The Quants: How a New Breed of Math Whizzes Conquered Wall Street and Nearly Destroyed It

by Scott Patterson

  The Medallion fund, however, was doing even worse than RIEF. It had dropped a whopping 17 percent for the month, a loss of roughly $1 billion. Like Muhammad Ali getting licked by Joe Frazier in Madison Square Garden in 1971, the greatest fund of all time was on the ropes, and seemed at risk of getting knocked out.

  Over at AQR, the mood was even more grim. Its traders were tense and tired. The hard, round-the-clock work was completely atypical for quants used to the rigid, structured, predictable flow of markets. Complete chaos wasn’t supposed to be part of the package.

  The fun was over. AQR had reserved a movie theater that Thursday night for a showing of The Simpsons Movie as an employee event. The reservation was canceled.

  Ken Griffin, meanwhile, sensed blood in the water. While the quantitative Tactical Trading fund run by Misha Malyshev was getting hammered, it represented only a fraction of Citadel’s massive girth.

  Late Thursday night, Griffin picked up the phone and called Cliff Asness. Griffin wasn’t calling as a friend. He wanted to know if AQR needed help.

  Asness knew what that meant. He was hearing from Griffin the grave dancer, the vulture investor of Amaranth and Sowood fame. It brought home how much trouble he was in. While the call was friendly, there was an air of tension between the two managers. “I looked up and saw the Valkyries coming and heard the Grim Reaper’s scythe knocking on my door,” Asness later joked about the call. But at the time, he wasn’t laughing.

  Friday morning at AQR, August 10. Asness glanced pensively at a candy-colored array of Marvel superhero figurines lined up along his east-facing window: Spider-Man, Captain America, the Hulk, Iron Man. Comic book heroes of his boyhood days on Long Island.

  The fund manager wished he had some kind of superhuman power over the markets to make it all stop. Make the bleeding stop. AQR’s Absolute Return Fund was down 13 percent for the month, its biggest drop in such a short stretch of time ever. It doesn’t make sense … it’s perverse. He walked to his desk and looked at the P&L on his computer screen, a red flash of negative numbers. The losses were astronomical. Billions gone.

  Through the eastern window of his office, Asness could see the blue shimmer of the city’s teeming marina beyond Steamboat Road. A decade earlier, a short drive down Steamboat had led directly to the headquarters of Long-Term Capital Management.

  If the losses continued, AQR would be seen as just another LTCM, a quant disaster that wreaked havoc on the financial system, Wall Street’s eggheads run amok all over again as their witchy black boxes turned into loose-cannon HAL 9000s destroying everything in sight.

  He didn’t want to let that happen. And he thought: Maybe there’s a way.

  Asness had been huddling with his top lieutenants in his office, including Mendelson, John Liew, and David Kabiller. They were getting ready to make a momentous decision. It wasn’t easy. The fate of the hedge fund hung in the balance.

  Throughout the week, AQR, like every quant fund, had been trying to figure out what was going on, searching for the elusive patient zero. By the end of Thursday, they had determined that almost every large quant hedge fund had taken down its leverage significantly. Every quant fund but one: GSAM.

  AQR had been frantically trying to get information about what was going on inside Goldman. But Goldman wouldn’t talk. Through careful analysis of the situation, AQR had determined that GSAM’s Global Equity Opportunities fund hadn’t completely deleveraged. That meant one of two things: either Goldman was going to inject a large amount of money into the fund to keep it afloat or it was about to implode in a vicious sell-off, causing the market to spiral even further out of control.

  If the latter were the case, certain disaster would have been in store for AQR and every other quant fund, as well as the broader market. Goldman’s GEO fund was massive, with roughly $10 billion in assets. If it had started selling on top of all the other losses quant funds had endured, a meltdown of epic proportions would have ensued, rocking investors everywhere.

  Like PDT, the team at AQR had planned to shrink its book even further that day. But Asness made a gut decision, one of the most important of his trading career: It’s time to buy.

