the man who crashed the world

Another entertaining piece of journalism by Michael Lewis, this time reporting from the nucleus of the financial crisis. The premise? Buried deep within the world’s largest insurance company lay the other side of the global bet on real estate and perpetual growth. Lewis interviewed the FP traders accused of underwriting the risky default swaps that nearly destroyed the world economy while siphoning-off juicy bonuses from the comfort of their gated Connecticut suburbs. What he turned up was a far bigger fish and an interesting chorus: “while they disagreed on this and that, they all were fairly certain that if it hadn’t been for A.I.G. F.P. the subprime-mortgage machine might never have been built, and the financial crisis might never have happened…”

The Man Who Crashed the World
Michael Lewis
in Vanity Fair, August 2009

Almost a year after A.I.G.’s collapse, despite a tidal wave of outrage, there still has been no clear explanation of what toppled the insurance giant. The author decides to ask the people involved—the silent, shell-shocked traders of the A.I.G. Financial Products unit—and finds that the story may have a villain, whose reign of terror over 400 employees brought the company, the U.S. economy, and the global financial system to their knees.


Six months ago, I received an odd phone call from a man named Jake DeSantis at A.I.G. Financial Products—the infamous unit of the doomed insurance company, staffed by expensively educated, highly paid traders, whose financial ineptitude is widely suspected of costing the U.S. taxpayer $182.5 billion and counting. At the time A.I.G. F.P.’s losses were reported, it became known that a handful of traders in this curious unit had sold trillions of dollars of credit-default swaps (essentially unregulated insurance policies) on piles of U.S. subprime mortgages, but its employees hadn’t yet become the leading examples of Wall Street greed. And so this was before Jake DeSantis and his colleagues found themselves suburban-Connecticut outcasts, before their first death threats, before the House of Representatives passed a bill because of them (taxing 90 percent of their large bonuses), before New York attorney general Andrew Cuomo announced he was going after their paychecks, and before Iowa senator Charles Grassley said that A.I.G.’s leaders should follow the Japanese example and “either do one of two things, resign or go commit suicide.”

DeSantis turned out to be a friend of a friend. He’d called because he didn’t know anyone else “in the media.” As a type he was instantly recognizable: a “quant,” a numbers guy who was allowed to take financial risks because of his superior math skills, but who had no taste for company politics or public exposure. He’d grown up in the Midwest, the son of schoolteachers, and discovered Wall Street as a scholarship student at M.I.T. The previous seven years he’d spent running A.I.G. F.P.’s profitable stock-market-related trades. He wasn’t looking for me to write about him or about A.I.G. F.P. He just wanted to know why the public perception of what had happened inside his unit, and the larger company, was so different from the private perception of the people inside it, who actually knew what had happened. The idea that the employees of A.I.G. F.P. had conspired to maximize their short-term gains at the company’s longer-term expense, for instance. He and the other traders had been required to defer about half of their pay for years, and intertwine their long-term interests with their firm’s. The people who lost the most when A.I.G. F.P. went down were the employees of A.I.G. F.P.: DeSantis himself had just watched more than half of what he’d made over the previous nine years vanish. The incentive system at A.I.G. F.P., created in the mid-1990s, wasn’t the short-term-oriented racket that helped doom the Wall Street investment bank as we knew it. It was the very system that U.S. Treasury secretary Timothy Geithner, among others, had proposed as a solution to the problem of Wall Street pay.

Even more oddly, the public explanation of A.I.G.’s failure focused on the credit-default swaps sold by traders at A.I.G. F.P., when A.I.G.’s problems were clearly broader. There was the mortgage-insurance unit in North Carolina, United Guaranty, that had taken on all sorts of silly risks in the past two years, lost several billion dollars, and replaced their C.E.O. There were the fund managers at A.I.G., the parent company, who had blown nearly $50 billion on trades in subprime mortgages—that is, they had lost more than A.I.G. F.P., whose losses stood around $45 billion. And there was a pattern: all of this stuff had happened since 2005, after an accounting scandal forced C.E.O. Maurice “Hank” Greenberg to resign. Greenberg, who had headed A.I.G. since 1968, was a bullying, omnipotent ruler—one of those bosses who did not so much build a company as tailor it to his character and render it incapable of being run by anyone else. After he was forced out, Greenberg said, “The new management wanted to prove that they could continue to grow without former management” and so turned a blind eye to all sorts of risks. So how come most of the senior management at A.I.G. was left in place by the U.S. Treasury after the bailout? Why were officials, both public and private, so intent on leading others to believe all the losses at A.I.G. had been caused by a few dozen traders in this fringe unit in London and Connecticut?

