Money and Power

by William D. Cohan

  Another part of Birnbaum’s hedging strategy had nothing to do with mortgages, or the ABX index, or credit-default swaps. Rather, it was a simple bet that the equity of the firms most heavily involved in the mortgage sector would fall. He made these bets by buying put options, whereby he paid a premium to a third party who was willing to take the opposite side of the trade. Birnbaum was betting the stocks would fall in price by a given date, and the seller of the put option was betting the opposite—that the price of the stocks would go up. According to a July 24 e-mail from Birnbaum, his put options had made a profit of $49 million since he had bought them. Among those companies whose stock he bet would fall were Bear Stearns, Moody’s, Washington Mutual, Capital One Financial, and National City.

  It is not clear from the note when Birnbaum started buying the puts, but it is clear that he began sometime before June 21, since that was the date, he wrote, that his group “paused” in “our equities trading while we worked with management and market risk to come up with quantitative limits for these positions.” He wrote that he thought “we are getting close” to an agreement on the limits but in the meantime he wanted approval “to opportunistically buy puts” on those companies with exposure to the mortgage market. He cited specifically thirteen companies he wanted to buy puts for, including Bear Stearns, Lehman Brothers, Merrill Lynch, Morgan Stanley, and Countrywide. Donald Mullen, then head of Goldman’s U.S. credit sales and trading, having joined Goldman from Bear Stearns in 2001, wrote to Sparks the day after receiving Birnbaum’s memo with a sharp rebuke: “He is too large [redacted]. Bruce [Petersen, another Goldman managing director] is going to discuss w[ith] him today.”

  ——

  ON AUGUST 9, evidence of the international spread of America’s subprime crisis showed up in Paris when BNP Paribas, France’s largest bank, blocked withdrawals from three investment funds, which had about $2 billion in assets on August 7, because the bank could no longer “fairly” value them due to a “complete evaporation of liquidity in certain market segments of the U.S. securitization market.” BNP’s action followed an August 3 announcement by Union Investment Management, Germany’s third-largest mutual fund manager, that it had stopped permitting withdrawals from one of its funds after investors pulled out 10 percent of the fund’s assets. Also on August 9, the European Central Bank injected £95 billion into the overnight lending market “in an unprecedented response to a sudden demand for cash from banks roiled by the subprime crisis,” Bloomberg reported, and more than the central bank had lent after the September 11 attacks.

  Hank Paulson, who had been treasury secretary for a year, had been worried about just such a “crisis in the financial markets” since he took his post. He kept his weekly breakfast appointment that day with Ben Bernanke, the chairman of the Federal Reserve, and managed to gobble up his usual bowl of oatmeal, orange juice, ice water, and Diet Coke. “Ben shared my concerns with the developments in Europe,” Paulson wrote later in his memoir, On the Brink.

  When he got back to his office he spoke with Wall Street CEOs, including Blankfein, Richard Fuld at Lehman, Stephen Schwarzman at Blackstone Group, and Stanley O’Neal at Merrill Lynch. “All these CEOs were on edge,” Paulson wrote. O’Neal, for one, remembered that call with Paulson. “If you had called me a couple of days ago I would have been more sanguine,” he told Paulson. “I’m not anymore.” Paulson asked why. “Because you had overnight secured lending fail between rated banks,” he told Paulson. “There’s something more going on, and it means there are potential risks [to the system, beyond] what we think we see on the surface.” O’Neal also knew that Merrill had tens of billions of dollars of CDOs marked at or near par, a ticking time bomb.

  On August 17, the Federal Reserve began to take its first steps to try to stanch the bleeding. The central bank cut interest rates by 50 basis points in recognition that “financial market conditions have deteriorated, and tighter credit conditions and increased uncertainty have the potential to restrain economic growth going forward.” The Fed pledged to “act as needed to mitigate the adverse effects on the economy arising from the disruptions in financial markets.” The Fed also announced that banks could borrow from the discount window “for as long as 30 days, renewable by the borrower,” in order for banks to have “greater assurance about the cost and availability of funding.” The new plan would remain in effect “until the Federal Reserve determines that market liquidity has improved materially.” The two-pronged approach of lowering interest rates and effectively substituting the Fed’s balance sheet for the balance sheets of the country’s financial institutions, whether troubled or not, arose from a Fed offsite in Jackson Hole, Wyoming, during the third week of August 2007. New York Fed president Tim Geithner dubbed this new approach to the growing crisis “the Bernanke Doctrine.”

