The Big Short: Inside the Doomsday Machine

by Michael Lewis

  "This is a recurrent theme whenever the market moves our way," wrote Burry. "People get sick, people are off for unspecified reasons."

  On June 20, Grinstein finally returned to tell him that Goldman Sachs had experienced "systems failure."

  That was funny, Burry replied, because Morgan Stanley had said more or less the same thing. And his salesman at Bank of America claimed they'd had a "power outage."

  "I viewed these 'systems problems' as excuses for buying time to sort out a mess behind the scenes," he said. The Goldman saleswoman made a weak effort to claim that, even as the index of subprime mortgage bonds collapsed, the market for insuring them hadn't budged. But she did it from her cell phone, rather than the office line, on which the conversations would have been recorded.

  They were caving. All of them. At the end of every month, for nearly two years, Burry had watched Wall Street traders mark his positions against him. That is, at the end of every month his bets against subprime bonds were mysteriously less valuable. The end of every month also happened to be when Wall Street traders sent their profit and loss statements to their managers and risk managers. On June 29, Burry received a note from his Morgan Stanley salesman, Art Ringness, saying that Morgan Stanley now wanted to make sure that "the marks are fair." The next day, Goldman followed suit. It was the first time in two years that Goldman Sachs had not moved the trade against him at the end of the month. "That was the first time they moved our marks accurately," he notes, "because they were getting in on the trade themselves." The market was finally accepting the diagnosis of its own disorder.

  The moment Goldman was getting in on his trade was also the moment the market flipped. Some kind of rout was now on: Everyone at once seemed eager to talk to him. Morgan Stanley, which had been, by far, the most reluctant to acknowledge negative news in subprime, now called to say it would like to buy whatever he had "in any size." Burry heard a rumor--soon confirmed--that a fund run by Goldman, called Global Alpha, had taken huge losses in subprime and that Goldman itself had rapidly turned from betting on the subprime mortgage market to betting against it.

  It was precisely the moment he had told his investors, back in the summer of 2005, that they only needed to wait for. Crappy mortgages worth three-quarters of a trillion dollars were resetting from their teaser rates to new, higher rates. A single pool of mortgages, against which Burry had laid a bet, illustrated the general point: OOMLT 2005-3. OOMLT 2005-3 was shorthand for a pool of subprime mortgage loans made by Option One--the company whose CEO had given the speech in Las Vegas that Steve Eisman had walked out of, after raising his zero in the air. Most of the loans had been made between April and July of 2005. From January to June 2007, the news from the pool--its delinquencies, its bankruptcies, its house foreclosures--had remained fairly consistent. The losses were much greater than they should have been, given the ratings of the bonds they underpinned, but the losses did not change a great deal from one month to the next. From February 25 to May 25 (the remittance data always came on the twenty-fifth of the month), the combined delinquencies, foreclosures, and bankruptcies inside OOMLT 2005-3 rose from 15.6 percent to 16.9 percent. On June 25 the total number of loans in default spiked to 18.68 percent. In July it spiked again, to 21.4 percent. In August it leapt to 25.44 percent, and by the end of the year it stood at 37.7 percent--more than a third of the pool of borrowers had defaulted on their loans. The losses were sufficient to wipe out not only the bonds Michael Burry had bet against but also a lot of the more highly rated ones in the same tower. That the panic inside Wall Street firms had begun before June 25 suggested to Michael Burry mainly that the Wall Street firms might be working with inside information about the remittance data. "The dealers often owned [mortgage] servicers," he wrote, "and might have been able to get an inside track on the deterioration in the numbers."

  In the months leading up to the collapse of OOMLT 2005-3--and all of the other pools of home loans he had bought credit default swaps on--Michael Burry noted several remarks from both Ben Bernanke and the Secretary of the U.S. Treasury, Henry Paulson. Each said, repeatedly, that he saw no possibility of "contagion" in the financial markets from the losses in subprime mortgages. "When I first started shorting these mortgages in 2005," Burry wrote in an e-mail, "I knew full well that it was not likely to pay out within two years--and for a very simple reason. The vast majority of mortgages originated the last few years had a rather ominously attractive feature called the 'teaser rate period.' Those 2005 mortgages are only now reaching the end of their teaser rate periods, and it will be 2008 before the 2006 mortgages get there. What sane person on Earth would confidently conclude in early 2007, smack dab in the midst of the mother of all teaser rate scams, that the subprime fallout will not result in contagion? The bill literally has not even come due."

