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

by Scott Patterson

  While such a vision seems unnerving, it appeared to many to be far more realistic, and certainly captured the nature of the wild ride that started in August 2007.

  Then there were the behavioral finance theories of Daniel Kahneman, who picked up a Nobel Prize for economics in 2002 (his colleague, Amos Tversk, had passed away years earlier). The findings of behavioral finance—often studies conducted on hapless undergraduate students in stark university labs—had shown time and again that people don’t always make optimal choices when it comes to money.

  A similar strand of thought, called neuroeconomics, was delving into the hardwiring of the brain to investigate why people often make decisions that aren’t rational. Some investors pick stocks that sound similar to their own name, for instance, and others pick stocks with recognizable ticker symbols, such as HOG (Harley-Davidson). Evidence was emerging that certain parts of the brain are subject to a “money illusion” that blinds people to the impact of future events, such as the effect of inflation on the present value of cash—or the possibility of a speculative bubble bursting.

  A small group of researchers at a cutting-edge think tank called the Sante Fe Institute, led by Doyne Farmer (the hedge fund manager and chaotician who briefly met Peter Muller in the early 1990s), was developing a new way to look at financial markets as an ecology of interacting forces. The hope was that by viewing markets in terms of competing forces vying for limited resources, much like Lo’s evolutionary vision, economists, analysts, and even traders will gain a more comprehensive understanding of how markets work—and how to interact with those markets—without destroying them.

  And while quants were being widely blamed for their role in the financial crisis, few—aside from zealots such as Taleb—were calling for them to be cast out of Wall Street. That would be tantamount to banishing civil engineers from the bridge-making profession after a bridge collapse. Instead, many believed the goal should be to design better bridges—or, in the case of the quants, better, more robust models that could withstand financial tsunamis, not create them.

  There were some promising signs. Increasingly, firms were adapting models that incorporated the wild, fat-tailed swings described by Mandelbrot decades earlier. J. P. Morgan, the creator of the bell curve–based VAR risk model, was pushing a new asset-allocation model incorporating fat-tailed distributions. Morningstar, a Chicago investment-research group, was offering retirement-plan participants portfolio forecasts based on fat-tailed assumptions. A team of quants at MSCI BARRA, Peter Muller’s old company, had developed a cutting-edge risk-management strategy that accounted for potential black swans.

  Meanwhile, the markets continued to behave strangely. In 2009 the gut-churning thousand-point swings of late 2008 were a thing of the past, but stocks were still mired in a ditch despite an early-year rally; the housing market looked as if it would keep cratering until the next decade. Banks had dramatically reduced their leverage and promised their new investor—the U.S. government—that they would behave. But there were signs of more trouble brewing.

  As early as the spring of 2009, several banks reported stronger earnings numbers than most expected—in part due to clever accounting tricks. Talk emerged about the return of big bonuses on Wall Street. “They’re starting to sin again,” Brad Hintz, a respected bank analyst, told the New York Times.

  Quant funds were also suffering another wave of volatility. In April, indexes that track quant strategies suffered “some of the best and worst days ever … when measured over approximately 15,000 days,” according to a report by Barclays quant researcher Matthew Rothman (formerly of Lehman Brothers).

  Many of the toxic culprits of the meltdown were dying away. The CDOs were gone. Trading in credit default swaps was drying up. But there were other potentially dangerous quant gadgets being forged in the dark smithies of Wall Street.

  Concerns about investment vehicles called exchange-traded funds were cropping up. Investors seemed to be piling into a number of highly leveraged ETFs, which track various slices of the market, from oil to gold mining companies to bank stocks. In March 2009 alone, $3.4 billion of new money found its way into leveraged ETFs. Quant trading desks at banks and hedge funds started tracking their behavior using customized spreadsheets, attempting to predict when the funds would start buying or selling. If they could predict the future—if they knew the Truth—they could anticipate the move by trading first.