I had no idea, was busy doing other things, and had no special interest in Jake DeSantis’s predicament. I listened politely, made my excuses—and went back to whatever it was I’d been doing. But then, on March 19, the new C.E.O. of A.I.G., Edward Liddy, went to Washington to testify. The story broke—or, rather, rebroke, as it had been reported two weeks earlier, without stirring much notice—that A.I.G. F.P. had just shelled out $450 million in bonuses to the 400 employees of A.I.G. F.P., including to Jake DeSantis. It must have been an otherwise slow news day because all hell broke loose, in a way it hadn’t before and hasn’t since in this financial crisis. The perception was that the very same people who had made these insane, greed-driven decisions that might cost the U.S. taxpayer $182.5 billion were still paying themselves big bucks! An exchange between C.E.O. Liddy and Florida congressman Alan Grayson captured the spirit of that moment:

grayson: Mr. Liddy, you said before that there’s 20 or 25 people who were involved in the credit default business. What are their names, please?

liddy: I don’t have their names at my disposal, sir. grayson: Well, I’m sure you remember a few of the names. I mean, they did cause your company to crash.

liddy: You know, I’ve been at the company, as you know, for six months. I don’t know all the people that were in AIG F.P., and many of them are gone.

grayson: Well, there or gone, it doesn’t really matter. I want to know who they are. Names, please.…

liddy: If it’s possible to provide you the names, we will. We will cooperate with you.

grayson: That’s good, but I want to know the names that you know right now.

liddy: I don’t know them, sir.

grayson: Not a single one. You’re talking about a group, a small group of people who caused your company to lose $100 billion, as you sit here today, you can’t give me one single name.

liddy: The single name I would give you is Joseph Cassano, who ran …

grayson: That’s a good start. You already gave that name. Give me another name.

liddy: I just don’t know them. I do not know those names. I don’t have them all at my command.

grayson: Well, how can you propose to solve the problems of the company that you’re now running if you don’t know the names of the people who caused that problem? … I would expect you’d at least know more than one name. How about two names? Give us one more name.

liddy: I’m just not going to do that, sir, because that will provide—that’ll be the—that could be a list of people that we could do—individuals who want to do damage to them could do that. It’s just not …

grayson: Well, listen, these same people could now be working right now today at Citibank. Is it more important to protect them, the ones who caused the $100 billion loss, or protect us? Which is more important to you right now?

For a brief moment you had a glimpse of how harshly financial people might be treated if Wall Street ever lost its political influence. Just days before, Larry Summers had gone on the morning talk shows to explain that a contract is a contract and the government couldn’t just go in and void it and take back A.I.G.’s paychecks, but that “every legal step possible to limit those bonuses is being taken by Secretary Geithner and by the Federal Reserve System.” Then Obama himself went out of his way to denounce the greed at A.I.G. F.P. and say he was looking for a way to get the bonus money back—and even that failed to slake the public anger. “On A.I.G.,” a journalist asked Obama at a press conference, “why did you wait—why did you wait days to come out and express that outrage? It seems like the action is coming out of New York and the attorney general’s office. It took you days to come public with Secretary Geithner and say, Look, we’re outraged. Why did it take so long?”

“It took us a couple of days because I like to know what I’m talking about before I speak,” Obama said testily. “All right?”

It’s unlikely that he actually did know what he was talking about, except in the broadest outlines. Nor, for that matter, did the people who had engineered the bailout. How could they? At no point did anyone from the U.S. Treasury or the U.S. Congress, or any of the various New York State authorities that had gotten involved, call them up, much less visit A.I.G. F.P.—as, say, someone might who was genuinely curious to know what, exactly, had happened there. Not even A.I.G. C.E.O. Ed Liddy had bothered to make the drive from Manhattan to Wilton, Connecticut, where many of the offending trades had been done, and most of the offending bonuses were being paid, to ask questions of the people still on the scene—people who could have told him a great deal about what had happened and why. Everyone seemed to be operating on whatever they read in the newspapers—and the people inside A.I.G. F.P., who had the best view of the action, did not appear to be talking to reporters. Depending on which account you read, you thought they had lost $40 billion, or $100 billion, or $152 billion. They had done this by selling credit-default swaps on subprime-mortgage bonds—which is to say they had insured Goldman Sachs, Deutsche Bank, Merrill Lynch, and the rest against Americans with weak credit histories defaulting on their mortgages. But why? Apparently, because they were greedy: the premiums they took in from the insurance allowed them to pay themselves big bonuses, which they’d grown so accustomed to that they now were reduced to stealing from the U.S. taxpayer. And that, it seemed, was that.

The day after Liddy’s testimony, I got another call from Jake DeSantis. (I was still the only person “in the media” with whom he felt any connection.) He was upset. He’d turned down offers of more money from other people. He’d stayed only because the company had begged him to help clean up the mess: the bonus he was paid was the result of profits he had generated by selling off his trades in global equities—profits which almost surely would have been losses had he not hung around. He’d had nothing to do with the trades that lost money; the handful of people who’d known about them, when they happened, were long gone, and even they had been guided by a certain understandable logic. Now A.I.G.’s new leader, who had accepted these bonuses and run them by both the Treasury and the Federal Reserve, flies down to Washington and tells the world that he found the bonuses “distasteful.”

“You go to church and you go to soccer practice and people look at you funny,” said DeSantis. “This is changing people’s views on who I am as a person.” He’d decided to resign, write a letter to Liddy, and, as he put it, “release it to the media.”