  Across Wall Street, subprime mortgage bond traders were long and wrong, and scrambling to sell their positions--or to buy insurance on them. Michael Burry's credit default swaps were suddenly fashionable. What still shocked him, however, was that the market had been so slow to assimilate material information. "You could see that all these deals were sucking wind leading up to the reset date," he said, "and the reset just goosed them into another dimension of fail. I was in a state of perpetual disbelief. I would have thought that someone would have recognized what was coming before June 2007. If it really took that June remit data to cause a sudden realization, well, it makes me wonder what a 'Wall Street analyst' really does all day."

  By the end of July his marks were moving rapidly in his favor--and he was reading about the genius of people like John Paulson, who had come to the trade a year after he had. The Bloomberg News service ran an article about the few people who appeared to have seen the catastrophe coming. Only one worked as a bond trader inside a big Wall Street firm: a formerly obscure asset-backed bond trader at Deutsche Bank named Greg Lippmann. FrontPoint and Cornwall were both missing from the piece, but the investor most conspicuously absent from the Bloomberg News article sat alone in his office, in Cupertino, California. Michael Burry clipped the article and e-mailed it around the office with a note: "Lippmann is the guy that essentially took my idea and ran with it. To his credit." His own investors, whose money he was doubling and more, said little. There came no apologies, and no gratitude. "Nobody came back and said, 'Yeah, you were right,'" he said. "It was very quiet. It was extremely quiet. The silence infuriated me." He was left with his favored mode of communication, his letter to investors. In early July 2007, as the markets crashed, he posed an excellent question. "One rather surprising aspect of all this," he wrote, "is that there have been relatively few reports of investors actually being hurt by the subprime mortgage market troubles....Why have we not yet heard of this era's Long-Term Capital?"

  CHAPTER NINE

  A Death of Interest

  Howie Hubler had grown up in New Jersey and played football at Montclair State College. Everyone who met him noticed his thick football neck and his great huge head and his overbearing manner, which was interpreted as both admirably direct and a mask. He was loud and headstrong and bullying. "When confronted with some intellectual point about his trades, Howie wouldn't go to an intellectual place," said one of the people charged with supervising Hubler in his early days at Morgan Stanley. "He would go to 'Get the hell out of my face.'" Some people enjoyed Hubler, some people didn't, but, by early 2004, what others thought didn't really matter anymore, because for nearly a decade Howie Hubler had made money trading bonds for Morgan Stanley. He ran Morgan Stanley's asset-backed bond trading, which effectively put him in charge of the firm's bets on subprime mortgages. Right up to the point the subprime mortgage bond market boomed, and changed what it meant to be an asset-backed bond trader, Hubler's career had resembled Greg Lippmann's. Like every other asset-backed bond trader, he'd been playing a low-stakes poker game rigged in his favor, since nothing had ever gone seriously wrong in the market. Prices fell, but they always came back. You could either like asset-backed bonds or y
ou could love asset-backed bonds, but there was no point in hating them, because there was no tool for betting against them.

  Inside Morgan Stanley, the subprime mortgage lending boom created a who-put-chocolate-in-my peanut-butter moment. The firm had been a leader in extending into consumer loans the financial technology used to package corporate loans. Morgan Stanley's financial intellectuals--their quants--had been instrumental in teaching the rating agencies, Moody's and S&P, how to evaluate CDOs on pools of asset-backed bonds. It was only natural that someone inside Morgan Stanley should also wonder if he might invent a credit default swap on an asset-backed bond. Howie Hubler's subprime mortgage desk was creating bonds at a new and faster rate. To do so, Hubler's group had to "warehouse" loans, sometimes for months. Between the purchase of the loans and the sale of the bonds made up of those loans, his group was exposed to falling prices. "The whole reason we created the credit default swap was to protect the mortgage desk run by Howie Hubler," said one of its inventors. If Morgan Stanley could find someone to sell it insurance on its loans, Hubler could eliminate the market risk of warehousing home loans.