  If not now, when? he thought. Goldman, he decided, wouldn’t let the system collapse if it had the wherewithal. Instead, the bank would do the smart thing, the rational thing: inject capital into the GEO fund. That would allow it to hold on to its positions. It wouldn’t have to deleverage.

  That meant it was time to get back in, throw more chips on the table. AQR put out the word to its traders and told them to be intentionally loud about the move. They wanted everyone to know that AQR, one of the big lumbering gorillas of quantdom, was back in action. Maybe that will make the bleeding stop, Asness thought.

  It was like a poker game, the highest-stakes hand he’d ever played. This time, it wasn’t just that wiseass Peter Muller who could call his bluff; it was the market itself that could ruin him. Asness was all in, and he knew it.

  Back in New York, Muller sat poker-faced and pensive as strategies to cope with the chaos raced through his mind. As he waited for the market to open Friday morning, he knew PDT was on the edge. The group had lost an inconceivable $600 million. If the losses intensified much more, Morgan could decide to shut it down. The group’s brilliant fourteen-year run, and possibly Muller’s career, restarted only months earlier, could be over.

  It didn’t look good. European markets were still a basket case. So was Asia. The tension mounted as the 9:30 a.m. start of regular trading in the United States neared. Muller, Simons, Asness, Weinstein, Griffin, and nearly every other quant manager in the world were glued to their screens as the minutes ticked by, sweating, nervous, sick with dread.

  Then something of a miracle happened. When U.S. trading began, quant strategies started to rally, hard. Muller gave the order: don’t sell. Other quants followed suit. There was an initial lull, and then their positions took off in a rocket-mad surge. By the end of the day, the gains were so strong that many quant managers said it was one of their best days ever. Whether the rebound was a result of AQR’s decision to leap back into the market is impossible to know for certain. But there’s little doubt that it helped turn the tide.

  Inside Goldman, rescue efforts had in fact been in full swing since Wednesday—a $3 billion infusion of cash that helped to stop the bleeding. The bailout, about $2 billion of which was Goldman’s own money, was targeted at the GSAM’s Global Equity Opportunities fund, which had also suffered a huge blow and had lost a stunning 30 percent, or $1.6 billion, for the month through August 9. Global Alpha and its North American Equity Opportunities fund were left to fend for themselves. By the end of August, Global Alpha’s assets had plunged to $6 billion, down from $10 billion the previous year, an enormous 40 percent decline for one of Wall Street’s elite trading groups.

  “There is more money invested in quantitative strategies than we and many others appreciated,” wrote Global Alpha’s managers in a report to battered investors later that month. A staggering amount of that money was sitting at Goldman Sachs Asset Management. Including the GEO fund and Global Alpha, GSAM had about $250 billion in funds under management, of which about $150 billion was in hedge funds.

  In a separate letter, Global Alpha’s managers explained that a big driver of the losses was the carry trade. “In particular,” they wrote, “we saw very poor performance in our currency selection strategies, both developed and emerging, as positions of ours that were aligned with carry traders were punished in the massive unwind of the worldwide carry trade.”

  They were chastened, but they still believed in their system. Acknowledging that the 23 percent decline that month had been “a very challenging time for our investors,” they said they “still hold to our fundamental investment beliefs: that sound economic investment principles coupled with a disciplined quantitative approach can provide strong, uncorrelated returns over time.”

  Asness issued his own letter late Friday—and he pointed the finger at copy
cats riding his coattails. “Our stock selection investment process, a long-term winning strategy, has very recently been shockingly bad for us and for all of those pursuing similar strategies,” he wrote. “We believe that this occurred as the very success of the strategy over time has drawn in too many investors.”

  When all of those copycats rushed for the exit at once, it led to “a deleveraging of historical proportions.”

  It was a black swan, something neither AQR nor any of the quants had planned for.

  Matthew Rothman, meanwhile, was frazzled to the bone. He’d spent most of Thursday and Friday explaining the situation to investors, clients of Lehman, confused CEOs of companies whose stock was getting crushed by the quant meltdown (“You do what to stocks? Why?”). He’d barely slept for two days.