It sounded like the sort of thing that might work on a TV show. “What does that mean: ‘Release it to the media’?” I asked. That, he said, was why he’d called me: he thought I knew how. Having no clue, I put him in touch with the editor of the New York Times op-ed page, who published Jake DeSantis’s letter of resignation on March 25 at the top of his page under the headline: dear a.i.g., i quit! In it Jake repeated what he’d told me, offered a bit of his life story, and confessed the size of his after-tax bonus ($742,006.40). He also explained that he had for a year spent up to 14 hours a day helping to dismantle the company and that he and the others who had nothing to do with the losses had agreed to do so based on the promise their contracts would be honored.

It’s never easy to prove that a piece of writing causes anything, but Jake’s letter was an instant sensation. Bits of it were reprinted in major publications around the world. Within a few days it was the most sent, most blogged, and most read item on the Times’s Web site, and remained so for the entire month. Three point nine million browsers clicked on it and read it, and the tone of the public discussion changed. New York attorney general Andrew Cuomo stopped saying he intended to hound the millions paid to the people who worked at A.I.G. F.P., and started saying he was more interested in the $12.9 billion A.I.G. had paid out to Goldman Sachs, and others, to cover the massive bets against U.S. subprime mortgages that they had made with A.I.G. The House tax bill stalled in the Senate. I didn’t really know Jake DeSantis, but I thought, That was just incredibly brave. He stepped out alone in front of the mob and compelled it to disband, at least for the moment. But I never heard back from him. After a few days of not being able to open a newspaper or go to a Web site without seeing some reference to Jake DeSantis and his letter, I phoned him. “Oh hey,” he said cheerily. “They published my letter.”

No shit, Jake.

“Has it worked out O.K.?” I asked.

“Oh yeah,” he said, “but I had to move my family out of our house.”

He had woken up the morning his piece ran to find media trucks jamming the end of his driveway. He took his family out back through the woods—“We live in the middle of nowhere”—and secreted them at a friend’s house. “I’ve been going back and standing on the porch down the road and pretending to be a gawking neighbor,” he said, “but they’re all still there blocking the end of the driveway. They’re waiting for me to come back, I guess.” His voice mail, he said, was also jammed. “All these media people keep calling,” he said. “Like who?” I asked. “We don’t watch TV, so I don’t know who they are,” he said. I pressed him. “Well, there’s one guy who has been calling a bunch. Matt Lauer. I don’t know who he’s with.” The only caller he could completely identify was Katie Couric: “She called our mayor personally and tried to butter her up to get her to tell her where I am,” he said. (A Couric producer says, “Katie placed a brief call to the mayor, expressed interest in the interview, and nothing further happened.”) He suspected, probably rightly, that the media wanted him to play the role of the greedy Wall Street trader who had stolen millions and now claimed to feel misunderstood. “O.K.,” he said. “I can do this and probably not make an ass of myself. But I can do nothing and not make an ass of myself. I’ll stick with that.”

With that, A.I.G. F.P. went dark again, which, I now realized, was a shame. DeSantis had established, sort of, what the people in his unit didn’t do. He’d left unexplained what exactly they did do.

Here is an amazing fact: nearly a year after perhaps the most sensational corporate collapse in the history of finance, a collapse that, without the intervention of the government, would have led to the bankruptcy of every major American financial institution, plus a lot of foreign ones, too, A.I.G.’s losses and the trades that led to them still haven’t been properly explained. How did they happen? Unlike, say, Bernie Madoff’s pyramid scheme, they don’t seem to have been raw theft. They may have been an outrageous departure from financial norms, but, if so, why hasn’t anyone in the place been charged with a crime? How did an insurance company become so entangled in the sophisticated end of Wall Street and wind up the fool at the poker table? How could the U.S. government simply hand over $54 billion in taxpayer dollars to Goldman Sachs and Merrill Lynch and all the rest to make good on the subprime insurance A.I.G. F.P. had sold to them—especially after Goldman Sachs was coming out and saying that it had hedged itself by betting against A.I.G.? Since I had him on the phone I asked Jake DeSantis for what Congressman Grayson had asked Edward Liddy: names. He obligingly introduced me to his colleagues in London and Connecticut, and they walked me through what had happened—all of them speaking to someone from the outside for the first time. All, for obvious reasons, were terrified of seeing their names in print, and asked not to be mentioned by name. That was fine by me, as their names are not what’s interesting. What’s interesting is their point of view on the event closest to the center of the financial crisis. For while they disagreed on this and that, they all were fairly certain that if it hadn’t been for A.I.G. F.P. the subprime-mortgage machine might never have been built, and the financial crisis might never have happened.

The Soul of a New Machine

A.I.G. F.P. was created back in 1987 by refugees from Drexel Burnham, led by a trader named Howard Sosin, who claimed to have a better model to trade and value interest-rate swaps. Nineteen-eighties financial innovation had all sorts of consequences, but one of them was a boom in the number of deals between big financial firms that required them to take each other’s credit risks. Interest-rate swaps—in which a party swaps a stream of income from a floating rate of interest for one from a fixed rate of interest—was one such innovation.

