The Big Short: Inside the Doomsday Machine

by Michael Lewis

  The long answer was that there were huge sums of money to be made, if you could somehow get them re-rated as triple-A, thereby lowering their perceived risk, however dishonestly and artificially. This is what Goldman Sachs had cleverly done. Their--soon to be everyone's--nifty solution to the problem of selling the lower floors appears, in retrospect, almost magical. Having gathered 100 ground floors from 100 different subprime mortgage buildings (100 different triple-B-rated bonds), they persuaded the rating agencies that these weren't, as they might appear, all exactly the same things. They were another diversified portfolio of assets! This was absurd. The 100 buildings occupied the same floodplain; in the event of flood, the ground floors of all of them were equally exposed. But never mind: The rating agencies, who were paid fat fees by Goldman Sachs and other Wall Street firms for each deal they rated, pronounced 80 percent of the new tower of debt triple-A.

  The CDO was, in effect, a credit laundering service for the residents of Lower Middle Class America. For Wall Street it was a machine that turned lead into gold.

  Back in the 1980s, the original stated purpose of the mortgage-backed bond had been to redistribute the risk associated with home mortgage lending. Home mortgage loans could find their way to the bond market investors willing to pay the most for them. The interest rate paid by the homeowner would thus fall. The goal of the innovation, in short, was to make the financial markets more efficient. Now, somehow, the same innovative spirit was being put to the opposite purpose: to hide the risk by complicating it. The market was paying Goldman Sachs bond traders to make the market less efficient. With stagnant wages and booming consumption, the cash-strapped American masses had a virtually unlimited demand for loans but an uncertain ability to repay them. All they had going for them, from the point of view of Wall Street financial engineers, was that their financial fates could be misconstrued as uncorrelated. By assuming that one pile of subprime mortgage loans wasn't exposed to the same forces as another--that a subprime mortgage bond with loans heavily concentrated in Florida wasn't very much like a subprime mortgage bond more concentrated in California--the engineers created the illusion of security. AIG FP accepted the illusion as reality.

  The people who worked on the relevant Goldman Sachs mortgage bond trading desk were all extremely intelligent. They'd all done amazingly well in school and had gone to Ivy League universities. But it didn't require any sort of genius to see the fortune to be had from the laundering of triple-B-rated bonds into triple-A-rated bonds. What demanded genius was finding $20 billion in triple-B-rated bonds to launder. In the original tower of loans--the original mortgage bond--only a single, thin floor got rated triple-B. A billion dollars of crappy home loans might yield just $20 million of the crappiest triple-B tranches. Put another way: To create a billion-dollar CDO composed solely of triple-B-rated subprime mortgage bonds, you needed to lend $50 billion in cash to actual human beings. That took time and effort. A credit default swap took neither.

  There was more than one way to think about Mike Burry's purchase of a billion dollars in credit default swaps. The first was as a simple, even innocent, insurance contract. Burry made his semiannual premium payments and, in return, received protection against the default of a billion dollars' worth of bonds. He'd either be paid zero, if the triple-B-rated bonds he'd insured proved good, or a billion dollars, if those triple-B-rated bonds went bad. But of course Mike Burry didn't own any triple-B-rated subprime mortgage bonds, or anything like them. He had no property to "insure" it was as if he had bought fire insurance on some slum with a history of burning down. To him, as to Steve Eisman, a credit default swap wasn't insurance at all but an outright speculative bet against the market--and this was the second way to think about it.

  There was also a third, even more mind-bending, way to think of this new instrument: as a near-perfect replica of a subprime mortgage bond. The cash flows of Mike Burry's credit default swaps replicated the cash flows of the triple-B-rated subprime mortgage bond that he wagered against. The 2.5 percent a year in premium Mike Burry was paying mimicked the spread over LIBOR* that triple-B subprime mortgage bonds paid to an actual investor. The billion dollars whoever had sold Mike Burry his credit default swaps stood to lose, if the bonds went bad, replicated the potential losses of an actual bond owner.

