When the SEC dropped its bombshell fraud charges against Goldman Sachs (GS), reporters"”including this one"”rushed to tease out the story of the complaint, a narrative of how a Goldman salesman snookered innocent investors. But as the case has evolved, the storyline has started to look more complicated. In the Daily Beast, for example, John Carney has raised questions about how naive the German investors whom Goldman is accused of bilking really were.
The role of another player in this"”ACA Capital, a Goldman partner and investor which the SEC accuses Goldman of misleading"”may be even more troubling. A longtime player in the relatively sleepy business of insuring municipal bonds, ACA had moved into what seemed to be a more lucrative line of designing and insuring derivatives.
ACA went public in November 2006, at the peak of the mortgage bond bubble and only weeks before its involvement in the deal at the center of the Goldman charges began. Before the end of 2007, barely a year after that public stock offering, it was effectively insolvent and was taken over by regulators.
The SEC charges center on a complex derivative, a CDO cryptically named ABACUS 2007-AC1, which would pay money to investors as long as homeowners kept making payments on mortgages linked to a certain set of high-risk bonds. If these bonds did badly, on the other hand, the deal would yield profits for John Paulson, a fund manager who ultimately made a fortune betting on the collapse of subprime mortgage bonds
ACA was a major loser in the ABACUS deal and is presented in the SEC complaint as being duped by its Goldman contacts. But ACA was also a partner of Goldman's in the ABACUS deal, whose role was to vet it for other investors. The SEC contends that a Goldman salesman, Fabrice Tourre, tricked ACA into believing that Paulson wanted to invest in mortgage bonds"”that he would be "long" on the deal when, in fact, he was betting against ACA and other investors.
To ask questions about the case may appear to some to be an apologia for Goldman. It is not. What is beyond dispute is that the ABACUS deal was a dud among duds, toxic to investors and linked to a portfolio of bonds that turned out to be worthless. If this were a real estate deal, then investors in this CDO were sold land in a distant Florida swamp.
A closer reading of the case, however, yields what may be a more nuanced story than that of the SEC complaint. The SEC charges paint a picture in which ACA, misinformed about Paulson's interests, was led to sign on to a deal that was built to fail"”filled with a "reference portfolio" of bonds that Paulson believed were ready to blow up, and marketed to unwary investors.
The SEC's theory of the case hinges largely on the idea that Goldman misled ACA, essentially its own marketing partner in the ABACUS deal. But a more thorough look at ACA's role reveals a narrative of at the least staggering incompetence, and presents some real questions about the facts and the SEC's strategy:
1. How misguided would ACA have to have been to think that John Paulson was on their side? Unbelievably so. Gregory Zuckerman's book The Greatest Trade Ever calls Paulson's interest in betting against the housing market an "open secret" on Wall Street. But really, it was no secret at all. It was the express and public aim of the fund, Paulson Credit Opportunities fund, which Paulson started in 2006. Check out this issue of a trade paper, Alternative Investment News, which says that Paulson's new "Credit Opportunities" fund would be short on the subprime U.S. housing market.
The idea that Paulson was betting on the bonds in the ABACUS portfolio rather than against them is based on a few mentions of Paulson as an "equity" investor in the CDO"”the buyer of the riskiest slice (or, in bank-speak, "tranche") of the deal, the value of which falls on even a small drop in the mortgage bond market. In fact, the complaint says, Goldman never intended to market that part of the deal to anyone.
One big issue for the SEC, however, will be explaining how, even if ACA ever thought this, it could still think so by the time the deal was actually presented to investors. ACA's own role takes up many pages of the deal presentation"”obtained by Naked Capitalism and posted here. That presentation makes clear the final structure of the deal and implies that the "equity" or "first loss" tranche would be sold. (On Page 14, the presentation notes that there would be no "coupon," or payment, to holders of a "first loss" tranche.) It's hard to understand how ACA could have missed this. 2. Even if ACA referred to Paulson as an "equity investor," does that prove that ACA thought he was on the same side as them? In a word, no. "Structured finance" is infamously complicated, and the SEC complaint presents a highly simplified version. In the SEC version, there are only two sides: the buyers of the securities and those who bet against them. On this view, Paulson must have been either "long," and on the same side as ACA, or "short" and betting against them.
It's clear even from the e-mails picked by the SEC that even if ACA did think Paulson would be an equity investor in the CDO, the folks at ACA saw their interests as potentially opposed to his. "I hope I didn't come across as too antagonistic," one ACA manager wrote to Goldman after one discussion with the Paulson fund, going to explain that while he could "understand Paulson's equity perspective," the deal would be "difficult" from ACA's standpoint.
