Goldman’s Abacus lies
The SEC suit against Goldman Sachs (full complaint here, and well worth reading) is explosive stuff. Essentially the SEC seems to have nailed down the kind of behavior that ProPublica was looking for in its story on the Magnetar Trade — a hedge fund which was short mortgages, in this case Paulson, was carefully picking nuclear waste to put into synthetic CDOs, unbeknownst to the final investors in those deals.
From the complaint:
GS&Co marketing materials for ABACUS 2007-AC1 – including the term sheet, flip book and offering memorandum for the CDO – all represented that the reference portfolio of RMBS underlying the CDO was selected by ACA Management LLC (“ACA”), a third-party with experience analyzing credit risk in RMBS. Undisclosed in the marketing materials and unbeknownst to investors, a large hedge fund, Paulson & Co. Inc. (“Paulson”), with economic interests directly adverse to investors in the ABACUS 2007-AC1 CDO, played a significant role in the portfolio selection process.
It seems here that ACA was somewhere between a useful idiot and an outright victim of Goldman’s Fabrice Tourre:
Tourre also misled ACA into believing that Paulson invested approximately $200 million in the equity of ABACUS 2007-AC1 (a long position) and, accordingly, that Paulson’s interests in the collateral section process were aligned with ACA’s when in reality Paulson’s interests were sharply conflicting.
Essentially what Goldman was doing here was using the respected ACA brand name (it wouldn’t remain respected for long) in order to attract big investors like Germany’s IKB: they even said in an email that “we expect to leverage ACA’s credibility and franchise to help distribute this Transaction.” But ACA was to a large extent a front for Paulson, as is evidenced in the name of the spreadsheet where it listed the proposed contents of the CDO:
On January 22, 2007, ACA sent an email to Tourre and others at GS&Co with the subject line, “Paulson Portfolio 1-22-10.xls.” The text of the email began, “Attached please find a worksheet with 86 sub-prime mortgage positions that we would recommend taking exposure to synthetically…
On February 5, 2007, Paulson sent an email to ACA, with a copy to Tourre, deleting eight RMBS recommended by ACA, leaving the rest, and stating that Tourre agreed that 92 bonds were a sufficient portfolio.
On February 5, 2007, an internal ACA email asked, “Attached is the revised portfolio that Paulson would like us to commit to – all names are at the Baa2 level. The final portfolio will have between 80 and these 92 names. Are ‘we’ ok to say yes on this portfolio?” The response was, “Looks good to me.”
I think that the SEC has the right defendant here. Paulson and ACA are both culpable, but it’s Goldman which was clearly central to the plan of deceiving investors into believing that the CDO was being managed by people who wanted it to make money, when in fact it was being structured by the biggest short-seller in the entire subprime market.
And although ACA should never have been so passive in terms of accepting the names given to it by Paulson, it did reasonably believe, because it was essentially lied to by Goldman Sachs, that Paulson was in the deal to make money on the long side:
On January 10, 2007, Tourre emailed ACA a “Transaction Summary” that included a description of Paulson as the “Transaction Sponsor” and referenced a “Contemplated Capital Structure” with a “% – %: pre-committed first loss” as part of the Paulson deal structure. The description of this % – % tranche at the bottom of the capital structure was consistent with the description of an equity tranche and ACA reasonably believed it to be a reference to the equity tranche. In fact, GS&Co never intended to market to anyone a “% – %” first loss equity tranche in this transaction…
On February 12, 2007, ACA’s Commitments Committee approved the firm’s participation in ABACUS as portfolio selection agent. The written approval memorandum described Paulson’s role as follows: “the hedge fund equity investor wanted to invest in the 0- 9% tranche of a static mezzanine ABS CDO backed 100% by subprime residential mortgage securities.”
The really crazy thing about this deal is that ACA not only managed the CDO, but also wrote the credit protection underlying most of it:
On or about May 31, 2007, ACA Capital sold protection or “wrapped” the $909 million super senior tranche of ABACUS 2007-AC1, meaning that it assumed the credit risk associated with that portion of the capital structure…
ACA Capital was unaware of Paulson’s short position in the transaction. It is unlikely that ACA Capital would have written protection on the super senior tranche if it had known that Paulson, which played an influential role in selecting the reference portfolio, had taken a significant short position instead of a long equity stake.
In a sense, then, IKB and the other investors in the deal were right to think that ACA was genuinely managing this structure to make money, and believed that it was sound. After all, if it all went to zero — as it eventually did — ACA stood to lose hundreds of millions of dollars.
But what happened here was that both IKB and ACA Capital placed their trust in ACA Management. And Goldman, armed with hefty CDO management fees and sleazy lies about Paulson’s role in the transaction, turned ACA Management from a bona fide fund manager into a useful idiot who could be relied upon to buy exactly the subprime securities that Paulson wanted to short.
With this suit, the SEC has finally uncovered the real scandal behind the Abacus deals. The NYT tried, back in December, but it didn’t quite get to the nub of the story — although Paulson was mentioned in the NYT story as someone who was generally short the subprime market, there was no indication that he played any role in structuring the deals. Neither was there any mention of ACA.
The scandal here is not that Goldman was short the subprime market at the same time as marketing the Abacus deal. The scandal is that Goldman sold the contents of Abacus as being handpicked by managers at ACA when in fact it was handpicked by Paulson; and that it told ACA that Paulson had a long position in the deal when in fact he was entirely short.
Goldman Sachs has lost more than $10 billion in market capitalization today, in the wake of these revelations. Good. It can go long markets and it can go short markets. But it can’t lie to its clients. That’s well beyond the pale.
Update: The Abacus pitch book can be seen here.