  On its surface, the booming market in side bets on subprime mortgage bonds seemed to be the financial equivalent of fantasy football: a benign, if silly, facsimile of investing. Alas, there was a difference between fantasy football and fantasy finance: When a fantasy football player drafts Peyton Manning to be on his team, he doesn't create a second Peyton Manning. When Mike Burry bought a credit default swap based on a Long Beach Savings subprime-backed bond, he enabled Goldman Sachs to create another bond identical to the original in every respect but one: There were no actual home loans or home buyers. Only the gains and losses from the side bet on the bonds were real.

  And so, to generate $1 billion in triple-B-rated subprime mortgage bonds, Goldman Sachs did not need to originate $50 billion in home loans. They needed simply to entice Mike Burry, or some other market pessimist, to pick 100 different triple-B bonds and buy $10 million in credit default swaps on each of them. Once they had this package (a "synthetic CDO," it was called, which was the term of art for a CDO composed of nothing but credit default swaps), they'd take it over to Moody's and Standard & Poor's. "The ratings agencies didn't really have their own CDO model," says one former Goldman CDO trader. "The banks would send over their own model to Moody's and say, 'How does this look?'" Somehow, roughly 80 percent of what had been risky triple-B-rated bonds now looked like triple-A-rated bonds. The other 20 percent, bearing lower credit ratings, generally were more difficult to sell, but they could, incredibly, simply be piled up in yet another heap and reprocessed yet again, into more triple-A bonds. The machine that turned 100 percent lead into an ore that was now 80 percent gold and 20 percent lead would accept the residual lead and turn 80 percent of that into gold, too.

  The details were complicated, but the gist of this new money machine was not: It turned a lot of dicey loans into a pile of bonds, most of which were triple-A-rated, then it took the lowest-rated of the remaining bonds and turned most of those into triple-A CDOs. And then--because it could not extend home loans fast enough to create a sufficient number of lower-rated bonds--it used credit default swaps to replicate the very worst of the existing bonds, many times over. Goldman Sachs stood between Michael Burry and AIG. Michael Burry forked out 250 basis points (2.5 percent) to own credit default swaps on the very crappiest triple-B bonds, and AIG paid a mere 12 basis points (0.12 percent) to sell credit default swaps on those very same bonds, filtered through a synthetic CDO, and pronounced triple-A-rated. There were a few other messy details*--some of the lead was sold off directly to German investors in Dusseldorf--but when the dust settled, Goldman Sachs had taken roughly 2 percent off the top, risk-free, and booked all the profit up front. There was no need on either side--long or short--for cash to change hands. Both sides could do a deal with Goldman Sachs by signing a piece of paper. The original home mortgage loans on whose fate both sides were betting played no other role. In a funny way, they existed only so that their fate might be gambled upon.

  The market for "synthetics" removed any constraint on the size of risk associated with subprime mortgage lending. To make a billion-dollar bet, you no longer needed to accumulate a billion dollars' worth of actual mortgage loans. All you had to do was find someone else in the market willing to take the other side of the bet.

  No wonder Goldman Sachs was suddenly so eager to sell Mike Burry credit default swaps in giant, $100 million chunks, or that the Goldman Sachs bond trader had been surprisingly indifferent to which subprime bonds Mike Burry bet against. The insurance Mike Burry bought was inserted into a synthetic CDO and passed along to AIG. The roughly $20 billion in credit default swaps sold by AIG to Goldman Sachs meant roughly $400 million in riskless profits for Goldman Sachs. Each year. The deals laste
d as long as the underlying bonds, which had an expected life of about six years, which, when you did the math, implied a profit for the Goldman trader of $2.4 billion.

