On January 31, as the two of them were planning on a weekend dinner date, Tourre wrote to Boukhtouche, “[W]ell I don’t know what you’ve done to the ABX markets, but you must have some sort of influence since today was a relatively calm day.” He then launched into a mini-tirade on the frustrations of life at Goldman Sachs. “[N]evertheless I am still stuck at work at 10 pm,” he continued, “but it’s been six years since I’ve been functioning on this @!$#@!$@$# schedule, so who cares!!! On top of which I have to ‘mentor’ others, in view of the fact that I am now considered a ‘dinosaur.’ In this business (at my firm the average longevity of an employee is about 2–3 years!!!) people ask me about career advice. I feel like I’m losing my mind and I’m only 28!!! OK, I’ve decided two more years of work and I’m retiring;).” That same day Goldman priced another CDO deal—Camber 7—and made a profit of $10 million. After taking an overnight flight to London, Dan Sparks relayed the good news to his boss, Tom Montag. “Need you to send message to [two members of the deal team] telling them what a great job they did,” Sparks wrote. “They structured like mad and travelled [sic] the world, and worked their tails off to make some lemonade from some big old lemons.”
Also on January 31, the ABACUS deal team at Goldman gave Schwartz, at ACA, an update. They suggested removing two names from the hundred-name portfolio because they were “both on negative credit watch by Moody’s” and in their place Goldman wrote that Paulson wanted to include two GSAMP deals. “We will continue our discussions with Paolo to confirm his agreement with the proposed transaction as structured and look forward to discussing the transaction and the draft Engagement Letter.”
On February 2, Tourre and ACA met again with Paulson at his offices to discuss the portfolio to be included in ABACUS. “I am at this [ACA] [P]aulson meeting, this is surreal,” Tourre wrote to a Goldman colleague, without elaborating. Later that day, ACA e-mailed a list of eighty-two mortgage-backed securities that Paulson and ACA agreed should be in ABACUS, plus a list of another twenty-one “replacement” bonds, and then sought Paulson’s approval. “Let me know if these work for you,” ACA wrote. Three days later, Paulson finalized the list of ninety-two bonds, with Tourre’s agreement, and sent an e-mail of them to ACA. That same day, ACA gave preliminary approval to the portfolio to be included in ABACUS. On February 8, Tourre wrote to Sparks that he was finishing up the engagement letters for the ABACUS deal that “will help Paulson short senior tranches off a reference portfolio of … subprime RMBS risk selected by ACA.” Tourre wanted to know if the trade needed the approval of the “Mortgage Capital Committee,” an internal Goldman group set up to approve such things, even though he thought there would be “no commitment for us to take down any risk.” Responded David Rosenblum: “Still reputation risk, so I suggest yes to MCC.” On February 20, Tourre updated his Goldman colleagues with the latest ABACUS thinking and said that he thought Goldman’s fee would likely be increased to $19 million, from $15 million. David Lehman responded to Tourre, “As u know, I am for doing this deal for them/with them, let’s just make sure we are charging enough for it given our axe as principal in this type of risk.” Lehman was concerned that Goldman get paid enough on the deal since it also wanted to sell similar securities to get short and that as an underwriter on the ABACUS deal might end up being long some of the mortgage-related risk when overall it wanted to be short. (Which is exactly what happened—Goldman did get stuck owning some of the long side of the deal when it could not sell the whole thing.) He encouraged Tourre to meet with Birnbaum and Swenson “live” as a “gut check and walk them through it.”
Tourre responded that he thought the team “will be on board with this” but conceded he needed to speak with Birnbaum. “It is really Josh I need to walk carefully through my thinking,” he wrote Lehman on February 21. “Sparks is really relying on us at this point. He is mostly focused on covering our single-names/idiosyncratic short trades to get better observability [sic]”—a reference to the ongoing efforts to establish “the big short.” Lehman responded, again, that it was important to “[w]alk Josh through the $, if that makes sense let’s go.” Lehman also said he needed reminding about whose idea ABACUS was in the first place. “My idea to broker the short,” Tourre replied. “Paulson’s idea to work with a manager. My idea to discuss this with ACA.” Finally, on February 26, after further discussion, Paulson and ACA reached agreement on the ninety bonds that would make up the reference portfolio for ABACUS. That same day, Goldman and ACA prepared a sixty-five-page “flip book,” or PowerPoint presentation, that would be used to market ABACUS to investors who might be willing to take the long side of the trade while Paulson took the short side.