  As originally conceived, in 2003, the subprime mortgage credit default swap was a one-off, nonstandard insurance contract, struck between Morgan Stanley and some other bank or insurance company, outside the gaze of the wider market. No ordinary human being had ever heard of these credit default swaps or, if Morgan Stanley had its way, ever would. By design they were arcane, opaque, illiquid, and thus conveniently difficult for anyone but Morgan Stanley to price. "Bespoke," in market parlance. By late 2004 Hubler had grown cynical about certain subprime mortgage bonds--and wanted to find clever ways to bet against them. The same idea had occurred to Morgan Stanley's intellectuals. In early 2003 one of them had proposed that they cease to be intellectuals and form a little group that he, the intellectual, would manage--a fact that the traders would quickly forget. "One of the quants actually creates all this stuff and they [Hubler and his traders] stole it from him," said a Morgan Stanley bond saleswoman who observed the proceedings up close. One of Hubler's close associates, a trader named Mike Edman, became the official creator of a new idea: a credit default swap on what amounted to a timeless pool of subprime loans.

  One risk of betting against subprime loans was that, as long as house prices kept rising, borrowers were able to refinance, and pay off their old loans. The pool of loans on which you've bought insurance shrinks, and the amount of your insurance shrinks with it. Edman's credit default swap solved this problem with some fine print in its contracts, which specified that Morgan Stanley was buying insurance on the last outstanding loan in the pool. Morgan Stanley was making a bet not on the entire pool of subprime home loans but on the few loans in the pool least likely to be repaid. The size of the bet, however, remained the same as if no loan in the pool was ever repaid. They had bought flood insurance that, if a drop of water so much as grazed any part of the house, paid them the value of the entire house.

  Thus designed, Morgan Stanley's new bespoke credit default swap was virtually certain one day to pay off. For it to pay off in full required losses in the pool of only 4 percent, which pools of subprime mortgage loans experienced in good times. The only problem, from the point of view of Howie Hubler's traders, was finding a Morgan Stanley customer stupid enough to take the other side of the bet--that is, to get the customer to sell Morgan Stanley what amounted to home insurance on a house designated for demolition. "They found one client to take the long side of the triple-B tranche of some piece of shit," says one of their former colleagues, which is a complicated way of saying that they found a mark. A fool. A customer to be taken advantage of. "That's how it starts--it drives Howie's first trade."

  By early 2005 Howie Hubler had found a sufficient number of fools in the market to acquire 2 billion dollars' worth of these bespoke credit default swaps. From the point of view of the fools, the credit default swaps Howie Hubler was looking to buy must have looked like free money: Morgan Stanley would pay them 2.5 percent a year over the risk-free rate to own, in effect, investment-grade (triple-B-rated) asset-backed bonds. The idea appealed especially to German institutional investors, who either failed to read the fine print or took the ratings at face value.

  By the spring of 2005, Howie Hubler and his traders believed, with reason, that these diabolical insurance policies they'd created were dead certain to pay off. They wanted more of them. It was now, however, that Michael Burry began to agitate to buy standardized credit default swaps. Greg Lippmann at Deutsche Bank, a pair of traders at Goldman Sachs, and a few others came together to hammer out the details of the contract. Mike Edman at Morgan Stanley was dragged kicking and screaming into their discussion, for the moment credit default swaps on subprime mortgage bonds were openly traded and standardized, Howie Hubler's group would lose their ability to peddle their murkier, more private version.

  It's now April 2006, and the subprime mortgage bond machine is roaring. Howie Hubler is Morgan Stanley's star bond trader, and his group of eight traders is generating, by their estimate, around 20 percent of Morgan Stanley's profits. Their profits have risen from roughly $400 million in 2004 to $700 million in 2005, on their way to $1 billion in 2006. Hubler will be paid $25 million at the end of the year, but he's no longer happy working as an ordinary bond trader. The best and the brightest Wall Street traders are quitting their big firms to work at hedge funds, where they can make not tens but hundreds of millions. Collecting nickels and dimes from the trades of unthinking investors felt beneath the dignity of a big-time Wall Street bond trader. "Howie thought the customer business was stupid," says one of several traders closest to Hubler. "It was what he'd always done, but he'd lost interest in it."* Hubler could make hundreds of millions facilitating the idiocy of Morgan Stanley's customers. He could make billions by using the firm's capital to bet against them.