  The worry was that with all the funds pouring money into the market at once—or pulling it out, since there were many ETFs that shorted stocks—a massive, destabilizing cascade could unfold. In a report on the products, Minder Cheng and Ananth Madhavan, two top researchers at Barclays Global Investors, said the vehicles could create unintended consequences and potentially pose systemic risk to the market. “There is a close analogy to the role played by portfolio insurance in the crash of 1987,” they warned.

  Another concern was an explosion in trading volume from computer-driven, high-frequency funds similar to Renaissance and PDT. Faster chips, faster connections, faster algorithms—the race for speed was one of the hottest going. Funds were trading at speeds measured in microseconds—or one-millionth of a second. In Mahwah, New Jersey, about thirty-five miles from downtown Manhattan, the New York Stock Exchange was building a giant data center three football fields long, bigger than a World War II aircraft carrier, that would do nothing but process computerized trades. “When people talk about the New York Stock Exchange, this is it,” NYSE co-chief information officer Stanley Young told the Wall Street Journal. “This is our future.”

  But regulators were concerned. The Securities and Exchange Commission was worried about a rising trend of high-frequency trading firms that were getting so-called naked access to exchanges from brokerages that lent out their computer identification codes. While high-frequency firms were in many ways beneficial for the market, making it easier for investors to buy and sell stocks since there always seemed to be a high-frequency player willing to take the other side of a trade, the concern was that a rogue fund with poor risk-management practices could trigger a destabilizing sell-off.

  “We consider this dangerous,” said one executive for a company that provided services to high-frequency trading firms. “My concern is that the next LTCM problem will happen in less than five minutes.”

  The world of high-frequency trading leapt into the media spotlight in July 2009 when Sergey Aleynikov, a quant who’d just quit a job writing code for Goldman Sachs, stepped off a plane at Newark Liberty Airport after a trip to Chicago. Waiting for him at the airport were FBI agents. Aleynikov was arrested and charged with stealing code from Goldman’s secretive high-frequency trading group, a charge he fought in court.

  Adding to the mystery was a connection to a powerful quant Chicago hedge fund: Citadel. Aleynikov had just taken a position at Teza Technologies, which had recently been founded by Misha Malyshev, who’d been in charge of Citadel’s highly lucrative Tactical Trading outfit. Six days after Aleynikov’s arrest, Citadel sued Malyshev and several of his colleagues—also former Citadel employees—alleging that they’d violated noncompete agreements and could also be stealing code, allegations the defendants denied.

  The suit spelled out previously unknown details about Citadel’s superfast trading operation. The Tactical Trading offices, which required special codes to enter, came equipped with ranks of cameras and guards to ensure no proprietary information was stolen. The firm had spent hundreds of millions to develop the codes over the years and alleged that Malyshev and his cohorts were threatening the investment.

  The suit also revealed that Tactical was a money-making machine, having raked in more than $1 billion in 2008, capitalizing on the market’s volatility, even as Citadel’s hedge funds lost about $8 billion. It raised questions about Griffin’s decision in late 2007 to spin off Tactical from the hedge fund operations, a move that effectively increased his stake in a unit that printed money at a time his investors were getting clobbered. Principals at C
itadel, mostly Griffin, owned about 60 percent of the $2 billion fund, according to people familiar with its finances.

  All of the controversy alarmed regulators and everyday investors, who’d been largely unaware of the lightning-fast trading that had become a central component of the Money Grid, strategies first devised in the 1980s by innovators such as Gerry Bamberger and Jim Simons and furthered in the following decade by the likes of David Shaw and Peter Muller. But there were legitimate concerns that as computer-driven trading reached unfathomable speeds, danger lurked.

  Many of these computer-driven funds were gravitating to a new breed of stock exchange called “dark pools”—secretive, computerized trading networks that match buy and sell orders for blocks of stocks in the frictionless ether of cyberspace. Normally, stocks are traded on public exchanges such as the Nasdaq and the NYSE in open view of anyone who chooses to look. Trades conducted through dark pools, as the name implies, are anonymous and hidden from view. The pools go by names such as SIGMA X, Liquidnet, POSIT, CrossFinder, and NYFIX Millennium HPX. In these invisible electronic pools, vast sums change hands beyond the eyes of regulators. While efforts were afoot to push the murky world of derivatives trading into the light of day, stock trading was sliding rapidly into the shadows.

