The Big Short: Inside the Doomsday Machine

by Michael Lewis

  * Even now, after the death of Lehman Brothers, LehmanLive remains the ghostly go-to source for the contents of many CDOs.

  * When the market cracked, Devaney went bust and was forced to sell his yacht, his plane, and his Renoir (for a nice profit) and defend himself against several nasty newspaper articles. "It takes an honest individual to admit that he was wrong," he wrote, in one of several rambling letters released over the PR Newswire. "I was long in 2007 and was wrong."

  "He was incredibly cynical about the market," said Charlie. "And he lost money. I never figured that out."

  * Two years later, Las Vegas would lead the nation in its rate of home foreclosures.

  * In Las Vegas they also met with David Wells, who ran subprime lending for a company called Fremont Investment & Loan. Wells also said he expected losses to run 5 percent. In September, nine months later, Fremont would announce that 30 percent of its subprime loans were in default. Its pools of loans would register losses higher than 40 percent--which is to say that, even after it sold the houses it foreclosed upon, it was out nearly half the money it loaned.

  * The "spread" on any bond is simply the difference between the interest rate it pays to the investor, and some putatively risk-free rate--say, the rate paid to investors in U.S. Treasury bonds.

  * A brief reminder: In thinking about these towers of debt, it's handy to simplify them into three floors: a basement, called the "equity," which takes the very first losses and is not an investment-grade security; the lower floor, called the "mezzanine," with triple-B rating; and the upper floor, with triple-A rating, and generally referred to as the "senior." In practice, the towers were far more finely sliced: a CDO might have fifteen different tranches, each with a slightly different rating, from triple-B-minus all the way up to triple-A: triple-B-minus, triple-B, A-minus, A, and so on. The double-A rating of the tranche shorted by Cornwall Capital implied that the underlying bonds, though slightly more risky than supposedly gold-plated triple-As, still had a less than 1 percent chance of defaulting.

  * A spokesman for S&P later doubted that any S&P employee would ever have said such a thing, as their model was capable of handling negative numbers.

  * On October 22, 2008, a former S&P subprime mortgage bond analyst named Frank Raiter would testify before the Committee on Oversight and Government Reform that the S&P managing director in charge of the surveillance of subprime mortgage bonds "did not believe loan-level data was necessary and that had the effect of quashing all requests for funds to build in-house data bases." Raiter introduced an e-mail from S&P's managing director of CDO ratings, Richard Gugliada, in which Gugliada said: "Any request for loan-level tapes is TOTALLY UNREASONABLE!! Most originators don't have it and can't provide it. Nevertheless we MUST produce a credit estimate.... It is your responsibility to provide those credit estimates and your responsibility to devise some method to do so."

  * A Connecticut-based hedge fund that lost $6.8 billion in bets on natural gas in early 2006 and blew up in spectacular fashion.

  * The distinction had become superficial. Alt-A borrowers had FICO credit scores above 680; subprime borrowers had FICO scores below 680. Alt-A loans were poorly documented, however; the borrower would fail to provide proof of income, for instance. In practice, Alt-A mortgage loans made in the United States between 2004 and 2008 totaling $1.2 trillion were as likely to default as subprime loans totaling $1.8 trillion.

  * A silent second is a second mortgage used, in the purchase of a house, to supplement a first mortgage. It is silent only to the guy who made the first loan, and who is less likely to be repaid, as the borrower is less likely to have any financial stake at all in his own home.

  * Just about everyone involved in the financial crisis stands to lose money if he is caught talking about what he saw and did. Obviously those still employed at the big Wall Street firms, but even those who have moved on, as they have typically signed some nondisclosure agreement. Morgan Stanley's former employees are not quite as spooked as those who worked at Goldman Sachs, but they're close.

  * Of all the conflicts of interests inside a Wall Street bond trading firm, here was both the most pernicious and least discussed. When a firm makes bets on stocks and bonds for its own account at the same time that it brokers them to customers, it faces great pressure to use its customers for the purposes of its own account. Wall Street firms like to say they build Chinese walls to keep information about customer trading from leaking to their own proprietary traders. Vincent Daniel of FrontPoint Partners offered the most succinct response to this pretense: "When I hear 'Chinese wall,' I think, You're a fucking liar."

  * Here it's useful to remember that selling a credit default swap on a thing leaves you with the same financial risk as if you owned it. If the triple-A CDO ends up being worth zero, you lose the same amount whether you bought it outright or sold a credit default swap on it.

  * The timing of Goldman's departure from the subprime market is interesting. Long after the fact, Goldman would claim it had made that move in December 2006. Traders at big Wall Street firms who dealt with Goldman felt certain that the firm did not reverse itself until the spring and early summer of 2007, after New Century, the nation's biggest subprime lender, filed for bankruptcy. If this is indeed when Goldman "got short," it would explain the chaos in both the subprime market and Goldman Sachs, perceived by Mike Burry and others, in late June. Goldman Sachs did not leave the house before it began to burn; it was merely the first to dash through the exit--and then it closed the door behind it.

  * There is some dispute about the conversations between Hubler and Cruz. The version of events offered by people close to Zoe Cruz is that she was worried about the legal risk of doing business with Bear Stearns's troubled hedge funds, and that Hubler never completely explained the risk of triple-A-rated CDOs to her, and led her to believe that Morgan Stanley stood no chance of suffering a huge loss--probably because Hubler himself didn't understand the risk. Hubler's friends claim that Cruz seized effective control of Hubler's trade and prevented him from ditching some large chunk of his triple-A CDOs. In my view, and in the view of Wall Street traders, Hubler's story line is far less plausible. "There's no fucking way he said, 'I have to get out now' and she said no," says one trader close to the situation. "No way Howie ever said, 'If we don't get out now we might lose ten billion dollars.' Howie presented her with a case for not getting out." The ability of Wall Street traders to see themselves in their success and their management in their failure would later be echoed, when their firms, which disdained the need for government regulation in good times, insisted on being rescued by government in bad times. Success was individual achievement; failure was a social problem.

  * It's too much to expect the people who run big Wall Street firms to speak plain English, since so much of their livelihood depends on people believing that what they do cannot be translated into plain English. What John Mack's trying to say, without coming right out and saying that no one else at Morgan Stanley had a clue what risks Howie Hubler was running, is that no one else at Morgan Stanley had a clue what risks Howie Hubler was running--and neither did Howie Hubler.

  + Another way to put the same question: How could Howie Hubler's bonds plunge from 100 to 7 and the reports you received still suggest that they were incapable of dramatic movement?

  * It's interesting to imagine how the disaster might have played out if AIG FP had simply continued to take all the risk. If Wall Street, following Goldman Sachs's lead, had dumped all of the risk of subprime mortgage bonds into AIG FP, the problem might well have been classified as having nothing to do with Wall Street and as being the sole responsibility of this bizarre insurance company.

  * Later revised to about $10 a share.

  * The case brought by the U.S. Department of Justice against Cioffi and Tannin sought to prove that the two men had knowingly deceived their investors, overlooking the possibility that they simply had no idea what they were doing, and failed to grasp the real risk of a triple-A-r
ated subprime-backed CDO. The case was weak, and turned on a couple of e-mails obviously ripped from context. A member of the jury that voted to acquit the Bear Stearns subprime bond traders told Bloomberg News afterward not only that she thought they were innocent as charged but that she would happily invest money with them.