Once upon a time Chrysler issued a bond through Morgan Stanley, and the only people who wound up with credit risk were the investors who had bought the Chrysler bond. Now Chrysler might sell its bonds and simultaneously enter into a 10-year interest-rate-swap transaction with Morgan Stanley—and just like that Chrysler and Morgan Stanley were exposed to each other. If Chrysler went bankrupt, its bondholders obviously lost; depending on the nature of the swap and the movement of interest rates, Morgan Stanley might lose, too. If Morgan Stanley went bust, Chrysler along with anyone else who had done interest-rate swaps with Morgan Stanley stood to suffer. Financial risk had been created, out of thin air, and it begged to be either honestly accounted for or disguised.

Enter Sosin, with his supposedly new and improved interest-rate-swap model (even though Drexel Burnham was not at the time a market leader in interest-rate swaps).

There was a natural role for a blue-chip corporation with the highest credit rating to stand in the middle of swaps and long-term options and the other risk-spawning innovations. The traits required of this corporation were that it not be a bank—and thus subject to bank regulation and the need to reserve capital against the risky assets—and that it be willing and able to bury exotic risks on its balance sheet. There was no real reason that company had to be A.I.G.; it could have been any AAA-rated entity with a huge balance sheet. Berkshire Hathaway, for instance, or General Electric. A.I.G. just got there first.

In a financial system that was rapidly generating complicated risks, A.I.G. F.P. became a huge swallower of those risks. In the early days it must have seemed as if it was being paid to insure against events extremely unlikely to occur—how likely was it that all sorts of companies and banks all over the globe would go bust at the same time? Its success bred imitators: Zurich Re F.P., Swiss Re F.P., Credit Suisse F.P., Gen Re F.P. All of these places were central to what happened in the last two decades; without them the new risks being created would have had no place to hide, but would have remained in full view of bank regulators. All of these places have been washed away by the general nausea now felt in the presence of complicated financial risks, but there was a moment when their existence seemed cartographically necessary to the financial world. And A.I.G. F.P. was the model for them all.

The division’s first 15 years were consistently, amazingly profitable—there wasn’t the first hint that it might be running risks that would cause it to lose money, much less cripple its giant parent. Its traders were able to claim that they were “hedged,” and even if the term was misleading, they never sold exactly the same thing as the thing they had bought—there was always some slight difference. The risks it ran were probably trivial in relation to its capital, because the risks that the financial system wanted to lay off on it were, in fact, not terribly risky. One indication of this is that, even in the middle of the calamity, the 95 percent of A.I.G. F.P. that had nothing to do with subprime-mortgage bonds continued to generate profits. By 2001, A.I.G. F.P. could be counted on to generate $300 million a year, or 15 percent of A.I.G.’s profits.

Meanwhile, the people who worked at A.I.G. F.P. got rich. Exactly how rich is hard to say, but there are plenty of hints. One is that a company lawyer—a mere lawyer!—took home a $25 million bonus at the end of one year. Another is that in 2005, when Howard Sosin and his wife divorced, she received more than $40 million of an estate valued at $168 million—and Sosin had left A.I.G. in 1993, receiving $182 million from the company! He had been replaced that year as C.E.O. by a gentler soul named Tom Savage, who had allowed Hank Greenberg to take some of the sugar out of F.P., but even then the small band of traders had, arguably, a sweeter deal than any money managers in the world. The typical hedge fund kept 20 percent of profits; the traders at A.I.G. F.P. kept 30 to 35 percent. The typical hedge fund or private-equity fund has to schlep around and raise money all the time, and post collateral with big Wall Street firms for all the trades they do. The traders at A.I.G. F.P. had essentially unlimited capital on tap from the parent company, along with the AAA rating, rent-free. For the people who worked there, A.I.G. F.P. was a financial miracle. They were required to leave 50 percent of their bonuses in the company, but they were happy to do so; many of them, viewing it as the best way to grow their own savings, invested far more than the minimum back in the company. When it collapsed, the employees lost more than $500 million of their own money.

How and why their miracle became a catastrophe, A.I.G. F.P.’s traders say, is a complicated story, but it begins simply: with a change in the way decisions were made, brought about by a change in its leadership. At the end of 2001 its second C.E.O., Tom Savage, retired, and his former deputy, Joe Cassano, was elevated. Savage is a trained mathematician who understood the models used by A.I.G. traders to price the risk they were running—and thus ensure that they were fairly paid for it. He enjoyed debates about both the models and the merits of A.I.G. F.P.’s various trades. Cassano knew a lot less math and had much less interest in debate.

It’s impossible to deliver the full flavor of a man’s character without talking to him, and relying instead upon a bunch of people who remain afraid of seeing their names in print. That Joe Cassano is the son of a police officer and was a political-science major at Brooklyn College seems, in retrospect, far less relevant than that he’d spent most of his career, both at Drexel and A.I.G. F.P., in the back office, doing operations. Across A.I.G. F.P. the view of the boss was remarkably consistent: a guy with a crude feel for financial risk but a real talent for bullying people who doubted him. “A.I.G. F.P. became a dictatorship,” says one London trader. “Joe would bully people around. He’d humiliate them and then try to make it up to them by giving them huge amounts of money.”