An important consideration here is that even if Paulson had invested in the equity tranche of the deal, it in no way precludes him also betting against the other tranches by buying a CDS. According to Zuckerman, this is what Paulson did on some of his CDO deals, and he might, at some point, have considered doing it here. This strategy is not unique to Paulson.
Buying the equity"”otherwise known as the "first loss" tranche"”and betting against other parts of the CDO is not a play unique to Paulson. (If you want to get into the weeds and try to understand why it makes sense, you can see that combination of "buying protection" and "retaining the first loss position" on Page 5 of the JP Morgan CDO handbook from all the way back in 2001.) We don't have to look very far for an example: In ProPublica's investigation of the hedge fund Magnetar, which also made a killing betting against CDOs, you can see that this combination is precisely what Magnetar did.
3. If there's a fraud here, is ACA the misled victim or a willing collaborator? ACA was the biggest investor in the ABACUS deal, and when the mortgage bond market collapsed, ACA wound up owing Goldman and Paulson about $841 million. That money was ultimately paid by the investment bank ABN Amro, which had insured ACA's liabilities against the risk that the losses on the CDO were so great that ACA couldn't pay. Other investors, such as the German bank IKB, also saw their stakes in the CDO wiped out.
Just as the relationship between Paulson & Co.'s interests and those of ACA is more complicated than whether both were "long" the deal, the link between the interests of ACA and those of investors like IKB is also more complex than the picture painted in the SEC complaint. ACA was the holder of the "super senior," or safest, slice of the CDO. That meant that losses on the bonds in the reference portfolio would hit other investors before ACA.
In essence, this means that while ACA was betting that mortgage bonds would hold up well enough to avoid losses on their own super senior tranche, ACA could have expected to make money even in a falling bond market. While it was ACA's role to vet the deal for other buyers of the CDO, ACA's interests could very have diverged from theirs from theirs. When defaults are high (but not too high), for instance, the holders of the safest tranches, who are protected against anything but a catastrophic collapse, stand to benefit even as other investors lose. Whether that's a scenario ACA contemplated will probably become a question as the case evolves.
4. Does who chose the reference portfolio in the ABACUS deal really matter? If it does, it's a whole lot less than it seems as first glance. The SEC complaint presents what at first appears to be the amazing fact that by January 29, 2008, 99 percent of the bonds in the ABACUS reference portfolio had been downgraded by ratings agencies. But a little bit of context helps. All the bonds included in the ABACUS portfolio were "mezzanine" mortgage bonds, very sensitive to fluctuations in the housing market and with a rating from the agency Moody's of "Baa2."
Moody's announced in January 2008 that it had downgraded almost two-thirds of all Baa rated subprime mortgage bonds. By April of that year, essentially all lower-rated subprime bonds had been downgraded. So while the bonds chosen for Paulson's CDO were miserable, they were only marginally less so than other, similar subprime bonds. Even a portfolio of the "best" bonds in the category would, in the end, have yielded a similar result. That ACA was willing to "vet" and approve a CDO built out of these bonds at all matters a lot more than exactly which clunkers made it in.
5. Did the SEC miss the forest for the trees? The SEC has focused its case on the narrow question of how much investors knew about how the bonds in the ABACUS portfolio were chosen. But in doing that, the regulators have essentially shied away from a simpler and bigger question: Why was Goldman marketing a CDO like this to investors when people inside Goldman believed that the market for mortgage bonds and CDOs was blowing up?
To get back to a real estate analogy, the SEC's focus on exactly what bonds made it into the ABACUS portfolio may be much like asking what particular pieces of land investors were offered in a swamp. Whether it was John Paulson, ACA, or Goldman who chose the exact plots of land, ultimately they were all located in the same bog.
One point that is made compellingly in the SEC complaint is that as the ABACUS deal was being put together, people inside Goldman believed the CDO business was already dead and a mortgage bond blow-up was near. As Fabrice Tourre put it in an e-mail, "The whole building is collapsing."
In focusing the case on ACA, the SEC has made this case largely about how much investors should have been told about how this complex derivative was constructed. That puts the SEC in the peculiar position of relying largely on the very people who helped Goldman sell the ABACUS deal to prove their case. And worse, it may mean that the case won't answer the really big question here: what responsibility Goldman Sachs had to tell investors that they were buying swampland in the first place.
Read Full Article »