Flash Boys: A Wall Street Revolt

by Michael Lewis

  With the purposelessness of the exercise hanging in the air, Rob said, “I just had a sick idea.” Rob’s idea was to license the technology to one of the exchanges. (Schwall had patented Thor for RBC.) The line between Wall Street brokers and exchanges had blurred. The big Wall Street banks now ran their own private exchanges. The stock exchanges, for their part, were making a bid to become brokers. The bigger ones now offered a service that enabled brokers to simply hand them their stock market orders, which they would then route. To their own exchange, of course, but also to others. The service was used mainly by small regional brokerage firms that didn’t have their own routers, but this brokerage-like service opened up, at least in Rob’s mind, a new possibility. If just one of the exchanges was handed the tool for protecting investors from market predators, the small brokers from around the country might flock to it, and it might become the mother of all exchanges.

  “Screw that,” said Brad. “Let’s just create our own stock exchange.”

  “We just sat there for a while,” said Rob. “Kind of staring at each other. Create your own stock exchange. What does that even mean?”

  A few weeks later Brad flew to Canada and sold his bosses on the idea of an RBC-led stock exchange. Then, in the fall of 2011, he canvassed a handful of the world’s biggest money managers (Janus Capital, T. Rowe Price, BlackRock, Wellington, Southeastern Asset Management) and some of its most influential hedge fund managers (David Einhorn, Bill Ackman, Daniel Loeb). They all had the same reaction. They loved the idea of a stock exchange that protected investors from Wall Street’s predators. They also thought that a new stock exchange, to be credibly independent of Wall Street, could not be created by a Wall Street bank. Not even a bank as nice as RBC. If Brad wanted to create the mother of all stock exchanges, he would need to quit his job and do it on his own.

  The challenges were obvious. He’d need to find money. He’d need to persuade a lot of highly paid people to quit their Wall Street jobs to work for tiny fractions of their current salaries—and possibly even supply the capital to pay themselves to work. “I was asking: Can I get the people I need? How long can we survive without getting paid? Will our significant others let us do this?” He also needed to find out if the nine big Wall Street banks that controlled nearly 70 percent of all stock market orders** would be willing to send those orders to a truly safe exchange. It would be far more difficult to start an exchange premised on fairness if the banks that controlled the vast majority of the customers’ orders were committed to unfairness.

  For a surprisingly long time, Brad had reserved final judgment about the biggest Wall Street banks. “I held out a degree of hope that the people at [each] bank who handled the clients’ orders were removed from the prop group,” he said. His hope sprang mainly from his own experience: At RBC, where he handled the clients’ orders, he barely knew the prop traders and had no idea what they were doing. There was a reason for this: RBC had not created a dark pool, because Brad had killed the idea. Still, he knew that each of the big Wall Street banks had its own internal politics, and that there were people in each of them who wanted to act in the long-term interests of their firms and do the right thing by their customers. His hope was that some of these people, in some of these places, had power.

  John Schwall’s private investigations put an end to that hope. By the fall of 2011 Schwall had become something like a connoisseur of the uses of LinkedIn to find stuff out about people in and around high-frequency trading. He’d put a face on high-frequency trading, or rather two faces. “I began to anticipate that certain people were in on the game,” said Schwall. “I’d connect to them so that I could see their network. There were maybe twenty-five guys I called kingpins—the people who actually knew what was going on.” At the very top of the food chain were a lot of white guys in their forties whose careers could be traced back, one way or another, to the early electronic stock exchanges born of the regulations passed after the crash of 1987—Wall Street guys who might have some technical background but whose identity was more trader than programming geek.

  The new players in the financial markets, the kingpins of the future who had the capacity to reshape those markets, were a different breed: the Chinese guy who had spent the previous ten years in American universities; the French particle physicist from FERMAT lab; the Russian aerospace engineer; the Indian PhD in electrical engineering. “There were just thousands of these people,” said Schwall. “Basically all of them with advanced degrees. I remember thinking to myself how unfortunate it was that so many engineers were joining these firms to exploit investors rather than solving public problems.” These highly trained scientists and technicians tended to be pulled onto Wall Street by the big banks and then, after they’d learned the ropes, to move on to smaller high-frequency trading shops. They behaved more like free agents than employees of a big corporation. In their LinkedIn profiles, for instance, they revealed all sorts of information that their employers almost certainly would not want revealed. Here Schwall stumbled upon the predator’s weakness: The employees of the big Wall Street banks felt no more loyalty toward the banks than the banks felt toward them.

  The employees of Credit Suisse offered the clearest example. Credit Suisse’s dark pool, Crossfinder, vied with Goldman Sachs’s Sigma X to be Wall Street’s biggest private stock exchange. Credit Suisse’s biggest selling point to investors was that it put their interests first and protected them from whatever it was that high-frequency traders were doing. Back in October 2009, the head of Advanced Execution Services (AES) at Credit Suisse, Dan Mathisson, had testified before a U.S. Senate Banking, Housing, and Urban Affairs Committee at a hearing on dark pools. “The argument that dark pools are somehow part of the high-frequency trading debate simply does not make sense,” he’d said. “High-frequency traders make their money by digesting publicly available information faster than others; dark pools hide order information from everyone.”