  As the contagion of the emerging credit crisis began being felt across the globe in the late summer of 2007, Goldman continued to rake in the profits from Birnbaum’s hedges. “Department-wide P&L for the week was $375mm,” Sparks wrote to Montag on July 29, and then added that the trading “P&L on the week was $234mm, with CMBS, CDOs and RMBS/ABX shorts all contributing.” Two days later, Montag updated Blankfein—in a mostly incomprehensible e-mail—on the profits and the market, as well as the firm’s ongoing efforts to cover short positions and reduce the VAR associated with Birnbaum’s hedging. In a presentation that Sparks prepared for Montag to give to Goldman’s Management Committee on August 6, Swenson and Birnbaum reported that it was “a phenomenal week for covering our Index shorts” with one desk buying “$3.3 [billion] of ABX index across various vintages and ratings over the past week,” with $1.5 billion being used to “cover shorts.”

  By the following week, though, the VAR police were back on the prowl. One of them pointed out in a widely circulated internal e-mail that Birnbaum’s trading group’s VAR seemed to be around $100 million, well above its $35 million limit. “[A]re you getting any more heat to cut/cover risk?” Birnbaum wrote to Deeb Salem, on August 9. Birnbaum wrote that he had asked about the VAR police only because he saw the “note about mortgages dropping back down to a permanent limit of [$]35mm (which we are way over). [T]his would mark a change of their recent policy to just keep increasing ou[r] limit. [M]akes me a little nervous that we may be told to do something stupid.” Salem quickly understood Birnbaum’s point. “[I] do think that is a real concern,” he replied. “[H]ow quickly can you work with [the VAR police] to get them to revise our VAR to a more realistic number?” Birnbaum replied that he had a meeting with them on Tuesday, where apparently he was able to get the VAR limit of $110 million extended until August 21. But, on August 13, when VAR for trading overall had increased to $159 million, from $150 million, Viniar was explicit. “No comment necessary,” he wrote. “Get it down.” Gary Cohn echoed Viniar’s comment two days later, after the trading VAR had increased to $165 million. “There is no room for debate,” he wrote. “We must get down now.”

  The concern about the rising VAR on the mortgage trading desk revealed a larger debate then percolating around Goldman: how to take advantage of the misery being felt by other firms as the mortgage markets started to collapse. The problem was that Birnbaum and company continued to see huge profit opportunities to buy when others were forced sellers, but this required putting more capital at risk, which increased the VAR and upset the police, as well as David Viniar and Gary Cohn. There may have been no room for debate, according to Cohn, but the debate was raging all the same.

  Sparks took a stab at trying to explain the opportunity up the chain of command. “Mortgage CDO market has continued to be hammered with combination of the large downward move in subprime RMBS, rating agencies action, and no liquidity,” he wrote to Montag, Viniar, Cohn, and others. He then gave them the example of how Goldman’s own Timberwolf deal, which had been marked at 80 cents on the dollar at the end of May, was then—on August 14—marked at around 20 cents on the dollar. “[I]t’s not just liquidity,” he wrote, “there are fundam
ental cashflow issues.” He then explained that the “best opportunity to make a bunch of money” in the near term was to buy the AAA ABX index as well as other residential mortgage-backed securities. He wrote that he thought that the market seemed to be overreacting and that the mortgage desk had been covering its shorts—at a big profit—but “we will likely come to you soon and say we’d like to get long billions” while also staying short the riskier part of the mortgage market. Cohn responded to Sparks that he wanted him to “talk to me before you go long,” suggesting that the decision would not be reached simply.

