Starting in late 2006, Goldman Sachs made trades that would pay off if the housing market tanked. Was this a massive bet that the housing market was going to crash, as Goldman's critics maintain? Or was it merely a hedge, an attempt by the firm to reduce its risk, as Goldman claims? This debate, which has raged for more than three years, has lately taken on a new significance. Sen. Carl Levin, D-Mich., stops just short of accusing Goldman's top executives of perjury for describing the position as a hedge in their testimony before the Permanent Subcommittee on Investigations, which he chairs. Goldman, for its part, says its executives were telling the truth.
To my mind, the perjury question is a distraction. Any claim that Goldman lied to Levin's subcommittee will probably be hard to prove, and, anyway, the subject obscures more fundamental questions about Goldman's behavior. The evidence that Goldman misled the government is murky. The evidence that Goldman misled its customers, on the other hand, is fairly compelling. And if that is indeed the case, then there are suitable remedies.
It was just over a year ago that top Goldman executives appeared before Levin's subcommittee. Plenty of people were already outraged by the fact that Goldman had established its short position while continuing to sell securities backed by subprime mortgages to its clients. At the hearing, CEO Lloyd Blankfein and CFO David Viniar both said that the short position was simply an attempt to reduce risk. Goldman "didn't have a massive short against the housing market, and we certainly didn't bet against our clients," said Blankfein. "Rather, we believe that we managed our risk as our shareholders and our regulators would expect." Vinair said: "We were primarily, although not consistently, short, and it was not a large short."
Yet emails the subcommittee released made it clear that individual traders at Goldman were quite bearish on the housing market and were seeking to profit from a downturn. In January 2007, a trader named Jonathan Egol wrote: "the mkt is dead." Two other Goldman traders, Michael Swenson and Joshua Birnbaum, bragged in performance reviews about the "extraordinary profits" they'd made and about a plan to not just "get flat, but get VERY short." In late 2007, Birnbaum wrote in a draft presentation (which argued that the mortgage traders should be compensated even more richly than Goldman traders usually are) that "the shorts were not a hedge." Yet another Goldman trader described the position as an "enormous directional short."
Then, this past April, the Levin subcommittee released a 639-page report alleging that Goldman's Structured Products Group, which was a trading desk within the larger mortgage department, "twice amassed and profited from large net short positions in mortgage related securities." (The quotation is from an accompanying press release.) The subcommittee says that in 2007 Goldman Structured Products Group produced more than $3.7 billion in revenues. At a press conference, Levin said, "Goldman clearly misled their clients and they misled Congress" in their 2010 testimony. He said that he wanted federal prosecutors to review whether to bring perjury charges against Goldman executives.
Over the last few months, the drumbeat has gotten steadily louder. In early May, Attorney General Eric Holder told the House Judiciary Committee that the Justice Department was looking at the sections of the subcommittee report that dealt with Goldman. In mid-May, Levin told the Financial Times that there was "real hope" that law enforcement authorities would act on the report. (He said he was "not going to judge whether they [Goldman executives] committed perjury," but they "obviously spent a lot of time parsing words.") Long-time financial services analyst Dick Bove wrote in a report that "pressure on the Justice department to bring a criminal lawsuit against Goldman [appeared to be] building to a high pitch." And the Manhattan District Attorney has subpoenaed Goldman.
Goldman stuck with its previous position, saying that the firm didn't have a "massive net short position because our short positions were largely offset by our long positions, and our financial results clearly demonstrate this point." Apart from that, the firm stayed quiet. Sources tell me the bet was that this would all blow over. When it didn't, Goldman launched a counteroffensive. A Wall Street Journal story cited "people familiar with the situation" saying that the subcommittee had "drastically overstated" Goldman's bet due to "sloppy" math. A New York Times piece pointed out a mistake in the subcommittee report that made the mortgage department's bet look far more important to Goldman's overall results than it really was.
The subcommittee can point to lots of evidence to bolster its position that Goldman was making a big bet against the housing market. But Goldman has one obvious defense: Its mortgage department didn't net large gains because Goldman took losses on the billions in mortgage-related products that it held in inventory. (Just as a clothing store has to have an inventory of merchandise to sell its clients, so does a securities firm carry inventory.) The losses on that inventory, the value of which declined precipitously as the crisis began, mostly offset the wins from the negative bet.
In other words, what was indeed a big short from the perspective of individual traders, or even one desk within Goldman, was more of a wash from the perspective of the entire firm. Because of the losses, Goldman contends that it had less than $500 million of net revenue from residential mortgage related products in 2007—about 1 percent of the firm's total net revenue of $46 billion. (The subcommittee argues that factoring in the losses gets you to a profit of $1.1 billion for the mortgage department in 2007. I'm told that the subcommittee included commercial real estate in its analysis.)
A skeptic would argue that Goldman is now lumping those losses in with its big wins in order to bolster its defense. But Goldman defined its mortgage-related revenues this way even before there was any controversy. On the firm's conference call for investors in the fall of 2007, Viniar said about the mortgage business, "We took significant markdowns on our long inventory positions during the quarter as we had in the previous two quarters. However, our risk bias from that market was to be short, and that net short position was profitable." Profitable, yes, under either Goldman or the subcommittee's calculation. Can Goldman plausibly argue that $500 million, or even $1.1 billion, in revenues isn't "massive" or "large" in a year when the firm's total profits were more than 40 times as great? I think it can.
The profit Goldman realized from its mortgage department isn't the real issue. The real issue is whether the firm reduced its exposure—sold off "cats and dogs," as Blankfein said in a February 2007 email—in a manner that was legal and ethical. The firm already paid $550 million last summer to settle charges brought by the Securities and Exchange Commission over the sale of a mortgage-based product called Abacus. The subcommittee's investigation lays out other instances of unsavory practices, including a draft presentation in which Goldman says it expects some securities it is marketing "to underperform."
Another, even more promising, topic for prosecutorial inquiry is how on earth Goldman and other Wall Street banks could truthfully have professed to customers that they had performed "due diligence" on the mortgages they bought (meaning they'd investigated the quality of the underlying mortgages before they packaged them up and sold them off). How could that possibly be true? There's talk that the SEC is actively investigating this.
It's possible Goldman's actions were legal, but only in a narrow sense of the word. Morgan Stanley took the other side of a Goldman-arranged deal in which Goldman would benefit if the underlying mortgages tanked. From the Senate report, it appears that Goldman used (and abused) the powers it had as the arranger of the deal to make the outcome as favorable as possible to Goldman. In another deal, Goldman sold securities to customers, including an Australian hedge fund and a Korean life-insurance company. A Goldman executive later described that particular offering as "one shitty deal."
Goldman's argument, which might be technically correct, is that such chicanery is not just OK but expected, because the buyers were "large financial institutions, insurance companies and hedge funds with a focus on this type of product" and that "these investors had access to highly detailed information that allowed them to conduct their own independent research and analysis." Translation: If you were dumb enough to buy what we sold, or not to see that we'd work the fine print to our advantage, that's your problem.
A corollary to such market-based analysis is that if customers do find Goldman's practices "troubling and abusive," as the subcommittee described them, then they would be well advised to stop doing business with Goldman, just as you'd stop doing business with an appliance store that won't take responsibility for selling you a defective refrigerator. Which would spell the end of Goldman more fittingly than a debatable perjury case. Convincing customers that they have a fair shot explains why Goldman completed a review of its business practices earlier this year. Will that suffice? Lloyd Blankfein, take note. The market giveth and the market taketh away.