The Quants Part 14

Speaking before the troops, roughly four hundred in all, Griffin was like a charged-up football coach preparing his team for the biggest game of their lives. After ticking off Citadel's accomplishments, Griffin s.h.i.+fted into corporate manager cliche mode. "Success has never been measured in home runs," he said, "it's the singles and doubles that got us here and will take us beyond. The best times are yet to come. Yes, there will be obstacles, but obstacles are opportunities for people who know how to get the job done. If you're up for that, you're in the right place."

The well-heeled crowd clapped and cheered. Griffin may be a hard driver, even a ruthless megalomaniac, but he was a winner and had made everyone in the room incredibly wealthy. Citadel seemed on the cusp of greatness. The downturn the economy was suffering from the collapse in the housing market would be short-lived, Griffin thought, a brief hiccup in the global economy's unstoppable growth cycle. Indeed, he believed he could already see the light at the end of the tunnel. Good times were coming soon.

There's an old Wall Street proverb about such opportunism: the light at the end of the tunnel is an oncoming train. Ken Griffin was stepping right in front of it.

"Aaargh!"

Cliff Asness leapt from the card table, grabbed the first lamp in reach, and smashed it against the wall. He stood seething before the wide hotel window. It was late 2007, and a light dusting of snow fell over New York. Christmas lights adorned a number of the apartment windows of the luxury high-rise across from the hotel.

"What the h.e.l.l is wrong with you?" Peter Muller said, looking up, startled, from his seat at the table.

There he goes again. Asness had lost another hand. Bad luck. But why did he care so much? What was with the temper tantrums? Asness's hedge fund made its money based on math, on cold-blooded rationality that ruthlessly cuts away the irrational, human element of trading. But when the chips were down at the poker table, Asness lost it.

Neil Chriss shook his head. "Cliff, you make and lose that much in a matter of minutes every day," he said. "I think some perspective is in order."

Why did Asness always take losing so personally? Why did he get so angry? He'd always had a temper, and hated to lose, especially to other quants.

"Screw it," Asness said, breathing heavily as he moved back to the card table.

In the previous year, Asness had been increasingly p.r.o.ne to outbursts. The pots seemed to keep rising, easily hitting five digits, sometimes more. Not that Asness couldn't afford it. He was the richest guy in the room. But his fortunes seemed to be dwindling every day as AQR hemorrhaged cash. And Asness's skills at the poker table seemed to track AQR's P&L-they waned at the same time his fund started to skid. Just my luck, he thought. Or lack of luck. It was nuts.

The quant poker games were brutal marathon sessions lasting until three, four in the morning. Not that Asness would stick around for the whole affair. He had two pairs of twins, born back-to-back in 2003 and 2004, waiting for him back at his mansion in Greenwich. He liked to call the sequential birth of the twins "a gross failure of risk control," referring to an overindulgence in fertility treatments.

Risk control seemed to get thrown out the window for the poker game, too, or so it might seem to an outsider. The buy-in was $10,000. For certain games attended by the more serious players of the group, such as Muller and Chriss, the buy-in could vault as high as fifty grand.

The players didn't toss it all on the table on the first hand, of course. They could stuff the chips in their pockets and keep them there all night, at least until their luck ran out, if that's how they wanted to play it. But who cared? Fifty grand was Monopoly money to them. It was all about who won and who lost. And although winners sometimes could take home winnings measured in six figures, it wasn't going to change any of their lives.

But Asness wasn't winning. He was losing. Just like AQR.

"Ante up," said Muller, dealing another hand.

Asness pulled a stack of chips from his pocket and tossed them into the pot. He watched the cards as they fell around the table. He looked up at Muller, who was staring blankly at his hand. He didn't know how Muller could stay so calm. He'd lost more than half a billion in August over the course of a few days, yet acted like it was another day on the beach in Hawaii. But AQR had lost even more, much more. Sure, things had bounced back-a lot-but the speed of the meltdown had been unnerving. And now, as the credit crisis ground on in late 2007, AQR was facing even more losses.

Picking up his cards, Asness winced. Nothing.