  Morgan Stanley management, for its part, always feared that Hubler and his small team of traders might quit and create their own hedge fund. To keep them, they offered Hubler a special deal: his own proprietary trading group, with its own grandiose name: GPCG, or the Global Proprietary Credit Group. In his new arrangement, Hubler would keep for himself some of the profits this group generated. "The idea," says a member of the group, "was for us to go from making one billion dollars a year to two billion dollars a year, right away." The idea, also, was for Hubler and his small group of traders to keep for themselves a big chunk of the profits this group generated. As soon as feasible, Morgan Stanley promised, Hubler would be allowed to spin it off into a separate money management business, of which he'd own 50 percent. Among other things, this business would manage subprime-backed CDOs. They would compete, for instance, with Wing Chau's Harding Advisory.

  The putative best and brightest on Morgan Stanley's bond trading floor lobbied to join him. "It was supposed to be the elite of the elite," said one of the traders. "Howie took all the smartest people with him." The chosen few moved to a separate floor in Morgan Stanley's midtown Manhattan office, eight floors above their old trading desks. There they erected new walls around themselves, to create at least the illusion that Morgan Stanley had no conflict of interest. The traders back down on the second floor would buy and sell from customers and not pass any information about their dealings to Hubler and his group on the tenth floor. Tony Tufariello, the head of Morgan Stanley's global bond trading and thus in theory Howie Hubler's boss, was so conflicted that he built himself an office inside Howie's group, and bounced back and forth between the second floor and the tenth.* Howie Hubler didn't want only people. He wanted, badly, to take with him his group's trading positions. Their details were complicated enough that one of Morgan Stanley's own subprime mortgage bond traders said, "I don't think any of the people above Howie fully understood the trade he had on." But their gist was simple: Hubler and his group had made a massive bet that subprime loans would go bad. The crown jewel of their elaborate trading positions was still the $2 billion in bespoke credit default swaps
Hubler felt certain would one day very soon yield $2 billion in pure profits. The pools of mortgage loans were just about to experience their first losses, and the moment they did, Hubler would be paid in full.

  There was, however, a niggling problem: The running premiums on these insurance contracts ate into the short-term returns of Howie's group. "The group was supposed to make two billion dollars a year," said one member. "And we had this credit default swap position that was costing us two hundred million dollars." To offset the running cost, Hubler decided to sell some credit default swaps on triple-A-rated subprime CDOs, and take in some premiums of his own.* The problem was that the premiums on the supposedly far less risky triple-A-rated CDOs were only one-tenth of the premiums on the triple-Bs, and so to take in the same amount of money as he was paying out, he'd need to sell credit default swaps in roughly ten times the amount he already owned. He and his traders did this quickly, and apparently without a great deal of discussion, in half a dozen or so massive trades, with Goldman Sachs and Deutsche Bank and a few others.

  By the end of January 2007, when the entire subprime mortgage bond industry headed to Las Vegas to celebrate itself, Howie Hubler had sold credit default swaps on roughly 16 billion dollars' worth of triple-A tranches of CDOs. Never had there been such a clear expression of the delusion of the elite Wall Street bond trader and, by extension, the entire subprime mortgage bond market: Between September 2006 and January 2007, the highest-status bond trader inside Morgan Stanley had, for all practical purposes, purchased $16 billion in triple-A-rated CDOs, composed entirely of triple-B-rated subprime mortgage bonds, which became valueless when the underlying pools of subprime loans experienced losses of roughly 8 percent. In effect, Howie Hubler was betting that some of the triple-B-rated subprime bonds would go bad, but not all of them. He was smart enough to be cynical about his market but not smart enough to realize how cynical he needed to be.