“One day he got me on the phone and was pissed off about a trade that had lost money,” says a Connecticut trader. “He said, ‘When you lose money it’s my fucking money. Say it.’ I said, ‘What?’ ‘Say “Joe, it’s your fucking money!”’ So I said, ‘It’s your fucking money, Joe.’”

“The culture changed,” says a third. “The fear level was so high that when we had these morning meetings you presented what you did not to upset him. And if you were critical of the organization, all hell would break loose.” Says a fourth, “Joe always said, ‘This is my company. You work for my company.’ He’d see you with a bottle of water. He’d come over and say, ‘That’s my water.’ Lunch was free, but Joe always made you feel he had bought it.” And a fifth: “Under Joe the debate and discussion that was common under Tom [Savage] ceased. I would say what I’m saying to you. But with Joe over my shoulder as the audience.” A sixth: “The way you dealt with Joe was to start everything by saying, ‘You’re right, Joe.’”

According to traders, Cassano was one of those people whose insecurities manifested themselves in a need for obedience and total control. “One day he came in and saw that someone had left the weights on the Smith machine, in the gym,” says a source in Connecticut. “He was literally walking around looking for people who looked buff, trying to find the guy who did it. He was screaming, ‘Who left the fucking weight on the fucking Smith machine? Who left the fucking weight on the fucking Smith machine?’” If that rings a bell it may be because you read The Caine Mutiny and recall Captain Queeg scouring the ship to find out who had stolen the strawberries. Even by the standards of Wall Street villains, whose character flaws wind up being exaggerated to fit the crime, Cassano was a cartoon despot.

Oddly, he was as likely to direct his anger at profitable traders as at unprofitable ones—and what caused him to become angry was the faintest whiff of insurrection.

Even more oddly, his anger had no obvious effect on the recipient’s paycheck; a trader might find himself routinely abused by his boss and yet delighted by his year-end bonus, determined by that same boss. Every one of the people I spoke with admitted that the reason they hadn’t taken a swing at Joe Cassano, before walking out the door, was that the money was simply too good. A man who valued loyalty and obedience above all other traits had not any tools to command them except money. Money worked, but only up to a point. If you were going to be on the other side of a trade from Goldman Sachs, you had better know what, exactly, Goldman Sachs was up to. A.I.G. F.P. could attract extremely bright people, whose success depended on precision of both calculation and judgment. It was now run, roughly, by a man who didn’t fully understand all the calculations and whose judgment was clouded by his insecurity. The few people willing to question that judgment wound up quitting the firm. Left behind were people who more or less accommodated Cassano. “If someone is a complete asshole,” one of them puts it to me, “you seek his approval in a way you don’t if he’s a nice guy.”

All of which raises an obvious question: Who put a man like Joe Cassano in charge of such an enterprise as A.I.G. F.P.? The simple answer is Hank Greenberg, the C.E.O. of A.I.G.; the more complicated one is A.I.G. F.P.’s board, consisting of many smart people, including Harvard economist Martin Feldstein. “Tom Savage proposed Joe to replace him,” says Greenberg, “and we had no reason to think he wasn’t able to do the job.”

A.I.G. F.P.’s employees for their part suspect that the only reason Greenberg promoted Cassano was that he saw in him a pale imitation of his own tyrannical self and felt he could control him. “So long as Greenberg was there, it worked,” says one trader, “because he watched everything Joe did. After the Nikkei collapsed [in the 1990s], a trader in Japan lost 20 million. Greenberg personally flew to Tokyo and took him into a room and grilled him until he was satisfied.” In March 2005, however, Eliot Spitzer forced Greenberg to resign. And, as one trader puts it, “the new guys running A.I.G. had no idea.” They thought the money machine ran on its own, and Cassano did nothing to discourage the view. By 2005, A.I.G. F.P. was indeed, in effect, his company.

But even here the story’s messier than its broad outlines. For a start, the guy who had the most invested in A.I.G. F.P. was Joe Cassano. Cassano had been paid $38 million in 2007, but left $36.75 million of that inside the firm. His financial interest in A.I.G. F.P. struck those who worked for him as secondary to his psychological investment: the firm was, by all accounts, Cassano’s sole source of self-worth, its success his lone status symbol. He wore crappy clothes, drove a crappy car, and spent all of his time at the office. He had made huge piles of money ($280 million!), but so far as anyone could tell he didn’t spend any of it. “Joe wasn’t a trader and now he wasn’t a risktaker, in his personal life,” says one of the traders. “With the money he didn’t have in the company he bought Treasury bonds.” He had no children, no obvious social ambition; his status concerns seemed limited to his place in the global financial order. He entertained a notion of himself as the street-smart guy who had triumphed over his social betters—which of course implied that he wasn’t quite sure that he had. “Joe had Goldman envy,” one trader tells me—which was strange, as Cassano’s brother and sister both worked for Goldman Sachs. “His whole life was F.P.,” another trader says. “Without F.P. he had nothing.” That was another reason, in addition to fear, that the highly educated, highly intelligent people who worked for Joe Cassano were slow to question whatever he was doing: he was the last person, they assumed, who would blow the place up.