He wasn't ready to give up yet, not even close. AQR held its traditional Christmas party at n.o.bu 57, a sw.a.n.k j.a.panese restaurant in midtown Manhattan. But there were signs that the bloom was off the rose. Spouses and guests weren't allowed, unlike previous years. The stressed Greenwich quants let loose, swilling sake and j.a.panese beer by the gallon. "It turned into a drunkfest," said one attendee.

The quants were also haunted by another fear: systemic risk. The August 2007 meltdown showed that the quants' presence in the market wasn't nearly as benign as they had believed. As with a delicate spiderweb, a tear in one part of the financial system, in this case subprime mortgages, could trigger a tear in another part-and even bring down the web itself. The market was apparently far more intertwined than they had ever realized.

MIT finance professor Andrew Lo, and his student Amir Khandani, published a definitive study of the meltdown in October 2007 called "What Happened to the Quants?" Ominously, they evoked an apocryphal Doomsday Clock for the global financial system. In August 2007, the clock ticked nearer to midnight, perhaps the closest it had come to financial Armageddon since Long-Term Capital's implosion in 1998.

"If we were to develop a Doomsday Clock for the hedge-fund industry's impact on the global financial system," they wrote, "calibrated to five minutes to midnight in August 1998, and fifteen minutes to midnight in January 1999, then our current outlook for the state of systemic risk in the hedge-fund industry is about 11:51 P.M P.M. For the moment, markets seem to have stabilized, but the clock is ticking."

One of the central concerns, Lo and Khandani explained, was the weblike interconnectedness of the system. "The fact that the ultimate origins of this dislocation were apparently outside the long-short equity sector-most likely in a completely unrelated set of markets and instruments-suggests that systemic risk in the hedge-fund industry has increased significantly in recent years," they wrote.

There was also the worry about what happened if high-frequency quant funds, which had become a central cog of the market, helping transfer risk at lightning speeds, were forced to shut down by extreme market volatility. "Hedge funds can decide to withdraw liquidity at a moment's notice," they wrote, "and while this may be benign if it occurs rarely and randomly, a coordinated withdrawal of liquidity among an entire sector of hedge funds could have disastrous consequences for the viability of the financial system if it occurs at the wrong time and in the wrong sector."

It wasn't supposed to be this way. The quants had always thought of themselves as financial helpers, greasing the churning wheels of the Money Grid. Now it seemed that they posed significant systemic risk-pus.h.i.+ng the world closer to doomsday. Sitting at the poker table, holding another dud hand, Asness shut his eyes and thought back to his days as the most brilliant student at the University of Chicago.

Where had it all gone wrong?

One quant gadfly had been predicting a crackup in the financial system for years: Na.s.sim Nicholas Taleb, the former hedge fund trader and author who'd squared off against Peter Muller at Neil Chriss's wedding several years back. In January 2008, Taleb arrived at AQR's office in Greenwich to give a lecture. Aaron Brown had asked him to explain his theories about why quant.i.tative models work fine in the physical world but are dangerous wizardry in the world of finance (a view Brown didn't necessarily share). gadfly had been predicting a crackup in the financial system for years: Na.s.sim Nicholas Taleb, the former hedge fund trader and author who'd squared off against Peter Muller at Neil Chriss's wedding several years back. In January 2008, Taleb arrived at AQR's office in Greenwich to give a lecture. Aaron Brown had asked him to explain his theories about why quant.i.tative models work fine in the physical world but are dangerous wizardry in the world of finance (a view Brown didn't necessarily share).

Taleb's audience was spa.r.s.e. Asness, drained, pa.s.sed. Aaron Brown had been a friend of Taleb's for years-Taleb had provided a blurb for Brown's book, The Poker Face of Wall Street The Poker Face of Wall Street-and was interested in what he had to say, even if he didn't agree with it.

"Hey, Na.s.sim, how goes it?"

"Not bad, my friend," Taleb said, stroking his beard. "I hear you have been having a bad time of it."

"You wouldn't believe it," Brown replied. "Or maybe you would, I don't know. I'd say we've definitely seen one of the blackest swans of all time. But things seem to be cooling off."