——
AS THE ABX index fell in the first few months of 2007, Goldman’s mortgage department minted money, some “several hundred million dollars,” according to the Wall Street Journal. But there were limits to how far Goldman wanted to push this bet, especially since the whole world seemed to be on the other side of it against the firm and a few other bold hedge-fund managers. There was also the matter of the losses Goldman would have to take as it lowered the value of its long positions on its books. Goldman knew the pain of the losses would hit well before the profits from the shorts could be realized. On February 2, Sparks wrote to Viniar, Ruzika, and Montag, “Gasvoda alerted me last night that we will take a write-down to some retained [long] positions next week as the loan performance data from a few second lien sub-prime deals just came in (comes in monthly) and it is horrible. The team is still working through the numbers, but the amount will likely be in the $20mm loss zone.” He informed his bosses that the team was also working on trying to return the souring mortgage loans to their originators, such as New Century, Washington Mutual, and Fremont, as was permitted by contract. That was becoming more difficult as these companies started to run into greater financial difficulties. “[T]here seem to be issues potentially including some fraud at origination,” Sparks wrote, “but resolution will take months and be contentious.” He concluded that his “[f]ocus is cleaning out rated bond positions and the put back process. Sorry for more bad news.” Montag wondered about Sparks’s “fraud at initiation” comment and Sparks replied, “We’ll be sorting through all the potential breaches of reps and warranties, and fraud at origination (appraisal, income, occupancy) would be a likely one. Fraud is usually borrower, appraiser or broker fraud—not necessarily fraud by the seller of the loans to us. But generally for reps & warranties the loan seller is responsible if fraud happened. The put backs will be a battle.”
On February 8, Sparks provided his bosses with another update and this time also posted Gary Cohn and Jon Winkelried, the Goldman co-presidents. Essentially, there was more bad news. “Subprime environment—bad and getting worse,” he wrote. “Every day is a major fight for some aspect of the business (think whack-a-mole). Trading position has basically squared”—news that the long risk had been finally mitigated—“plan to play from short side. Loan business is long by nature and goal is to mitigate. Credit issues are worsening on deals and pain is broad (including investors in certain GS-issued deals)”—oops! “Distressed opportunities will be real, but we aren’t close to that time yet.” He also addressed a question from Winkelried about whether Goldman was establishing the new trading prices or “chasing them down,” meaning following the leads of other traders. “We have been chasing them down based on loan performance data as it comes out,” he continued. At the end of the same day, Cohn asked Sparks for an update. After reviewing the situation throughout the day with the traders and the controllers, Sparks wrote to Cohn—soon after 11:00 p.m.—that the two constituents agreed Goldman’s loss on the securities should be reflected at $28.4 million, higher than the original thought that it would be a $22 million loss.
Just before midnight, Tourre e-mailed Serres in London. He sent to her an internal analysis by a vice president in mortgage credit trading about the increasingly gloomy outlook for subprime mortgage-backed securities. “The
re has been an increase in early delinquencies and defaults in the Subprime market, most notably those deals backed by … collateral originated in mid-to-late 2005 and 2006,” it read. “We have seen this trend in our 2006 Subprime … deals,” including GSAMP S2. The analysis went on in that depressing vain for several more pages. “You should take a look at this,” Tourre wrote to his girlfriend.