The more subtle change inside A.I.G. F.P. occurred not long after Cassano assumed control. In 1998, A.I.G. F.P. had entered the new market for credit-default swaps: it sold insurance to banks against the risk of defaults by huge numbers of investment-grade public corporations. As Gillian Tett tells it in her new book, Fool’s Gold, bankers at J. P. Morgan, having invented credit-default swaps, went looking for an AAA-rated company to assume the bulk of the risk associated with them, and discovered A.I.G. The relationship began innocently enough, by Wall Street standards. The risk in these early deals was indeed small: it was unlikely that large numbers of investment-grade companies in different countries and different industries would default on their debt at the same time. (Even now A.I.G. F.P.’s $450 billion portfolio of corporate credit-default swaps, which dwarfs the $75 billion portfolio of subprime-mortgage credit-default swaps, has avoided losses.) But it made explicit what until then had only been implicit: A.I.G. F.P. was the most receptive dumping ground for new risks created by big Wall Street firms.

And in the early 2000s, the big Wall Street firms performed this fantastic bait and switch in two stages. Stage One was to apply technology that had been dreamed up to re-distribute corporate credit risk to consumer credit risk. The banks that used A.I.G. F.P. to insure piles of loans to IBM and G.E. now came to it to insure much messier piles that included credit-card debt, student loans, auto loans, prime mortgages, and just about anything else that generated a cash flow. “The problem,” as one trader puts it, “is that something else came along that we thought was the same thing as what we’d been doing.” Because there were many different sorts of loans, to different sorts of people, the logic applied to corporate credit seemed to apply to this new pile of debt: it was sufficiently diverse that it was unlikely to all go bad at once. But then, these piles, at least at first, contained almost no subprime-mortgage loans.

Toward the end of 2004, that changed dramatically—but just how dramatically A.I.G. F.P. was extremely slow to realize. In the run-up to the financial crisis there were several moments when an intelligent, disinterested observer might have realized that the system was behaving strangely. Maybe the most obvious of these was the effects of U.S. monetary policy on borrowing and lending. The combination of the dot-com bust and the 9/11 attacks had led Alan Greenspan to pump money into the system, and to lower interest rates. In June 2004 the Fed began to contract the money supply, and interest rates rose. In a normal economy, when interest rates rise, consumer borrowing falls—and in the normal end of the U.S. economy that happened: from June 2004 to June 2005 prime-mortgage lending fell by half. But in that same period subprime lending doubled—and then doubled again. In 2003 there had been a few tens of billions of dollars of subprime-mortgage loans. From June 2004 until June 2007, Wall Street underwrote $1.6 trillion of new subprime-mortgage loans and another $1.2 trillion of so-called Alt-A loans—loans which for some reason or another can be dicey, usually because the lender did not require the borrower to supply him with the information typically required before making a loan. The subprime sector of the financial economy clearly was responding to different signals than the others—and the result was booming demand for housing and a continued rise in house prices. Perhaps the biggest reason for this was that the Wall Street firms packaging the loans into bonds had found someone to insure against what turned out to be the rather high risk that they’d go bad: Joe Cassano.

A.I.G. F.P. was already insuring these big, diversified, AAA-rated piles of consumer loans; to get it to insure subprime mortgages was only a matter of pouring more and more of the things into the amorphous, unexamined piles. They went from being 2 percent subprime mortgages to being 95 percent subprime mortgages. And yet no one at A.I.G. said anything about it—not C.E.O. Martin Sullivan, not Joe Cassano, not Al Frost, the guy in A.I.G. F.P.’s Connecticut office in charge of selling his firm’s credit-default-swap services to the big Wall Street firms. The deals, by all accounts, were simply rubber-stamped by Cassano and then again by A.I.G. brass—and, on the theory that this was just more of the same, no one paid them special attention. It’s hard to know what Joe Cassano thought and when he thought it, but the traders inside A.I.G. F.P. are certain that neither Cassano nor the four or five people overseen directly by him, who worked in the unit that made the trades, realized how completely these piles of consumer loans had become, almost exclusively, composed of subprime mortgages.

The Big Switch

Gene Park worked in the Connecticut office and sat close enough to the credit-default-swap traders to have a general idea of what they were up to. In mid-2005 he’d read a front-page story in The Wall Street Journal about the mortgage lender New Century. He noted how high its dividend was and thought he might like to buy some of its stock for himself. As he dug into New Century, however, Park saw that it owned all these subprime mortgages—and he could see from its own statements that the quality of the loans was frightening. Just after that he got a phone call from a penniless, jobless old college friend who had been offered a package of loans to buy a house he couldn’t afford. At the same time, Park saw Al Frost announcing new credit-default-swap deals at an alarming rate. A year before, Frost might have had one half-billion-dollar deal each month; now he was doing 20, all on piles of consumer loans. “We were doing every single deal with every single Wall Street firm, except Citigroup,” says one trader. “Citigroup decided it liked the risk and kept it on their books. We took all the rest.” When traders asked Frost why Wall Street was suddenly so eager to do business with A.I.G., says a trader, “he would explain that they liked us because we could act quickly.” Park put two and two together and guessed that the nature of these piles of consumer loans insured by A.I.G. F.P. was changing, that they contained a lot more subprime mortgages than anyone knew, and that if U.S. homeowners began to default in sharply greater numbers A.I.G. didn’t have anywhere near the capital required to cover the losses. He told Andy Forster, Cassano’s right-hand man in London, who brought this up at a meeting, but Cassano dismissed the concerns as overblown.