Taleb quickly set up his PowerPoint demonstration and began to talk. One of the first slides in the talk showed a clip from a Wall Street Journal Wall Street Journal article from August 11 about Matthew Rothman's description of the quant meltdown. article from August 11 about Matthew Rothman's description of the quant meltdown.

"Matthew Rothman is used to working with people who pride themselves on their rationality," the article said. "He's a 'quant,' after all, one of a legion of Ph.D.'s on Wall Street who use emotionless rules of mathematics to pick trading positions. But this week, he caught a whiff of panic."

Taleb's slide was t.i.tled "Fallacy of Probability." Rothman had described the quant meltdown as something models predicted would happen once in ten thousand years-but it had taken place every day for several days in a row. To Taleb, that meant something was wrong with the models.

"These so-called financial engineers experience events that can only happen once, in the history of mankind, according to the laws of probability, every few years," he told the room (full, of course, of financial engineers). "Something is wrong with this picture. Do you see my point?"

Another slide showed a giant man sitting on the right side of a scale, tipping it heavily, while a group of tiny people scatter and fall on the left side. The slide read "Two Domains: Type 1-Mild 'Mediocristan' (Gauss, etc.); Type 2-Wild 'Extremistan.'"

The slide was a key to Taleb's vision about extreme events in the market and why the mathematics used in the physical world-the science used to put a man on the moon, fly an airplane across the ocean, microwave a sandwich-doesn't apply in the world of finance. The physical world, he said, is "Mediocristan." Bell curves are perfect for measuring the heights or weights of people. If you measure the height of a thousand people, the next measurement isn't likely to change the average.

In finance, however, a sudden swing in prices can change everything. This is Taleb's world of "Extremistan." Income distributions, for instance, exhibit signs of Extremistan, a discovery Benoit Mandelbrot had made more than half a century before. Measure the wealth of a thousand people plucked off the street. On a typical day, the distribution will be normal. But what if you select Bill Gates, the richest man in the world? The distribution is suddenly, ma.s.sively skewed. Market prices can also change rapidly, unexpectedly, and ma.s.sively.

Taleb continued to address the spa.r.s.e audience for another thirty minutes. He talked about fat tails. Uncertainty. Randomness. But he could tell his audience was fried. They didn't need to be told about black swans. They'd just seen one, and it had terrified them.

Still, few could believe that the downturn would get much worse. A year of unfathomable volatility was just beginning. Later that January, it emerged that a thirty-one-year-old rogue trader at Societe Generale, a large French bank, lost $7.2 billion on complex derivatives trades. The trader, Jerome Kerviel, used futures contracts tied to European stock indexes to build up a staggering $73 billion worth of positions that were basically one-way bets that the market would rise. After the bank discovered the trades, which Kerviel covered up by hacking its risk-control software, it decided to unwind them, triggering a staggering global market sell-off. In response to the volatility, the Federal Reserve, which didn't know about the SocGen trades, slashed short-term rates by three-quarters of a point, a bold move that frightened investors because it smacked of panic.

Still, even as the system teetered on the edge, many of the smartest investors in the world couldn't see the destructive tsunami heading directly for them. The implosion of Bear Stearns in March was a wake-up call.

The Doomsday Clock was ticking.

Around 1:00 p.m. on March 13, 2008, Jimmy Cayne sat down at a card table in Detroit and began to craft his strategy. The seventy-four-year-old chairman of Bear Stearns, seeded fourth in a group of 130 in the IMP Pairs category of the North American Bridge Champions.h.i.+p, was squarely focused on the cards in his hand.

Bridge was an obsession for Cayne, a product of Chicago's hardscrabble South Side, and he wasn't going to let his company's troubles get in the way of one of the most important compet.i.tions of the year.

At the same time, back at Bear's New York headquarters on Madison Avenue, about forty of the firm's top executives had gathered in a twelfth-floor dining area to strategize. Everyone knew trouble was brewing. Bear's anemic stock price told the story all too clearly. But no one was sure exactly how bad it was. Around 12:45 P.M P.M., Bear chief executive Alan Schwartz appeared to a.s.sure the troops that all was well.