  He called up a friend who lived in Napa Valley, an hour’s drive from San Francisco. “I’ve had a crazy week,” he said. “Mind if I stay at your place for the weekend?” Rothman spent the days visiting wineries and relaxing. It was one of the last moments he’d have for such a break in a very long time.

  Over the weekend, Alan Benson came into Saba’s office to go over his positions and bumped into Weinstein, who was trying to keep up to speed on the chaos that had enveloped the markets. Saba’s quant equities desk had lost nearly $200 million. Weinstein was clearly upset and told Benson to keep selling. By the time Benson was done, his positions had been cut in half.

  On Monday, Goldman Sachs held a conference call to discuss the meltdown and the $3 billion infusion into the GEO fund. “The developments of the last few days have been unprecedented and characterized by remarkable speed and intensity across global markets,” said David Viniar, Goldman’s chief financial officer. “We are seeing things that were 25-standard-deviation events, several days in a row.”

  It was the same out-of-this-world language the quants used to describe Black Monday. According to quant models, the meltdown of August 2007 was so unlikely that it could never have happened in the history of the human race.

  The meltdown by the quant funds was over, at least for the moment. But it was only the first round of a collapse that would bring the financial system to its knees. The following week, the turmoil in the financial markets only worsened. A global margin call was under way, and spreading.

  On Thursday morning, August 16, Countrywide Financial said it needed to tap $11.5 billion in bank credit lines, a sign that it couldn’t raise money on the open market. About the same time, in London, about $46 billion in short-term IOUs issued outside the United States were maturing and had to be rolled over into new debt. Typically this happens almost automatically. But that morning, no one was buying. Only half of the debt was sold by the end of the day. The Money Grid was breaking down.

  The yen continued to pop, surging 2 percent in a matter of minutes midday in New York that Thursday, a move that can crush a currency trader leaning the wrong direction. Treasuries were also soaring as panicked investors continued to buy the most liquid assets, a swing one trader at the time called “an extraordinarily violent move.”

  “These shocks reflected one of the most perilous days for global capital markets, the circulatory system of the international economy, since the 1997–98 crisis that began in Asia, spread to Russia and Brazil and eventually to the U.S.-based hedge fund Long-Term Capital Management,” stated a front-page article in the Wall Street Journal.

  Stock investors were pummeled by whiplash swings that saw the Dow industrials dip and surge by hundreds of points in the space of a few minutes. It was dizzying. The meltdown that had begun in subprime mortgages and spread to quant hedge funds was now visible to everyone—including the Federal Reserve.

  Early that Friday morning, stock markets were in free fall. At one point, futures tied to the Dow industrials were indicating that the market would open more than 500 points lower.

  Then, shortly after 8:00 a.m. Eastern time, the Fed lowered interest rates on its so-called discount window, through which it makes direct loans to banks, to 5.75 percent from 6.25 percent. The central bank hoped that by cutting rates through the window, it would encourage banks to make loans to customers that had previously been squeezed. Banks had been cutting off certain clients, such as hedge funds, that they feared held large portfolios of subprime mortgages. The fear about who was holding toxic assets was spreading. The Fed also signaled that it would likely lower the federal funds rate, the more important rate it charges banks for overnight loans, when it met again in September.

  It was a highly unusual move, and it worked. Stock futures surged dramatically and markets opened sharply higher.

  For the time being, the deleveraging appeared to have stopped. The quants had stared into the abyss. If the selling had continued—which likely would have happened if Goldman Sachs hadn’t bailed out its GEO fund—the results could have been catastrophic, not only for the quants but for everyday investors, as the sell-off spilled into other sectors of the market. Just as the implosion of the mortgage market triggered a cascading meltdown in quant funds, the losses by imploded quant funds could have bled into other asset classes, a crazed rush to zero that could have put the entire financial system in peril.

  The most terrifying aspect of the meltdown, however, was that it revealed hidden linkages in the Money Grid that no one had been aware of before. A collapse in the subprime mortgage market triggered margin calls in hedge funds, forcing them to unwind positions in stocks. The dominoes started falling, hitting other quant hedge funds and forcing them to unwind positions in everything from currencies to futures contracts to options in markets around the world. As the carry trade unraveled, assets that had benefited from all the cheap liquidity it had spun off began to lose their mooring.