  Wall Street's newest technique for squeezing profits out of the bond markets should have raised a few questions. Why were supposedly sophisticated traders at AIG FP doing this stuff? If credit default swaps were insurance, why weren't they regulated as insurance? Why, for example, wasn't AIG required to reserve capital against them? Why, for that matter, were Moody's and Standard & Poor's willing to bless 80 percent of a pool of dicey mortgage loans with the same triple-A rating they bestowed on the debts of the U.S. Treasury? Why didn't someone, anyone, inside Goldman Sachs stand up and say, "This is obscene. The rating agencies, the ultimate pricers of all these subprime mortgage loans, clearly do not understand the risk, and their idiocy is creating a recipe for catastrophe"? Apparently none of those questions popped into the minds of market insiders as quickly as another: How do I do what Goldman Sachs just did? Deutsche Bank, especially, felt something like shame that Goldman Sachs had been the first to find this particular pay dirt. Along with Goldman, Deutsche Bank was the leading market maker in abstruse mortgage derivatives. Dusseldorf was playing some kind of role in the new market. If there were stupid Germans standing ready to buy U.S. subprime mortgage derivatives, Deutsche Bank should have been the first to find them.

  None of this was of any obvious concern to Greg Lippmann. Lippmann did not run Deutsche Bank's CDO business--a fellow named Michael Lamont did. Lippmann was merely the trader responsible for buying and selling subprime mortgage bonds and, by extension, credit default swaps on subprime mortgage bonds. But with so few investors willing to make an outright bet against the subprime bond market, Lippmann's bosses asked Lippmann to take one for the team: in effect, to serve as a stand-in for Mike Burry, and to make an explicit bet against the market. If Lippmann would buy credit default swaps from Deutsche Bank's CDO department, they, too, might do these trades with AIG, before AIG woke up and stopped doing them. "Greg was forced to get short into the CDOs," says a former senior member of Deutsche Bank's CDO team. "I say forced, but you can't really force Greg to do anything." There was some pushing and pulling with the people who ran his firm's CDO operations, but Lippmann found himself uncomfortably short subprime mortgage bonds.

  Lippmann had at least one good reason for not putting up a huge fight: There was a fantastically profitable market waiting to be created. Financial markets are a collection of arguments. The less transparent the market and the more complicated the securities, the more money the trading desks at big Wall Street firms can make from the argument. The constant argument over the value of the shares of some major publicly traded company has very little value, as both buyer and seller can see the fair price of the stock on the ticker, and the broker's commission has been driven down by competition. The argument over the value of credit default swaps on subprime mortgage bonds--a complex security whose value was derived from that of another complex security--could be a gold mine. The only other dealer making serious markets in credit default swaps was Goldman Sachs, so there was, in the beginning, little price competition. Supply, thanks to AIG, was virtually unlimited. The problem was demand: investors who wanted to do Mike Burry's trade. Incredibly, at this critical juncture in financial history, after which so much changed so quickly, the only constraint in the subprime mortgage market was a shortage of people willing to bet against it.

  To sell investors on the idea of betting against subprime mortgage bonds--on buying his pile of credit default swaps--Greg Lippmann needed a new and improved argument. Enter the Great Chinese Quant. Lippmann asked Eugene Xu to study the effect of home price appreciation on subprime mortgage loans. Eugene Xu went off and did whatever the second smartest man in China does, and at length returned with a chart illustrating default rates in various home price scenarios: home prices up, home prices flat, home prices down. Lippmann looked at it...and looked again. The numbers shocked even him. They didn't need to collapse; they merely needed to stop rising so fast. House prices were still rising, and yet default rates were approaching 4 percent; if they rose to just 7 percent, the lowest investment-grade bonds, rated triple-B-minus, went to zero. If they rose to 8 percent, the next lowest-rated bonds, rated triple-B, went to zero.

  At that moment--in November 2005--Greg Lippmann realized that he didn't mind owning a pile of credit default swaps on subprime mortgage bonds. They weren't insurance; they were a gamble; and he liked the odds. He wanted to be short.

  This was new. Greg Lippmann had traded bonds backed by various consumer loans--auto loans, credit card loans, home equity loans--since 1991, when he had graduated from the University of Pennsylvania and taken a job at Credit Suisse. He'd never before been able to sell them short, because they were impossible to borrow. The only choice he and every other asset-backed bond trader ever had to make was whether to like them or to love them. There was never any point in hating them. Now he could, and did. But hating them set him apart from the crowd--and that represented, for Greg Lippmann, a new career risk. As he put it to others, "If you're in a business where you can do only one thing and it doesn't work out, it's hard for your bosses to be mad at you." It was now possible to do more than one thing, but if he bet against subprime mortgage bonds and was proven wrong, his bosses would find it easy to be mad at him.