The next day, the push to sell Goldman’s long positions in mortgage securities continued. A list of around thirty of the long positions Goldman owned and wanted to sell fast was circulated. “Below are our updated RMBS axes,” the memo stated, using the argot of traders looking to unload positions. “The focus continues to be on moving credit positions. Again, these are priority positions that should be a focus for everyone before quarter-end.… Let all of the respective desks know how we can be helpful in moving these bonds.” At the end of the day, Walter Scott informed the mortgage group that “this week, a total of [$]169+mm in axe positions were sold” but that “obviously we need to continue to push credit positions across subprime and second liens. We are working with both the desks and the strats to do so.” Kevin Gasvoda replied, “Great job syndicate and sales, appreciate the focus.”
A few hours later, just before midnight, Gasvoda sent Montag a detailed accounting of the financial risks in Goldman’s mortgage portfolio. He explained that the firm had taken around $70 million of write-downs on the overall mortgage portfolio in the first five weeks of 2007, with another $70 million more likely to come. He said the losses had all been in those securities in HPA—home price appreciation—sensitive sectors. “They’ve crumbled under HPA slowdown as these are the most levered borrowers.” Gasvoda told Montag that to “mitigate” these losses, Goldman had stopped buying subprime second liens in the summer of 2006 and instead focused on prime mortgages and what were known as Alt-A mortgages, those between subprime and prime. Goldman had also focused on selling new mortgage-backed securities “at any clearing levels,” or whatever price the market would bear, to get rid of them and had given traders, such as Birnbaum, the authority to sell any remaining “retained bonds.”
Three days later, on Sunday, Montag passed Gasvoda’s analysis on to Winkelried and Blankfein. “Very good writeup of our positions in each sector[,] hedges we have on and potential for further write-down over next six months,” he wrote. Fourteen minutes later, Blankfein responded, wanting to know the “short summary of our risk” and what “further writedowns” would be. Montag replied with a summary of Gasvoda’s analysis, although it was not that easy to follow. “If things got no worse,” he concluded, “the desk—perhaps in wishful mode—feels they have gains we haven’t shown. [T]hey did make [$]21[mm] on Friday outside of write down.” Blankfein replied in short order, “Tom, you refer to losses stemming from residual positions in old deals. Could/should we have cleaned up these books before, and are we doing enough right now to sell off cats and dogs in other books throughout the division[?]” This question got Montag thinking. “Should we have done before?” he replied, rhetorically. “Most likely.” He then explained the steps the firm had taken but added that he thought the cleanup had been ongoing “for years” and took credit himself for moving “residuals out of origination and into traders” for them to sell.
But there were plenty of crosscurrents buffeting the mortgage market in February 2007 and plenty of people who disagreed with Goldman’s decision to get short the mortgage market. For instance, the next day—February 12—Gyan Sinha, a senior managing director at Bear Stearns in charge of the firm’s market research regarding asset-backed securities and collateralized debt obligations, held a conference call for some nine hundred investors where he spelled out his beliefs about the market’s reaction to the news that New Century, the mortgage lender, was having financial trouble. To that point, Sinha had been very well respected and had even testified in front of Congress about the subprime mortgage market. “It’s time to buy the [ABX] index,” he said, adding that based on his modeling, “the market has overreacted” and predictions of rising problems in the mortgage market should be taken “with a large grain of salt.” Many investors shared Sinha’s view.
Two days later, on Valentine’s Day, New Century announced that a wave of shareholder lawsuits had been filed against it and that after two weeks of tough negotiations, Goldman Sachs had agreed to a three-month extension of a line of credit to the company that had been set to expire the next day. Goldman had extracted its pound of flesh by insisting on the ability to get out of the agreement “at the first hint of trouble.” At 6:33 that morning, Sparks, who had long been worried about this kind of problem, wrote himself an e-mail, titled “Risk,” to help keep track of the increasingly volatile events. “Bad week in subprime,” he wrote. He noted that “originators”—such as New Century—“are really in a bad spot. Thinly capitalized, highly levered, dealing with significant loan put-backs … now having trouble selling loans above par when it cost them 2 points to produce. Will have to … really tighten credit standards which will cut volume significantly.” He wondered what the next area of “contagion” might be and answered himself that it would be CDOs, “which have been the buyer of most single name mezz[anine] subprime risk for the past year.” He noted that Goldman was doing four things to reduce its risk: finding warehouse “risk partners,” giving the secondary trading desk at Goldman—essentially Birnbaum and company—the ultimate authority “so all risk housed and managed by traders,” buying protection on CDOs, and “executing deals.”