Oddly, this dramatic increase in the amount of risk A.I.G. F.P. was assuming came at exactly the moment when it lost the reason for its existence. The day after Hank Greenberg was forced to resign, in March 2005, the credit-rating agencies downgraded A.I.G. from AAA to AA. The AAA rating was the competitive advantage; without it, the natural course of action would have been to close or dramatically shrink A.I.G. F.P.’s business. Instead, Cassano grew it.

Toward the end of 2005, Cassano promoted Al Frost, then went looking for someone to replace him as the ambassador to Wall Street’s subprime-mortgage-bond desks. As a smart quant who understood abstruse securities, Gene Park was a likely candidate. That’s when Park decided to examine more closely the loans that A.I.G. F.P. had insured. He suspected Joe Cassano didn’t understand what he had done, but even so Park was shocked by the magnitude of the misunderstanding: these piles of consumer loans were now 95 percent U.S. subprime mortgages. Park then conducted a little survey, asking the people around A.I.G. F.P. most directly involved in insuring them how much subprime was in them. He asked Gary Gorton, a Yale professor who had helped build the model Cassano used to price the credit-default swaps. Gorton guessed that the piles were no more than 10 percent subprime. He asked a risk analyst in London, who guessed 20 percent. He asked Al Frost, who had no clue, but then, his job was to sell, not to trade. “None of them knew,” says one trader. Which sounds, in retrospect, incredible. But an entire financial system was premised on their not knowing—and paying them for their talent!

By the time Joe Cassano invited Gene Park to London for the meeting in which he would be “promoted” to the job of creating even more of these ticking time bombs, Park knew he wanted no part of it. He announced that, if he was made to take the job, he’d quit. (Had he taken it he would now be a magazine cover.)

This, naturally, infuriated Joe Cassano, who, says one trader, thought Park was being lazy, dreaming up reasons not to do the deals that would require work. Confronted with the new development—his company was insuring not consumer credit generally but subprime mortgages—Cassano didn’t blink. He simply claimed that the fact was irrelevant: for the bonds to default, U.S. house prices had to fall, and Cassano didn’t believe house prices could ever fall everywhere in the country at once. After all, Moody’s and S&P still rated this stuff AAA!

Still, Cassano agreed to meet with all the big Wall Street firms and discuss the logic of their deals—to investigate how a bunch of shaky loans could be transformed into AAA-rated bonds. Together with Park and a few others, Cassano set out on a series of meetings with Morgan Stanley, Goldman Sachs, and the rest—all of whom argued how unlikely it was for housing prices to fall all at once. “They all said the same thing,” says one of the traders present. “They’d go back to historical real-estate prices over 60 years and say they had never fallen all at once.” (The lone exception, he said, was Goldman Sachs. Two months after their meeting with the investment bank, one of the A.I.G. F.P. traders bumped into the Goldman guy who had defended the bonds, who said, Between you and me, you’re right. These things are going to blow up.) The A.I.G. F.P. executives present were shocked by how little actual thought or analysis seemed to underpin the subprime-mortgage machine: it was simply a bet that U.S. home prices would never fall. Once he understood this, Joe Cassano actually changed his mind. He agreed with Gene Park: A.I.G. F.P. shouldn’t insure any more of these deals. And at the time it didn’t really seem like all that big of an issue. A.I.G. F.P. was generating around $2 billion year in profits. At the peak, the entire credit-default-swap business contributed only $180 million of that. He was upset, it seemed, mainly that he had been successfully contradicted.

What no one realized was that it was too late. A.I.G. F.P.’s willingness to assume the vast majority of the risk of all the subprime-mortgage bonds created in 2004 and 2005 had created a machine that depended for its fuel on subprime-mortgage loans. “I’m convinced that our input into the system led to a substantial portion of the increase in housing prices in the U.S. We facilitated a trillion dollars in mortgages,” says one trader. “Just us.” Every firm on Wall Street was making fantastic sums of money from this machine, but for the machine to keep running the Wall Street firms needed someone to take the risk. When Gene Park informed them that A.I.G. F.P. would no longer do so—Hello, my name is Gene Park and I’m closing down your business—he became the most hated man on Wall Street.

The big Wall Street firms solved the problem by taking the risk themselves. The hundreds of billions of dollars in subprime losses suffered by Merrill Lynch, Morgan Stanley, Lehman Brothers, Bear Stearns, and the others were hundreds of billions in losses that might otherwise have been suffered by A.I.G. F.P. Unwilling to take the risk of subprime-mortgage bonds in 2004 and 2005, the Wall Street firms swallowed the risk in 2006 and 2007. Lending standards had fallen, property values had risen, and the more recent loans were thus far riskier than the earlier ones, but still they gobbled them up—for if they didn’t, the machine would have ceased to function. The people inside the big Wall Street firms who ran the machine had made so much money for their firms that they were now, in effect, in charge. And they had no interest in anything but keeping it running. A.I.G. F.P. wasn’t an aberration; what happened at A.I.G. F.P. could have happened anywhere on Wall Street … and did.