No one was buying it. Bear Stearns, founded in 1923, was teetering on the edge of collapse as its trading clients pulled billions from the bank in a white-hot panic. Insiders at the firm knew it was serious when one of its most cherished clients pulled out more than $5 billion in the first half of March. The client: Renaissance Technologies. Then another top client bolted for the exits with $5 billion more in hand: D. E. Shaw.

The quants were killing Bear Stearns.

To this day, former Bear Stearns employees believe the firm was taken out in a ruthless mugging. During its final week as a public ent.i.ty, it had $18 billion in cash reserves on hand. But once Bear's blood was in the water, jittery clients who traded with Bear weren't willing to wait around to see what happened. The worry was that the bank would implode before they could pull out their money. It wasn't worth the risk. There were other investment banks more than willing to take their funds, such as Lehman Brothers.

By March 15, 2008, a Sat.u.r.day, Bear Stearns was nearly finished. Federal Reserve and Treasury Department officials and bankers from J. P. Morgan roamed the halls of its midtown Manhattan skysc.r.a.per like scavengers picking over a cadaver. Bear executives were frantic, worried about a shotgun wedding and jamming the phones for any kind of last-minute salvation. Nothing worked. On Sunday, Cayne and the rest of Bear's board agreed to sell the eighty-five-year-old inst.i.tution to J. P. Morgan for $2 a share. A week later, the deal was boosted to $10 a share.

For a time, optimistic investors believed the death of Bear marked the high-water mark of the credit crisis. The stock market shot up. The system had seen its moment of crisis and come through largely unscathed. Or so it seemed.

d.i.c.k Fuld was putting on a cla.s.sic performance. The Lehman Brothers CEO, known as the "Gorilla" for his heavy-browed Cro-Magnon glare, monosyllabic grunts, and fiery rampages, had been ranting for more than a half hour to a roomful of managing directors. was putting on a cla.s.sic performance. The Lehman Brothers CEO, known as the "Gorilla" for his heavy-browed Cro-Magnon glare, monosyllabic grunts, and fiery rampages, had been ranting for more than a half hour to a roomful of managing directors.

Fuld screamed. Jumped up and down. Shook his fists in defiance.

It was June 2008. Lehman's stock had been getting hammered all year as investors fretted about the firm's shaky balance sheet. Now it was getting worse. The firm had just posted a quarterly loss of $2.8 billion, including $3.7 billion in write-downs for toxic a.s.sets such as mortgages and commercial real estate investments. It was the bank's first quarterly loss since 1994, when it was spun off from American Express. Despite the losses, Fuld and his lieutenants had kept a straight face publicly, insisting everything was fine. It wasn't.

Fuld had called a meeting of the firm's managing directors to clear the air and explain the situation. He began with an announcement: "I spoke to the board this weekend," he said. Some in the room wondered whether he'd offered to resign. Just a week earlier, the sixty-two-year-old CEO had scrambled the firm's executive office, replacing president Joe Gregory with his longtime partner Herbert "Bart" McDade. Was it Fuld's turn to fall on his sword? Some in the room hoped so.

"I told them," Fuld said, "I'm not taking a bonus this year."

The room seemed to release an audible sigh of despair. Fuld quickly started going through the math, laying out how strong Lehman was, how solid its balance sheet remained. He talked about how the firm would crush the short sellers who'd been pile-driving Lehman's stock into dust.

Someone raised a hand. "We hear everything you're saying, d.i.c.k. But talk is cheap. Acting is louder than words. When are you going to buy a million shares?"

Fuld didn't miss a beat. "When Kathy sells some art."

Fuld was referring to his wife, Kathy Fuld, known for her expensive art collection. Was he joking? Some wondered. Fuld wasn't laughing. There was the cla.s.sic furrowed brow. It was a moment when some of Lehman's top lieutenants started to wonder in earnest whether Lehman was in fact doomed. Their CEO seemed detached from reality. When Kathy sells some art? When Kathy sells some art?

There was a yell, a smash, the sound of gla.s.s breaking and cras.h.i.+ng to the floor. AQR's researchers and traders jolted in their seats, looking up shocked from their computer screens toward John Liew's office, where the sudden crash had come from, breaking the standard office calm typified by the constant low hum and snick of quants typing furiously at keyboards. a yell, a smash, the sound of gla.s.s breaking and cras.h.i.+ng to the floor. AQR's researchers and traders jolted in their seats, looking up shocked from their computer screens toward John Liew's office, where the sudden crash had come from, breaking the standard office calm typified by the constant low hum and snick of quants typing furiously at keyboards.