None of this was communicated to Goldman’s clients, of course.
Later that morning, Sparks summarized Goldman’s “risk reduction program” for his bosses, including Montag, Viniar, Ruzika, and Gary Cohn. He copied Winkelried on the memo. The strategy consisted, he wrote, of selling the ABX index, of buying CDS on individual tranches of mortgage-backed securities, buying CDS—some $3 billion worth—on the “super-senior portions of BBB/BBB- index” and which seemed to be significantly in the money. “This is good for us position-wise,” he wrote, “[but] bad for accounts who wrote that protection [to us] (M[organ] S[tanley] Prop[rietary trading desk], Peleton [a hedge fund], ACA, Harvard) but could hurt our CDO pipeline position as CDOs will be harder to do.” Cohn sent Montag’s e-mail on to Blankfein, without comment.
——
ON FEBRUARY 17, the Wall Street Journal interviewed Lew Ranieri and reported that the “rumpled 60-year-old says he is worried about the proliferation of risky mortgages and convoluted ways of financing them. Too many investors don’t understand the dangers.… The problem, he says, is that in the past few years the business has changed so much that if the U.S. housing market takes another lurch downward, no one will know where all the bodies are buried. ‘I don’t know how to understand the ripple effects throughout the system today,’ he said during a recent seminar.” The growing problem was that 40 percent of the subprime borrowers in 2006 were not required to produce pay stubs or other proof of their net worth, according to Credit Suisse Group, and lenders were relying more and more on computer models to estimate the value of homes. “We’re not really sure what the guy’s income is and … we’re not sure what the home is worth,” Ranieri said. “So you can understand why some of us become a little nervous.” He worried further that with so many mortgages being packaged by Wall Street into CDOs and sold in slices to investors all over the world, U.S. home mortgage risks were being spread to a “much less sophisticated community.” The Journal made it clear that “Mr. Ranieri isn’t predicting Armageddon. Some of the riskier new types of mortgages probably will perform ‘horribly’ in terms of defaults, leading to losses for some investors. But, he says, the ‘vast majority’ of mortgages outstanding are based on sounder lending principles and should be fine.”
Most Wall Street investors and executives were not sure just what to do. Were the cracks in the mortgage market, as reflected in the decline of the ABX, a buying opportunity, as Bear’s Sinha suggested? Or were the cracks the first small fissu
res in what soon would be a spectacular collapse of the market for mortgages and mortgage-backed securities? Major proponents of the glass-is-half-full thinking were the two Bear Stearns hedge-fund managers, Ralph Cioffi and Matthew Tannin. Apparently unbeknownst to many of their investors who thought Cioffi and Tannin had invested in less risky securities, the two Bear Stearns hedge funds—which together had around $1.5 billion of investor money riding—were heavily invested in mortgage-backed securities, including the synthetic CDOs Goldman had been selling. Like their Bear colleague Sinha, Cioffi and Tannin were generally of the view that the dip in the ABX index was a buying opportunity.
On February 21, Tannin sent an e-mail to his colleagues at Bear in which he sounded quite happy about all the doom and gloom over subprime mortgages in the marketplace. He cited such a negative report from a rival hedge-fund manager and said, “This piece is mostly unhelpful and more than a bit misleading. Scare mongering. I used to fly into a rage when I would read this stuff but now it makes me happy. We need some caution and naysayers in our market—it keeps spreads wider. So I’m glad this has been printed.” A week later, Cioffi wrote to the team that he was thinking of “very selectively buying at these levels” since that “in and of itself would stabilize the markets.” Tannin responded that he thought it would be a good idea. “Fear + illiquidity + a CDO ready and waiting ⁼ a good trade.” That same day, Ben Bernanke, the chairman of the Federal Reserve, testified on Capitol Hill that he did not believe a “housing downturn” was a “broad financial concern or a major factor in assessing the state of the economy.”
Money and Power Page 73