  In the righteous spirit of a man bearing an inconvenient truth, Greg Lippmann, a copy of "Shorting Subprime Mezzanine Tranches" tucked under his arm, launched himself at the institutional investing public. He may have begun his investigation of the subprime mortgage market in the spirit of a Wall Street salesman, searching less for the truth than for a persuasive-sounding pitch. Now, shockingly, he thought he had an ingenious plan to make customers rich. He'd charge them fat fees to get in and out of their credit default swaps, of course, but these would prove trivial compared to the fortunes they stood to make. He was no longer selling; he was dispensing favors. Behold. A gift from me to you.

  Institutional investors didn't know what to make of him, at least not at first. "I think he has some kind of narcissistic personality disorder," said one money manager who heard Lippmann's pitch but did not do his trade. "He scared the shit out of us," said another. "He comes in and describes this brilliant trade. It makes total sense. To us the risk was, we do it, it works, then what? How do we get out? He controls the market; he may be the only one we can sell to. And he says, 'You have no way out of this swimming pool but through me, and when you ask for the towel I'm going to rip your eyeballs out.' He actually said that, that he was going to rip our eyeballs out. The guy was totally transparent."

  They loved it, in a way, but decided they didn't want to experience the thrill of eyeball removal. "What worked against Greg," this fund manager said, "was that he was too candid."

  Lippmann faced the usual objections any Wall Street bond customer voiced to any Wall Street bond salesmen--If it's such a great trade, why are you offering it to me?--but other, less usual ones, too. Buying credit default swaps meant paying insurance premiums for perhaps years as you waited for American homeowners to default. Bond market investors, like bond market traders, viscerally resisted any trade that they had to pay money to be in, and instinctively sought out trades that paid them just for showing up in the morning. (One big bond market investor christened his yacht Positive Carry.) Trades where you fork over 2 percent a year just to be in them were anathema. Other sorts of investors found other sorts of objections. "I can't explain credit default swaps to my investors" was a common response to Greg Lippmann's pitch. Or "I have a cousin who works at Moody's and he says this stuff [subprime mortgage bonds] is all good." Or "I talked to Bear Stearns and they said you were crazy." Lippmann spent twenty hours with one hedge fund guy and thought he had him sold, only to have the guy call his college roommate, who worked for some home builder, and change his mind.

  But the most common response of all from investors who heard Lippmann's argument was, "I'm convin
ced. You're right. But it's not my job to short the subprime market."

  "That's why the opportunity exists," Lippmann would reply. "It's nobody's job."

  It wasn't Lippmann's, either. He was meant to be the toll booth, taking a little from buyers and sellers as they passed through his trading books. He was now in a different, more opinionated relationship to his market and his employer. Lippmann's short position may have been forced upon him, but by the end of 2005 he'd made it his own, and grown it to a billion dollars. Sixteen floors above him inside Deutsche Bank's Wall Street headquarters, several hundred highly paid employees bought subprime mortgage loans, packaged them into bonds, and sold them off. Another group packaged the most repellent, unsalable tranches of those bonds, and CDSs on the bonds, into CDOs. The bigger Lippmann's short position grew, the greater the implicit expression of contempt for these people and their industry--an industry quickly becoming Wall Street's most profitable business. The running cost, in premiums Lippmann paid, was tens of millions of dollars a year, and his losses looked even bigger. The buyer of a credit default swap agreed to pay premiums for the lifespan of the underlying mortgage bond. So long as the underlying bonds remained outstanding, both buyer and seller of credit default swaps were obliged to post collateral, in response to their price movements. Astonishingly, the prices of subprime mortgage bonds were rising. Within a few months, Lippmann's credit default swap position had to be marked down by $30 million. His superiors repeatedly asked him to explain why he was doing what he was doing. "A lot of people wondered if this was the best use of Greg's time and our money," said a senior Deutsche Bank official who watched the growing conflict.