provide dummy variables for the unknowable (future interest rates for ARMS, probability of prepayment, default, severity given default) for the investor to play with and generate a pro forma report. The fact that you have to translate the defined terms into variables and formulas that you show to investors only means that as an issuer you have to take care that you really understand the dealer’s structure. Why that should be a problem is not really at all clear.
THB will close this rant with the report of a howler heard today.
What about re-REMICs? There might be a couple of hundred underlying bonds, each with its own pool and own paydown rules. Surely, that’s too hard!.
Who prices these? Madame Rue?
The Business Law Section of the American Bar Association’s 117-page comment letter has been filed. Rather than attempting a summary here, The Headless Bouncer refers you to the helpful table of contents.
The Headless Bouncer has read much in the comments about the virtues and importance of the purveyors of ABS analytics and their vital role in the securitization market. Variously, they render the issuer-provided waterfall computer model superfluous, are superior to any possible issuer-provide model, state-of-the-art, reliable, trustworthy and the general, all-around way to go.
Mindful of the experience of undue reliance on the oracular opinions of rating agencies, THB wonders whether there is an appropriate regulatory regime that would provide a degree of oversight of the activities of the ilk of Intex. One of the many joys of the securities law is that there’s always a treasure trove of unexplored possibilities that await new application.
One jewel is §202(a)(11) of the ever-popular Investment Advisers Act of 1940 that defines an investment adviser as including any person:
who, for compensation and as part of a regular business, issues or promulgates analyses or reports concerning securities
Compensation. Check. Regular business. Check. Analyses or reports concerning securities. Check.
Exempt?
Not a bank. Check. Services not incidental to a profession. Check. Not a broker-dealer. Check. Not a publisher of a financial publication of general and regular circulation. Probably check. Services do not relate solely to exempt securities under Section 3(a)(12) of the Exchange Act. Check. Not exempted by Commission order. THB is not sure on this one, but notes that what the Commission hath done under its discretionary authority it may undo.
The Headless Bouncer has been raising his voice in protest of the traditional form of notwithstanding the third sentence of the second preceding paragraph disclosures of ABS structures and in support of the SEC’s proposed waterfall computer program.
Today’s mail brought a counterexample of narrative description of a complex financial instrument that, had it been the norm, might have made both the protest and the support unnecessary.
Dr. Dapeng Hu and Dr. Kishore Yalamanchili of Blackrock describe the PrimeX indices in the current issue of American Securitization, available online (members only) in lucid terms such as
Price of Funded Index = PV(premium writedown) = PV(premium) + PV(cash bond) – PV(guaranteed bond)
THB earnestly recommends that if you are not an institutional member of ASF (shame on you), you should give your friends at Blackrock a shout for a reprint.
“This is a bad thing?” asks The Headless Bouncer?
“A load of codswallop,” snorts The Headless Bouncer in his best working class English accent. “CDO.”
Parting words
The Headless Bouncer is heading back to another project while the Commission chews on the comments and decides whether to proceed with its various proposals, including the waterfall computer program. That one seems like a no-brainer, but THB is likely biased, since for him everything by definition is a no-brainer.