  A vicious feedback loop ensued, causing billions to evaporate in a matter of days. The selling cycle had stopped before major damage had been inflicted—but there was no telling what would happen the next time around, or what hidden damage lurked within the system’s mostly invisible plumbing.

  The unwind that week had been so unusual, so unexpected, that several of the rocket scientists at Renaissance Technologies gave it its own name: the August Factor. The August Factor represented a complete reversal of quant strategies, the Bizarro World in which up was down and down was up, bad assets rose and good assets fell, ignited by a mass deleveraging of funds with overlapping strategies. It was an entirely new factor, with strong statistical properties unlike any that had ever been seen in the past—and, hopefully, would never be seen again.

  But there were new, far more destructive disruptions coming. Indeed, a financial storm of unprecedented fury was already under way. In the next two years, the relentless deleveraging that first hit obscure places such as PDT’s New York office and AQR’s Greenwich headquarters spread throughout the financial system like a mutating virus, pushing the financial system to the edge of a cliff. Trillions were lost, and giant banks failed.

  Looking back, however, many quants would see the dramatic, domino-like meltdown of August 2007, one that scrambled the most sophisticated models in the world, as the strangest and most unexplainable event of the entire credit crisis.

  “In ten years, people may remember August ’07 more than they remember the subprime crisis,” Aaron Brown observed. “It started a chain reaction. It’s very interesting that there was this tremendous anomalous event before the great crisis.”

  Cliff Asness paced back and forth, alone in his corner office at AQR, wringing his hands. It was late November 2007, and AQR was on its heels all over again, suffering huge losses.

  What had happened? The fund had ripped higher after the August meltdown, making back almost all the losses of that insane week. Everything looked fine. For a while he even dared to dream the IPO might be on the table again. September was okay, and so was October.

  In November, the nightmare started all over again. AQR was slammed as a number of quant strategies were hit. The global margin call continued to batter the financial system. Subprime CD
O assets continued to collapse, and investors were coming to realize that far more banks than they had imagined were holding toxic assets. Morgan Stanley fessed up to a $7.8 billion loss, assigning most of the blame to Howie Hubler’s desk. Mortgage giant Freddie Mac revealed a $2 billion loss. HSBC, one of Europe’s biggest banks, took $41 billion in assets it had held in special investment vehicles—those offbalance-sheet entities that ferried subprime mortgages through the securitization pipeline—onto its balance sheet, a symptom of the frozen credit markets. Citigroup, Merrill Lynch, Bear Stearns, and Lehman Brothers also started to show even more severe strains from the crisis.

  AQR was getting hit on all sides. Asness’s precious value stocks were plunging. Currencies and interest rates were all over the map. A big bet he’d made on commercial real estate turned south in dramatic fashion, losing hundreds of millions in the course of a few weeks.

  Less than three years earlier, back in the golden days of the Wall Street Poker Night at the St. Regis, the quants had been one of the most powerful forces in the market, the Nerd Kings of Wall Street. Asness and Muller had stood shoulder to shoulder, the poker trophies in their hands like symbols of their shared ability to make the right calculations to collect pots of money. Now it seemed to have been only act two of a three-act Greek tragedy about hubris. They were getting crushed by a market gone mad. It wasn’t right. It wasn’t fair.

  Asness sat down at his desk and stared at his computer screen. More red numbers. He hauled back and lunged with a roar, punching the screen with his fist. The screen cracked and flipped back off his desk, falling to the floor, destroyed.

  Asness shook his head, gazing out the window at the browning foliage of Greenwich beyond. He knew he wasn’t the only hedge fund taking a beating as 2007 drew to a close. The global financial crisis was metastasizing like a cancer. A reckoning for the high-flying industry was under way, pummeling even the savviest operators. AQR had long been considered one of the smartest, most advanced funds in the business. But starting in August 2007, it all seemed to be coming undone. All the math, all the theory—none of it worked. Whatever AQR did to try to right its ship proved the wrong move as wave after wave of deleveraging ripped across the system.