  That, Schwall thought, because Brad had explained it all to him, was simply wrong. It was true that when, say, a pension fund gave a Wall Street bank an order to buy 100,000 shares of Microsoft, and the Wall Street bank routed the order to the dark pool, the wider world was not informed. But that was just the beginning of the story. The pension fund did not know the rules of the dark pool, and could not see how the buy order was handled inside of it. The pension fund would not be able to say, for example, whether the Wall Street bank allowed its own proprietary traders to know of the big buy order, or if those traders had used their (faster than the dark pool) market connections to front-run the order on the public exchanges. Even if the Wall Street bank resisted the temptation to trade for itself against its own customers, there was virtually no chance they resisted the temptation to sell access to the dark pool to high-frequency traders. The Wall Street banks did not disclose which high-speed trading firms had paid them for special access to their dark pools, or how much they had paid, but selling that access was standard practice.

  Raising, again, the obvious question: Why would anyone pay for access to the customers’ orders inside a Wall Street bank’s dark pool? The straight answer was that a customer’s stock market order, inside a dark pool, was fat and juicy prey. The order was typically large, and its movements were especially predictable: Each Wall Street bank had its own detectable pattern for handling orders. The order was also slow, because of the time it was forced to spend inside the dark pool before accessing the wider market. As Brad had put it, “You could front-run an order in a dark pool on a bicycle.” The pension fund trying to buy 100,000 shares of Microsoft could, of course, specify that the Wall Street bank not take its orders to the public exchanges at all but simply rest it, hidden, inside the dark pool. But an order hidden inside a dark pool wasn’t very well hidden. Any decent high-frequency trader who had paid for a special connection to the pool would ping the pool with tiny buy and sell orders in every listed stock, searching for activity. Once they’d discovered the buyer of Microsoft, they’d simply wait
for the moment when Microsoft ticked lower on the public exchanges and sell it to the pension fund in the dark pool at the stale, higher “best” price (as Rich Gates’s tests had demonstrated). It was riskless, larcenous, and legal—made so by Reg NMS. The way Brad had described it, it was as if only one gambler were permitted to know the scores of last week’s NFL games, with no one else aware of his knowledge. He places bets in the casino on every game and waits for other gamblers to take the other side of those bets. There’s no guarantee that anyone will do so; but if they do, he’s certain to win.

  In his investigation of the people who managed Credit Suisse’s dark pool, one of the first things Schwall noticed was the guy in charge of electronic trading: Josh Stampfli, who had joined Credit Suisse after seven years spent working for Bernie Madoff. (Madoff had pioneered the idea of paying brokers for the right to execute the brokers’ customers’ orders, which should have told people something but apparently did not.) This, of course, only heightened Schwall’s suspicions, and sent him digging around in old articles in trade journals about Credit Suisse’s dark pool.†† There he found references and allusions that made sense only if Credit Suisse had planned, right from the start, to be deeply involved with high-frequency trading firms. For instance, in April 2008 a guy named Dmitri Galinov, a director and the head of liquidity strategy at Credit Suisse, had told the Securities Technology Monitor that many of Credit Suisse’s “clients” had placed computer servers in Weehawken, New Jersey, to be closer to Credit Suisse’s dark pool. The only people who put servers next to dark pools in Weehawken were Ronan’s old clients—the high-frequency trading firms. No stock market investor went to such lengths to shave microseconds off trading time.

  “Client,” to Credit Suisse, appeared to Schwall to be a category that included “high-frequency trading firms.” Schwall’s suspicion that Credit Suisse wanted to service HFT while not seeming to do so grew after he read an interview Dan Mathisson gave to the New York Times in November 2009.

  Q: Who are your clients at CrossFinder [sic] and how do they benefit from using a dark pool as opposed to just going through a broker and trading on the exchange?

  A: Our clients are mutual funds, pension funds, hedge funds and some other large broker-dealers, so it is always institutional clients . . .

  All the large high-frequency trading firms, Schwall knew, were “broker-dealers.” They had to be, to gain the special access they had to the public stock exchanges. So Mathisson had not ruled out dealing with them. The only reason he would not explicitly rule out dealing with them, Schwall assumed, was that he was dealing with them.

  The LinkedIn searches became a new obsession. The former Madoff employee’s profile led him to the people who worked for the former Madoff employee, who led him to the people who worked for them, and so on. Even as Credit Suisse tried to appear as if it had nothing to do with high-frequency trading, its employees begged to differ. Schwall dug out dozens of examples of Credit Suisse’s computer programmers boasting on their résumés about “building high-frequency trading platforms” and “implementing high-frequency trading strategy,” or of experience as a “quantitative trader on equity and equity derivatives: high-frequency trading.” One guy explained that he had “managed on-boarding of all high-frequency clients to Crossfinder.” Another said he had built the Credit Suisse Crossfinder dark pool and now worked in high-frequency trading market making. Credit Suisse claimed that its dark pool had nothing to do with high-frequency trading, and yet it somehow employed, in and around its dark pool, a mother lode of high-frequency trading talent.