  Increasingly, hedge funds had been crafting new systems to game the pools, hunting for price discrepancies between them in the eternal search for arbitrage or even causing price changes with dubious tactics and predatory algorithms. Hedge funds were “pinging” the dark pools with electronic signals like submarines hunting prey, searching for liquidity. The behavior was largely invisible, and light-years ahead of regulators.

  Dark pools were also opening up to superfast high-frequency trading machines. The NYFIX Millennium pool had narrowed its response time to client orders to three milliseconds. A flier Millennium sent to potential clients claimed that traders with “ninja skills succeed” in dark pools at “dangerously high speeds for the unprepared.”

  Mom and Pop’s retirement dreams, meet Ninja Hedge Fund.

  Whether such developments posed a broader risk to the financial system was unknown. Users of the technology said faster trades boosted “liquidity,” making trading easier and cheaper. But as the financial panic of 2007 and 2008 had shown, liquidity is always there when you don’t need it—and never there when you do.

  Meanwhile, congressmen, President Obama, and regulators were making noise about reining in the system with new rules and regulations. Progress had been made in setting up a clearinghouse for credit default swaps to keep better track of the slippery contracts. But behind the scenes, the financial engineers were hard at work devising new methods to operate in the shadows.

  Just look: exotic leveraged vehicles marketed to the masses worldwide, hedge funds gaming their returns, lightning-fast computerized trading robots, predatory ninja algorithms hunting liquidity in dark pools …

  Here come the quants.

  Notes

  1 ALL IN

  Simons had pocketed $1.5 billion: Alpha, May 2006.

  That night at the St. Regis: Several details of the poker event were gleaned from MFA News 2, 1 (Spring 2006).

  In 1990, hedge funds held $39 billion: Based on data from Hedge Fund Research, a Chicago research group.

  2 THE GODFATHER: ED THORP

  Just past 5:00 A.M.: I conducted numerous interviews with Ed Thorp and exchanged many emails. Many details about Ed Thorp’s blackjack career, including a description of his foray into blackjack in 1961, were found in his colorful book Beat the Dealer: A Winning Strategy for the Game of Twenty-One (Vintage, 1962).

  Other details were found in the excellent Fortune’s Formula: The Untold Story of the Scientific Betting System That Beat the Casinos and Wall Street, by William Poundstone (Hill and Wang, 2005). I confirmed details I used from this book with Thorp.

  The strategy was from a ten-page article: “Getting a Hand: They Wrote the First Blackjack Book but Never Cashed In,” by Joseph P. Kahn, Boston Globe, February 20, 2008.

  He’d kept his roulette strategy largely secret: “The Invention of the First Wearable Computer,” by Edward O. Thorp (http://graphics.cs.columbia.edu/courses/mobwear/resources/thorp-iswc98.pdf.)

  Science-fiction writer Arthur C. Clarke: Voice Across the Sea, by Arthur C. Clarke (HarperCollins, 1975).

  3 BEAT THE MARKET

  On a typical day of desert sun: Much like the blackjack chapter, many details of this chapter derive from interviews with Thorp, Fortune’s Formula, and Thorp’s second book, Beat the Market: A Scientific Stock Market System. That book is out of print, but Thorp kindly provided a Web-accessible version.

  Huge, sudden leaps: For more than a decade, Nassim Taleb has been criticizing quant models for leaving out huge market events, or black swans, and he deserves much credit for warning about such shortcomings of the models. I had numerous conversations with Taleb while writing this book.

  Gerry Bamberger discovered stat arb: The section on the discovery of statistical arbitrage is based almost entirely on interviews with Gerry Bamberger, Nunzio Tartaglia, and several other members of the original Morgan Stanley group that discovered and spread stat arb across Wall Street. Previous mention of this group can be found in Demon of Our Own Design, by Richard Bookstaber (John Wiley & Sons, 2007).

