The Fine Print of Goldman’s Subprime Bet
The vampire squid haters won’t like this column.
For the past several weeks, I have been trying to understand if Lloyd C. Blankfein, Goldman Sachs’s chief executive, could have perjured himself — as Senator Carl Levin has suggested — when he testified last year in front of the Senate’s Permanent Subcommittee on Investigations and declared, “We didn’t have a massive short against the housing market.”
Based on the subcommittee’s report, which was referred to the Justice Department, I wrote a column raising questions about Mr. Blankfein’s comments. At the time, his testimony seemed ridiculous in the face of evidence that Mr. Levin presented, which showed that the firm had regularly made large bets against the subprime market.
But upon further reporting — talking with executives at Goldman, who pointed me to other documents, and with officials in Washington, and then poring through the report, following the footnotes to the original sources and then cross-referencing them against other public records — I have come to a different and perhaps unsatisfying conclusion for those readers looking for a big scalp: Mr. Blankfein wasn’t lying.
That’s not to suggest Goldman always behaved well. There are other assertions in the subcommittee’s report that detail some pretty egregious activity by certain executives.
But after comparing the report with publicly available filings and documents, there are enough questions about the accuracy of certain parts of the Senate report to raise some red flags.
Take this example: The report unequivocally states that in 2007, Goldman “reported net revenues of $11.6 billion, of which $3.7 billion was generated by the structured products group in the mortgage department, primarily as a result of its subprime investment activities.”
The sentence was meant to show that Goldman’s shorting of the housing market had provided a large percentage of the firm’s revenue that year. The sentence in the report even included a handy citation for the information.
But in 2007, Goldman Sachs reported net revenue of $45.98 billion, not $11.6 billion. That’s a big difference.
When I spoke to Robert L. Roach, a counsel and chief investigator for the Senate subcommittee who helped draft the report, he said, “We made a mistake.” He said it was a “typo” and insisted that “we weren’t trying to cook anything.”
The figure was Goldman’s net earnings, not its net revenue. He pointed out that later in the report — 93 pages later in a different context — the report accurately described Goldman’s net revenue as $45.98 billion.
Mr. Roach protested that Goldman had never brought the mistake to his attention in the two months since the report was published. There has clearly been much antagonism between Goldman and the various bodies investigating the firm, even as Goldman says it is cooperating. Staff members of the Financial Crisis Inquiry Commission complained when Goldman dumped hundreds of thousands of documents on it; Goldman said privately that the commission was engaging in political theater.
Yet there are other sections in the Senate report that appear to go beyond sloppiness. The report says that Goldman’s structured products group made $3.7 billion, mostly by going short in 2007. That is correct. But the report omitted the total net revenue for the mortgage department, which included structured products.
According to a document Goldman Sachs provided to the subcommittee and made public on its Web site a year ago, Goldman had “less than $500 million of net revenue from residential mortgage-related products — approximately 1 percent of the firm’s overall net revenues.”
In other words, while one part of the department had gone short, another part had gone long.
So when Mr. Blankfein contended that the firm was “not consistently or significantly net ‘short the market’ in residential mortgage-related products in 2007 and 2008” the numbers — if you believe them — are on his side.
Mr. Roach said he was unaware of where the figure of less than $500 million came from, although in the document that Goldman provided the subcommittee, a series of bar charts broke down its exposure for every quarter of 2007 and 2008. Mr. Roach pointed to a different document that showed the firm had made $1.13 billion.
This isn’t meant to say that part of the firm didn’t go short — it did and the firm has repeatedly said so. But the suggestion that the short was a huge directional bet by the firm to profit off a real estate collapse may not completely stand up.
One document the subcommittee cited as evidence that Goldman had been “massively short” was a presentation by Josh Birnbaum. Mr. Birnbaum, a Goldman trader who ran the structured products group, tried to persuade his bosses that he and his group were deserving of a bigger bonus because of their successful short positions. To make his point, Mr. Birnbaum said that “the shorts were not a hedge,” a quote that Mr. Levin’s report brandished as proof that Goldman was lying about its short position.
But it’s hard to give much weight to the quote. Left out of the subcommittee’s report was the answer Mr. Birnbaum gave under oath during his testimony. When asked whether the shorts had been hedged, Mr. Birnbaum said, “I was not aware of what the firm as a whole was — what the firm’s position on mortgages was.”
After his testimony, when he realized that the government had him saying “the shorts were not a hedge” in a presentation, he doubled back. He tried to suggest in written testimony that he believed what he had originally written in his presentation but that his other testimony had also been accurate — a hard circle to square.
More important, if Goldman made only $500 million in net revenue from its residential mortgages when Mr. Birnbaum’s unit made $3.7 billion from shorts, it is clear that it also had huge long positions. Had it been “massively short,” the firm should have made much more than $500 million.
Mr. Levin’s office referred questions about the report to the permanent subcommittee.
The senator also made use of Goldman’s “top sheets” as evidence the firm was short the residential housing market. The report says that a top sheet from June 25, 2007, shows that the firm was short $13.9 billion. That would be quite a big short position.
But in studying the document, the subcommittee may have mixed apples and oranges. It added in $4.1 billion worth of short positions for commercial real estate to residential real estate. And the subcommittee ignored the footnote on the bottom of the document that the “top sheet” had not included long positions in other parts of the business that people close to the firm said were in excess of $5 billion.
Subtract those positions, and Goldman had a net short position of less than $5 billion in residential mortgages — not nearly triple that.
Goldman haters will think that this column is an apology or spin for the firm. That wasn’t the point. Of course, some of what Goldman and others did ahead of the financial crisis is deeply troubling — and possibly even illegal — and they should be punished if a crime was committed. But on this particular score — about Mr. Blankfein’s testimony — the evidence is far from convincing.
As even Mr. Roach acknowledged, “It’s about how you define ‘massive’ and ‘large’ — and I’m not trying to be cute.”