Through the windows, they could see their boss, Cliff Asness, peering back at them, smiling sheepishly. He opened the door.

"Everyone is okay," he said. "Calm down."

It was another outburst. Asness had hurled a hard object at the wall, scoring a direct hit on a framed picture in Liew's office, shattering the gla.s.s. Asness had already destroyed several computer screens as well as an office chair as AQR's fortunes continued to sour. It was late summer 2008. The mood in the office had grown tense. The carefree days of just over a year ago were long gone, replaced by paranoia, fear, and worry. Some believed the firm was losing direction, but no one dared challenge the mercurial boss. Asness had surrounded himself with yes-men, some complained, and brooked no deviation from the carefully wrought models that had made him wildly rich. "It will all come back," he said repeatedly, like a mantra. "When the insanity goes away."

Others weren't so sure, and some employees were getting increasingly alarmed by the fund manager's outbursts. "He was losing it a little more and more every day," said a former employee. "He was off his rocker. Things were spinning out of control."

A key player had abandoned s.h.i.+p. Earlier in the year, Mani Mahjouri, one of AQR's whiz kids who'd been with the firm since 2000, had quit. He'd grown tired of Asness's tongue-las.h.i.+ngs, the battle-axe emails. Manjouri had been an idol of the younger quants at AQR. A student of Ken French's at MIT in the 1990s with degrees in math, physics, and finance, he was on the verge of becoming a partner, living proof that a young gun could rise to the top in a culture dominated by Goldman veterans. He was also the cla.s.s clown of the fund, turning his office into a haunted house during Halloween, decorating the cubicles of a researcher on his or her birthday with balloons and party hats (unbeknownst to the researcher), and hacking into the target's email and writing: "Today is my birthday, please come to my cubicle and celebrate with me," sending the message to the entire firm-hugely embarra.s.sing to certain antisocial quants.

The fun and games were over. Mahjouri was gone. The IPO was history, a bad reminder of better days. As the summer of 2008 neared an end, few of the quants at AQR could fathom-probably Asness most of all-that the pain was about to get much worse.

By September 9, d.i.c.k Fuld's confidence in Lehman Brothers was visibly shaken. In his thirty-first-floor office at the bank's midtown Manhattan headquarters, equipped with a shower, library, and expansive views of the Hudson, the Wall Street mogul raged against his tormentors like Ahab on the deck of the 9, d.i.c.k Fuld's confidence in Lehman Brothers was visibly shaken. In his thirty-first-floor office at the bank's midtown Manhattan headquarters, equipped with a shower, library, and expansive views of the Hudson, the Wall Street mogul raged against his tormentors like Ahab on the deck of the Pequod Pequod. That morning, news broke that Lehman's white knight, the Korea Development Bank, had decided not to purchase a stake in the bank. Making matters worse, if not catastrophic, J. P. Morgan's cochief of investment banking, Steven Black, called Fuld and told him that Morgan needed $5 billion in extra collateral and cash. Lehman had been margin-called. It was a dagger. Lehman's shares were in free fall, down more than 40 percent.

"We've got to act fast so this financial tsunami doesn't wash us away," Fuld said to his underlings, a manic tone in his voice.

But it was too late. The firm that Fuld had joined in 1969 was in a death spiral. Over the weekend of September 13, 2008, Lehman's fate was decided among a select group of individuals at the Federal Reserve's concrete fortress on Liberty Street in downtown Manhattan. Fuld wasn't even present. Instead, Hubert McDade and Alex Kirk, a fixed-income expert, sat at the table with Treasury secretary Paulson and New York Fed president Tim Geithner, President Obama's future Treasury secretary.

Fuld machine-gunned phone calls to the meeting, frantically making offers, spinning new deals. "How about this? How about this?"

Nothing worked. London banking giant Barclays, run by Bob Diamond, briefly considered ponying up some cash for Lehman, so long as the Fed backed the deal as it had with Bear. Paulson said no.