As recently as August 2007, A.I.G. F.P. traders were feeling almost smug: all these loans made in 2006 and 2007 were going bad, but the relatively more responsible 2005 vintage that they had insured didn’t look as if it would suffer any credit losses. They were, they thought, the smart guys at the poker table. Joe Cassano even went on an investor conference call and said, famously, “It is hard for us, without being flippant, to even see a scenario within any kind of realm of reason that would see us losing $1 on any of those transactions.”

Say It Ain’t So, Joe

What no one realized is that Joe Cassano, in exchange for the privilege of selling credit-default swaps on subprime-mortgage bonds to Goldman Sachs and Merrill Lynch and all the rest, had agreed to change the traditional terms of trade between A.I.G. and Wall Street. In the beginning, A.I.G. F.P. had required its counter-parties simply to accept its AAA credit: it refused to post collateral. But in the case of the subprime-mortgage credit-default swaps, Cassano had agreed to several triggers, including A.I.G.’s losing its AAA credit rating, that would require the firm to post collateral. If the value of the underlying bonds fell, it would fork over cash, so that, for instance, Goldman Sachs would not need to be exposed for more than a day to A.I.G. Worse still, Goldman Sachs assigned the price to the underlying bonds—and thus could effectively demand as much collateral as it wanted. In the summer of 2007, the value of everything fell, but subprime fell fastest of all. The subsequent race by big Wall Street banks to obtain billions in collateral from A.I.G. was an upmarket version of a run on the bank. Goldman Sachs was the first to the door, with shockingly low prices for subprime-mortgage bonds—prices that Cassano wanted to dispute in court, but was prevented by A.I.G. from doing so when he was fired. A.I.G. couldn’t afford to pay Goldman off in March 2008, but that was O.K. The U.S. Treasury, led by the former head of Goldman Sachs, Hank Paulson, agreed to make good on A.I.G.’s gambling debts. One hundred cents on the dollar.

A pair of ancient maples shade Joe Cassano’s London home. It’s a tasteful, almost inconspicuous place on a square, one of the best in London, just around the corner from Harrods. Only the living-room drapes are left open, to let in the spring light. Four black porcelain elephants decorate the windowsill. Behind them a shadow moves through the room.

Cassano resigned from A.I.G. F.P. early last year, but he didn’t simply leave. He continued to turn up at his desk and spend the day staring at his Bloomberg TV. The traders thought it strange; only later did they learn that A.I.G. was still paying him $1 million a month to consult. As far as anyone could tell, he had nothing to do. And then one day he simply stopped showing up. From time to time they spotted him cycling past their Mayfair office. Every now and again some British newspaper snapped a picture of him exiting his house with his racing bike. Apart from that he had as good as vanished. His absence is as frustrating as it is expected—the people best positioned to explain this financial disaster have all similarly vanished from view. It would be nice if Joe Cassano came out of hiding and tried to explain what he did and why, but there is little chance of that.

The people still left inside A.I.G. F.P. like to list just how many things had to go wrong for their business to implode. Any one of a number of things might have sufficed to avert their catastrophe: our political leaders might have decided against the Wall Street argument not to regulate credit-default swaps; the ratings agencies might have resisted the Wall Street argument to rate subprime bonds AAA; Wall Street banks, in 2006 and 2007, might have declined to replace A.I.G. F.P. in the role of subprime risktaker of last resort; and on and on. Their list is mostly a catalogue of large, impersonal forces. But impersonal forces require people to conspire with them. Joe Cassano was the perfect man for these times—as responsible for a series of disastrous trades as a person in a big company can be. He discouraged the dissent of subordinates who understood them better than he did. He acted with the approval of A.I.G., but he also must have known that A.I.G. wasn’t able to evaluate his trades. Once he was persuaded to stop insuring subprime-mortgage bonds, the logical course of action was to reverse the deals he had already done. In 2006 he might have found a way to do this, if he had been willing to accept the costs involved, but he wasn’t. Had he been, the machine he helped to create would have kept running—by then it had a life of its own—and the losses would have simply wound up more concentrated inside the big banks. But he’d have saved his company. No one would be blaming Jake DeSantis for blowing up the world.

And yet the A.I.G. F.P. traders left behind, much as they despise him personally, refuse to believe Cassano was engaged in any kind of fraud. The problem is that they knew him. And they believe that his crime was not mere legal fraudulence but the deeper kind: a need for subservience in others and an unwillingness to acknowledge his own weaknesses. “When he said that he could not envision losses, that we wouldn’t lose a dime, I am positive that he believed that,” says one of the traders. The problem with Joe Cassano wasn’t that he knew he was wrong. It was that it was too important to him that he be right. More than anything, Joe Cassano wanted to be one of Wall Street’s big shots. He wound up being its perfect customer.

Michael Lewis is a Vanity Fair contributing editor.