  By August 20, Sparks had begun to further flesh out the trade. In an e-mail to Cohn, Winkelried, Viniar, Montag, and Mullen, titled “Big Opportunity,” he reviewed for his bosses the ongoing market meltdown. “We are seeing large liquidations,” he wrote, brought on by a need for liquidity and that was “fear and technically driven.” He mentioned that CIT—the large commercial lender—had called and wanted Goldman to buy $10 billion worth of its loans. “We think it is now time to start using balance sheet,” he continued, “and it is a unique opportunity with real upside … there’s the opportunity for us to make 5–10+ points if we have a longer term hold.” Winkelried responded to Sparks, in part, “Clearly [an] opportunity.”

  The next day, Birnbaum wrote to many of the same executives with his version of what Sparks had described the day before: “The mortgage department thinks there is currently an extraordinary opportunity for those with dry powder to add AAA subprime risk in either cash or synthetic form,” he continued. He suggested that the trade would reduce the mortgage department’s VAR by $75 million and that the arbitrage opportunity implied by the trade could result in big profits. He thought there would be plenty of distressed sellers to provide the supply and that he intended to “share this trade quietly with selected risk partners.”

  The proposal was a bold one, and the profit potential huge if Birnbaum and company were correct. But there also seemed to be some concern at the firm’s highest levels that the group’s recent success had made them a bit cocky. “It would help to manage these guys if u would not answer these guys and keep bouncing them back to Tom [Montag] and I,” Mullen wrote to Winkelried and Cohn. Cohn responded, “Got that and am not answering” but then had to admit the trade had merit. “I do like the idea but you call,” he replied to Mullen. Montag then weighed in. “Just to be clear,” he wrote, “[t]his is buy and hold not buy and sell strategy,” suggesting that the firm’s capital would be committed for some time. Cohn got that. In the end, Sparks and Birnbaum got the green light to “opportunistically … buy assets” at the same time that the mortgage trading group was “significant[ly] covering [its] short positions,” according to a presentation given to the Goldman board of directors in September 2007.

  By the end of August (and Goldman’s third quarter), there was no denying the Birnbaum juggernaut. His structured products trading group was carrying the mortgage business at the firm and keeping it profitable at a time when Goldman’s main competitors on Wall Street were struggling mightily. According to a September 17 presentation delivered to the Goldman board of directors, during the firm’s third quarter, the mortgage origination machine at the firm lost some $200 million—mostly from writing down the value of soured loans—while Birnbaum and company racked up revenue, which was close to pure profit, of $731 million. Indeed, of the $735 million of gross revenues made by Goldman’s mortgage business in the third quarter of 2007, $731 million, or 99.5 percent, came from Birnbaum’s desk. What’s more, of the $1.017 billion of gross revenues the mortgage business generated through the first nine months of 2007, $955 million—or close to Birnbaum’s “bilsky”—came from his desk.

  Through all the turmoil, Goldman was minting money. In the third quarter of 2007, Goldman’s revenues were $12.3 billion—its second-highest quarter of revenue ever—and its net income was $2.9 billion. Goldman’s return on equity was 31.6 percent, nearly unheard of for a public company. On the earnings call, Viniar spoke specifically about the firm’s performance in mortgages. “The mortgage sector continues to be challenged and there was a broad decline in the value of mortgage inventory during the third quarter,” he said. “As a result, we took significant markdowns on our long inventory positions during the quarter, as we had in the previous two quarters. Although we took these marks, our risk bias in that market was to be short and that net short position was profitable. I would also note that you’ve heard me express our generally negative views on the outlook for mortgages since the beginning of the year, so you could correctly assume that we’ve been very aggressive in reducing our long mortgage exposure and conservatively marking down our long mortgage positions.”

  Peter Eavis, a writer at Fortune, was the first to pick up on what he called Goldman’s “stunning strategy” of making a “huge, shrewd bet” while “the credit markets went sour” and “seems to have won big.” Eavis was not sure how Goldman pulled it off, but he observed cannily on September 20 that “[a]s the credit markets fell apart over the summer, causing the prices of hundreds of billions of mortgage-backed bonds to plunge, Goldman Sachs had already positioned itself so that it would profit massively from a decline in those securities.” He noted that Goldman’s third-quarter earnings “were far above expectations” with the “chief contributor to earnings blowout” being “trades that made money from price drops in mortgage-backed securities.” When he tried to figure out how much Goldman had made from the bearish bets—he did not know who at Goldman had been responsible for them—he observed that Viniar, on the third-quarter conference call, “declined to give a number for the amount of money Goldman made on its mortgage short.”