  Morgan had hired Shaw: The account of Shaw’s departure from Morgan are based on interviews with Nunzio Tartaglia and others who were at APT.

  4 THE VOLATILITY SMILE

  Sometime around midnight: Many details of Black Monday were found in numerous Wall Street Journal articles written during the time, including “The Crash of ’87—Before the Fall: Speculative Fever Ran High in the 10 Months Prior to Black Monday,” by James B. Stewart and Daniel Hertzberg, December 11, 1987.

  Other details, including the description at the opening of the chapter, were found in An Engine, Not a Camera: How Financial Models Shape Markets, by Donald MacKenzie (MIT Press, 2006), and The Age of Turbulence: Adventures in a New World, by Alan Greenspan (Penguin 2007), 105.

  On the evening of September 11, 1976: The best description of the invention of portfolio insurance that I know of can be found in Capital Ideas: The Improbable Origins of Modern Wall Street, by Peter L. Bernstein (John Wiley & Sons, 2005). Another source is “The Evolution of Portfolio Insurance,” by Hayne E. Leland and Mark Rubinstein, published in Portfolio Insurance: A Guide to Dynamic Hedging, edited by Donald Luskin (John Wiley & Sons, 1988).

  “Even if one were to have lived”: The age of the universe is 13.5 billion years, not 20 billion.

  When German tanks rumbled into France: Some details of Mandelbrot’s life come from a series of interviews with Mandelbrot in the summer of 2008. Many also come from the book The (Mis)Behavior of Markets: A Fractal View of Financial Turbulence, by Benoit Mandelbrot and Richard L. Hudson (Basic Books, 2006).

  “I realized that the existence of the smile”: My Life as a Quant, by Emanuel Derman (John Wiley & Sons, 2004), 226.

  A squad of fifty armed federal marshals: Certain details come from Den of Thieves, by James Stewart (Simon & Schuster, 1991).

  He also worked as a consultant: I learned the fascinating story of Thorp’s discovery of the Madoff fraud in several interviews with Thorp in December 2008 as the fraud was discovered. I confirmed his story with the firm involved and through several pertinent documents.

  5 FOUR OF A KIND

  In 1990, Ed Thorp took a call: The details of Thorp’s connection to Citadel was told to me by Thorp, Frank Meyer, Justin Adams, and Ken Griffin.

  Griffin set up shop: I learned a number of details about Griffin’s and Citadel’s history during a single interview with Griffin and numerous interviews with people who worked for him. Other details were found in the following articles: “Citadel’s Griffin: Hedge Fund Superstar,” by Marcia Vickers, Fortune, April 3, 2007; “Citadel Returns 26 Percent, Breaks Hedge Fund Mold, Sees IPO,” by Katherine Burton, Bloomberg News Service, April 29, 2005; “Will a H
edge Fund Become the Next Goldman Sachs?” by Jenny Anderson, New York Times, April 4, 2007.

  When he was ten years old: I learned a number of details about Peter Muller’s life in a series of interviews with people who knew him. Other details, such as the trip to Europe, were taken from an essay he wrote in the book How I Became a Quant, edited by Richard R. Lindsey and Barry Schachter (John Wiley & Sons, 2007).

  The muscular professor strode: I’ve never attended a Fama course, but I did speak with a number of people who took his courses, including Cliff Asness, and I conducted several interviews with Fama. This is a depiction of what his course may have been like and what he may have said based on those interviews.

  As a child, Clifford Scott Asness: I learned a number of details about Asness’s life in a series of interviews with Asness and people who knew him. Other details about Asness’s early life were taken from an essay he wrote in the book How I Became a Quant.

  I conducted numerous interviews with current and former employees of AQR. Other sources include “Beta Blocker: Profile of AQR Capital Management,” by Hal Lux, Institutional Investor, May 1, 2001, and, “The Quantitative, Data-Based, Risk-Massaging Road to Riches,” by Joseph Nocera, New York Times Magazine, June 5, 2005.

  Born near the end of the Great Depression: Fama interview.

  Kendall found no such patterns: Quoted from An Engine, Not a Camera: How Financial Models Shape Markets, 63.