Derivatives traders, frantic about the demise of one of the world's largest banks, convened at the New York Fed's office Sat.u.r.day night. The goal was to create a game plan for settling trades in case Lehman imploded. Among those traders was Boaz Weinstein. Deutsche Bank had done a significant amount of trading through Lehman, and Weinstein was concerned about the impact of a Lehman collapse on his positions. He seemed calm and relaxed, as poker-faced as ever. Beneath his calm exterior, Weinstein was nervous, realizing that he could be facing the biggest test of his trading career.

Sunday morning a consortium of bankers briefly cobbled together a deal to back a Barclays-led buyout, but the plan fell apart. Regulators in the United Kingdom had gotten cold feet and wouldn't sign off. It spelled doom for Lehman. Sunday night McDade returned to Lehman's midtown headquarters and told Fuld the bad news. Lehman would have to file for bankruptcy.

"I want to throw up," Fuld moaned.

That Sunday, Lehman quant Matthew Rothman was furious. His bank was teetering. And still his bosses wanted him to fly to Europe for a series of quant conferences in London, Paris, Milan, Frankfurt, and Zurich? Idiots Idiots.

He checked his schedule. He was on tap to give the keynote speech for Lehman's quant.i.tative conference in London the following day. The previous week he'd sent an email to the team in Europe organizing the conference: "We may be filing for bankruptcy; there's a good chance we may not even be here." The response: You're crazy You're crazy.

Rothman's boss, Ravi Mattu, was bombarded with complaints about Rothman. He's not a team player. He's psycho. Of course Lehman Brothers isn't going to declare bankruptcy! He's not a team player. He's psycho. Of course Lehman Brothers isn't going to declare bankruptcy!

Rothman was incredulous. He wanted to be available for his team in case anything happened. Like a platoon sergeant in a foxhole, he didn't want to leave his troops when the s.h.i.+t hit the fan. And he could see it coming. So he reached a compromise: he'd take the red-eye to London for Monday's conference, then head straight back to Heathrow for the red-eye back to New York. It would suck, but at least he'd be around in case anything happened.

Sunday afternoon, he took a car from his home in Montclair, New Jersey, to JFK Airport. All along the way he was checking his BlackBerry, looking for news or emails from his colleagues at Lehman. At the airport, as he was checking into the terminal, he sent one last email to Ravi: "Do you want me to go to this conference?"

Just as Rothman was about to go through security, he got a response. "Cancel the trip."

Rothman's first emotion was relief. Then it hit him: Lehman was dying. Numb with the realization of what had happened, he took a cab back home to Montclair and immediately hopped in his wife's station wagon, leaving his '96 Honda Civic behind. He needed the extra room, he thought. For boxes.

Nearly every Lehman employee who could make it streamed toward the bank's New York office that night. Ranks of cameras and news teams lined Seventh Avenue. Rumors were flying around that the bank would shut its doors at midnight. There was no time to lose.

Inside the bank, there was relative calm. Employees were busy packing up their belongings. It was a surreal scene, like a wake. Another rumor came down: the computer systems were going to shut down. Everyone started sending emails, saying their goodbyes, attaching email addresses where they could be reached in the future, "It's been great to work with all of you," et cetera. Rothman sent his own email, picked up his belongings, and carried them out to his wife's station wagon.

Monday morning, chaos gripped Wall Street. Lehman had declared bankruptcy. Merrill Lynch had disappeared into Bank of America. American International Group, the world's largest insurer, teetered on the edge of collapse.

Outside Lehman's office, hordes of cameramen perched like vultures, pouncing on any bedraggled, box-laden Lehman employee scurrying from the building. Satellite dishes stacked atop vans lined the western verge of Seventh Avenue at the feet of Lehman's spotlit skysc.r.a.per. Mutating pixels of images and colors slithered robotically across the bank's twenty-first-century facade, a triple-deck stack of ma.s.sive digital screens. A bulky man in a blue suit and candy-striped tie, bald with a bushy white mustache-a beat cop dressed for a funeral-protected the doors to the besieged building.