  Not surprisingly, Eavis had not known—could not have known—that Birnbaum and company had been shorting the mortgage market for the previous nine or so months. What he did figure out, correctly, was that a decision would take time to make and to implement. “Amassing a large bearish position in mortgages would have required planning and direction from a senior level,” he wrote. “On the conference call, Viniar said the bet was executed across the whole mortgage business, implying that it wasn’t the work of one swashbuckling trader or trading desk. Of course, the prescience of the short sale would seem to confirm the view that Goldman is the nimblest, and perhaps smartest, brokerage on Wall Street.” He noted that Bear Stearns’s “mortgage business suffered considerably in the quarter” and Morgan Stanley “wasn’t as well hedged to bond losses” (in fact, according to Michael Lewis’s The Big Short, Morgan Stanley would go on to lose $9 billion betting incorrectly on the outcome of the mortgage meltdown).

  Lucas van Praag, Goldman’s longtime head of public relations, sent the Fortune article around to the firm’s top executives. Winkelried, for one, did not appreciate it. “Once again[,] they completely miss the franchise strength and attribute it all to positions and bets,” he wrote to the executive team. Blankfein elaborated on Winkelried’s observation. “Also, the short position wasn’t a bet,” he wrote. “It was a hedge. I.e., the avoidance of a bet. Which is why for a part it subtracted from VAR, not added to VAR.” (This was a very subtle argument, making a fine distinction—and Blankfein kept making it over and over again, even though few people could follow his logic. The fact of the matter was that Goldman had bet—correctly—that the mortgage market would collapse.) In an e-mail a few minutes later sent from his BlackBerry, Peter Kraus, a longtime Goldman partner and then the co-head of the firm’s investment management division, provided Blankfein a particularly unique—perhaps even myopic—insight. He explained that since the third-quarter earnings announcement he had met with more than ten prospective and current clients. “The institutions don’t and I wouldn’t expect them to, make any comments like [‘]ur good at making money for urself but not us,[’]” he wrote to Blankfein. “The individuals do sometimes, but while it requires the utmost humility from us in response[,] I feel very strongly it binds clients even closer to the firm, because the alternative of [‘]take ur money
to a firm who is an under performer and not the best,[’] just isn’t reasonable. Clients ultimately believe association with the best is good for them in the long run.” Needless to say, Blankfein did not respond to Kraus.

  ——

  ON OCTOBER 11, Moody’s—one of the three large bond-rating agencies—downgraded $32 billion of publicly traded mortgage debt that had been originally issued in 2006, the second large and sweeping ratings downgrade by Moody’s in six weeks. Swenson shared the news with Montag and Mullen. “This will eventually filter into downgrades in CDOs,” he wrote, adding that one of the ABX indexes sold off “by a point” after the news, meaning more profits for Goldman. “ABS [asset-backed securities] Desk P and L will be up between [$]30 and [$]35mm today,” he added. Mullen responded, “Nice day.” There was also a discussion of other, more technical consequences of the Moody’s downgrade that Swenson thought would result in interest payments on some of the bonds “being shut off,” or not being paid, which would lower dramatically the value of the securities and mean big profits for anyone—like Goldman—betting they would lose value. “Sounds like we will make some serious money,” Mullen responded. Concluded Swenson, “Yes we are well positioned.” Actually, the profit news that day was even better than Swenson had originally thought. Instead of the mortgage trading desk making between $30 million and $35 million as a result of the Moody’s downgrades, the desk actually ended up making $110 million, $65 million of which came from “yesterday’s downgrades which lead to the selloff in aa[-rated bonds] through bbb[minus-rated bonds] today,” Swenson wrote to Mullen, taking the proverbial “victory lap” that Wall Streeters are so well trained to do, even at Goldman Sachs, the champion of teamwork. “Great Day!” Mullen replied.