Insider's Game

Many is the time I would review a write-up of a new deal and scribble in the margins, “Get to the bleeping point!” Unless you can articulate, up front, exactly what assets we would be lending against, and what circumstances would cause us to lose money (i.e. a quick-and-dirty breakeven analysis), you don’t really know what you’re talking about. And if you don’t have a good grasp of that issue, everything else you have to say is superfluous, a waste of time.

This lack of common sense is pervasive, extending far beyond the financial services industry. (When, over the last seven years, have you ever heard a journalist ask, “How many troops do we have to replace those currently deployed in Iraq?”) In certain markets, most notably, CDOs, this lack of common sense was institutionalized. It’s evident in the deal book for Abacus 2007 AC-1, at the center of the S.E.C.’s case against Goldman.

What risks are investors assuming? The presentation doesn’t say. There’s a reference portfolio of 90 subprime mortgage bonds, on pages 55 and 56, which ostensibly would be insured via credit default swaps for the benefit of Goldman. But, as the small print says,

“Goldman Sachs neither represents nor provides any assurances that the actual Reference Portfolio on the Closing Date or any future date will have the same characteristics as represented above.”

According to my bias, everything else in the 66-page presentation is superfluous. And the real reference portfolio for Abacus 2001 AC-1 remains, to my knowledge at this point in time, hidden from public view.

But if we assume that no one pulled a bait-and-switch, then the evidence of Goldman’s corrupt intent was always hiding in plain sight. Eyeballing the list of 90 subprime reference obligations, I happened to recognize a few that were notorious.

J.P. Morgan Mortgage Acquisition Corp. 2006-FRE1 (JPMAC 2006-FRE1) was a billion-dollar subprime bond that imploded right away. About 13% of its loans were in foreclosure (either in the foreclosure process or as real estate owned, known as REO) as of April 26, 2007, when Abacus 2007 AC-1 closed. Because JPMAC 2006-FRE1 had such a high level of serious delinquencies at that time, no cash flow could be could be applied to any principal repayment for 10 of the 11 tranches that were senior to the BBB tranche, which Goldman shorted. It was obvious that BBB tranche, in the bottom 7% of the capital structure, would default. Goldman wasn’t assuming any kind of risk at all. There was virtual certainty that it would collect on the credit default swap.

The same certainty applied to Argent Securities Trust Series 2006-W1, which had a 10% foreclosure rate, toMorgan Stanley Abs Capital I Inc. Trust 2006-WMC2, which had an 8% foreclosure rate, and to Structured Asset Investment Loan Trust 2006-4, which also had an 8% foreclosure rate. Each of those deals had already tripped up the delinquency trigger in its respective cash flow waterfall structure. In other words, there really wasn’t any doubt that Goldman would collect on its credit default swaps.

These are just a couple of high profile deals that I happened to recognize. They may not be representative of the actual overall portfolio. But common sense tells me all I need to know. This deal was designed to provide a windfall to Goldman at the expense of some unwitting suckers.

Goldman has asserted that the portfolio selection did not matter, that all subprime bonds of that 2006 vintage performed badly. Exactly. As explained here previously, the real estate bubble concealed a multitude of sins. In 2005, people who could not afford their mortgages would still sell their homes and recover some equity. Home flipping schemes were profitable, and promptly paid off the loans. But when home appreciation stopped in 2006, those same types of borrowers, who had lost their equity, were walking away handing over the keys to lenders. Goldman and John Paulson saw the spike in delinquencies and figured out what was going on, which was why they were aggressively shorting those deals.

All this gets back to the myth of the sophisticated investor. The reference portfolio of credit default swaps was static, with no substitutions or reinvestments allowed. So after the deal closed, it didn’t matter whether the investment manager were a subsidiary of ABN Amro or two guys with a Bloomberg terminal.

Pre-closing, the most important question was: How likely is it that Goldman will collect on its swaps? The motivations of ACA Management, the nominal Portfolio Selection Agent, were not all that relevant.

But the deal book clearly demonstrates Goldman’s intent to distract attention away from the underlying substance of the deal. In other Goldman deals, notably Anderson Mezzanine Funding 2007, Abacus 2006-13 and Abacus 2006-17, the opportunities for abusive self-dealing were conspicuous. Abacus 2007 AC-1 is clearly organized to seduce investors with an illusory sense of comfort, that ACA Management, an independent third party and subsidiary of a global bank, would offset Goldman’s motivation to shaft investors. The irony, of course, is that Goldman worked in tandem with John Paulson to minimize any possibility that investors escape unharmed.

Still, from looking at ACA Management’s organization chart and its staff biographies, I never would have expected that they could not have known, as of the CDO’s closing date on April 26, 2007, that Goldman’s windfall was a sure thing. They must have reviewed the current performance reports on 90 different bonds. How could they miss it? Somebody at ACA was afflicted with a pretty big case of willful blindness.