A man in a scruffy white jacket and green cap shuffled back and forth in front of the building's entrance, eyes shooting toward the turnstiles. "The capitalist order has collapsed," he shouted, waving a fist as cameramen snapped the idle shot. "The whole scam is falling apart." Security men quickly shooed him away.

In the executive suite on the thirty-first floor, d.i.c.k Fuld looked down upon the spectacle below. His global banking empire lay in ruins beneath his feet. To protect himself from fuming employees, Fuld, who'd taken home a $71 million paycheck in 2007, had staffed up on extra bodyguards. A garish painting of Fuld, perched on the sidewalk outside the building, was covered in angry messages scrawled in marker, pen, pencil. "Greed took them down," read one. Another: "d.i.c.k, my kids thank you."

Credit markets were frozen as trading began. Investors were struggling to make sense of Lehman's collapse and the black cloud hanging over AIG. Later that Monday, rating agencies slashed AIG's credit. Since AIG had relied on its triple-A rating to insure a number of financial a.s.sets, including billions in subprime bonds, the change pushed it toward the edge of insolvency. Rather than let AIG fail, the U.S. government stepped in with a ma.s.sive bailout.

A unit of AIG, called AIG Financial Products, was behind the blowup. Known as AIG-FP, the unit had gobbled up $400 billion of credit default swaps, many of which were tied to subprime loans. AIG-FP's headquarters were in London, where it could sidestep tricky U.S. banking laws. It had a AAA rating, which made its business attractive to nearly every investor imaginable, from hedge funds to highly regulated pension funds. The sterling rating also allowed it to sell products cheaper than many compet.i.tors.

AIG-FP had sold insurance on billions of dollars of debt securities tied to a.s.set-backed CDOs packed with everything from corporate loans to subprime mortgages to auto loans to credit-card receivables. Since AIG-FP had such a high credit rating, it didn't have to post a dime of collateral on the deals. It could just sit back and collect fees. It was a form of infinite leverage based on AIG's good name. The collateral was the soul and body of AIG itself.

The models that gauged the risk of those positions were constructed by Gary Gorton, a quant who also taught at Yale University. The models were replete with estimates about the likelihood that the bonds AIG was insuring would default. But defaults didn't torch AIG-FP's balance sheet: AIG-FP was killed by margin calls. If the value of the underlying a.s.set insured by the swaps declined for whatever reason, the protection provider-AIG-FP-would have to put up more collateral, since the risk of default was higher. Those collateral calls started to soar in the summer of 2007. Goldman Sachs, for instance, had demanded an extra $8 billion to $9 billion in collateral from AIG-FP.

It was a case of model failure on a ma.s.sive scale. AIG had rolled the dice on a model and had c.r.a.pped out.

Meanwhile, the hasty exodus of Lehman employees that Sunday night had proven premature. The following week, Barclays purchased Lehman's investment banking and capital markets units, which included Rothman's group. But the damage had been done, and regulators were scrambling to contain it.

On Thursday, September 18, Fed chairman Ben Bernanke, Treasury secretary Hank Paulson, and a select group of about sixteen top legislators, including New York senator Chuck Schumer, Arizona senator Harry Reid, and Connecticut senator Chris Dodd, gathered around a polished conference table in the offices of House Speaker Nancy Pelosi. Bernanke began to talk. The credit markets had frozen, he explained, likening the financial system to the arteries of a patient whose blood had stopped flowing.

"That patient has had a heart attack and may die," Bernanke said in a somber tone to the dead-quiet room. "We could have a depression if we don't act quickly and decisively."

Bernanke spoke for about fifteen minutes, outlining a looming financial Armageddon that could destroy the global economy. The Money Grid was collapsing. The elected officials who had been faced with terrorist attacks and war were stunned. The loquacious Chuck Schumer was speechless. Chris Dodd, whose state pulled in billions from hedge fund taxes, turned talc.u.m white.

The infusion of cash came quickly. The government stepped in with an $85 billion bailout of AIG, which soared to about $175 billion within six months. In the following weeks, the Treasury Department, led by Hank Paulson, former CEO of Goldman Sachs, unveiled a plan to pump $700 billion into the financial system to jolt the dying patient back to life. But no one knew if it would be enough.

Andy Lo's Doomsday Clock was nearing midnight.

A FLAW