  By the time he’d finished, Schwall had built the entire Credit Suisse dark pool organization chart. “He’s got these people charts,” said Brad incredulously. “It’s like one of those FBI boards, with the drug kingpins.” Looking over Schwall’s charts on Credit Suisse, the bank that went to the most trouble to sell itself as safe to investors, Brad decided that the game was probably over inside all the big Wall Street banks. All of them, one way or another, were probably using the unequal speeds in the market to claim their share of the prey. He further assumed that the big Wall Street banks must have stumbled upon his solution to high-frequency front-running, and must have chosen not to use it, because they had too great a stake in the profits generated by that front-running. “It became very obvious to me why we were the first to discover Thor, because we weren’t,” he said. “What that meant to me was that the problem was going to be much, much harder to solve. It also told me why the clients were so in the dark, because the clients rely on brokers for information.” Creating an exchange designed to protect the prey from the predator would mean starting a war on Wall Street—between the banks and the investors they claimed to represent.

  Schwall’s private investigations also revealed to Brad just how little the technical people understood of their role in the financial world. “It’s not like you are building a bridge connecting two pieces of land,” he said. “You can’t see the effects of what you are doing.” The openness with which the Credit Suisse technologists described their activities made him aware of a larger, almost charming obliviousness. “I was totally shocked when John started to pull out these résumés,” he recalled. “The banks had adopted a policy of saying as little as possible about what they were actually doing. They’d fire people for being quoted in the newspaper, but in their LinkedIn pages those same people said whatever they wanted.” From the way the engineers described their roles in the new financial system, he could see that they had no clue about the injustices of that system. “It told me that these tech guys were completely oblivious to what they were working on,” he said. “They were tying these things they were working on—helping the bank to make markets in their dark pools; building automated systems for the bank to use with its customers—in a way you never would if you understood what the banks were doing. It’s like saying on your LinkedIn profile, ‘I have all the skills of a robber and I know this one house intimately.’ ”

  Schwall had started out looking for the villains who were committing crimes against the life savings of ordinary Americans, fully aware of their own villainy. He wound up finding, mainly, a bunch of people who had no idea of the meaning of their own lives. In his searches, Schwall noticed something else, though at first he didn’t know what to make of it: A surprisingly large number of the people pulled in by the big Wall Street banks to build the technology for high-frequency trading were Russians. “If you went to LinkedIn and looked at one of these Russian guys, you would see he was linked to all the other Russians,” said Schwall. “I’d go to find Dmitri and I’d also find Misha and Vladimir and Tolstoy or whatever.” The Russians came not from finance but from telecom, physics, medical research, university math departments, and a lot of other useful fields. The big Wall Street firms had become machines for turning analytically minded Russians into high-frequency traders. Schwall filed that fact away for later, as something perhaps worth thinking about.

  _____________

  * It is irritating to read about an American bank that insists on calling itself a banc. The banc in this case was pushed to do so, as the securities divisions within American banks (here, Bank of America) are prohibited by regulators from referring to themselves as banks.

  † A year later, in 2012, Wall Street Journal reporter Scott Patterson would write an excellent history of the early electronic traders called Dark Pools.

  ‡ “There’s a culture in the SEC of not getting into a dialogue with any individual who comes in,” says a staffer who listened to Brad Katsuyama’s presentation. “They don’t want to give any one person an unfair peek at the way the SEC thinks. But it’s a very defensive culture. And there were people in the room who had written some of the rules he was implicitly criticizing.”

  § In early 2013, one of the largest high-frequency traders, Virtu Financial, publicly boasted that in five and a half years of trading it had experienced just one day when it hadn’t made money, and that the loss was caused by “human er
ror.” In 2008, Dave Cummings, the CEO of a high-frequency trading firm called Tradebot, told university students that his firm had gone four years without a single day of trading losses. This sort of performance is possible only if you have a huge informational advantage.

  ¶ A former employee of Citadel who also once had top secret security clearance at the Pentagon says, “To get into the Pentagon and into my area, it took two badge swipes. One to get into the building and one to get into my area. Guess how many badge swipes it took me to get to my seat at Citadel? Five.”

  ** Those nine banks, in order of their (fairly evenly distributed) 2011 market share, from highest to lowest: Credit Suisse, Morgan Stanley, Bank of America, Merrill Lynch, Goldman Sachs, J.P. Morgan, Barclays, UBS, Citi, Deutsche Bank.

  †† Stampfli has not been charged with any wrongdoing.

  CHAPTER FIVE

  PUTTING A FACE ON HFT

  Sergey Aleynikov wasn’t the world’s most eager immigrant to America, or, for that matter, to Wall Street. He’d left Russia in 1990, the year after the fall of the Berlin Wall, but more in sadness than in hope. “When I was nineteen I haven’t imagined leaving it,” he says. “I was very patriotic about Russia. I cried when Brezhnev died. And I always hated English. I thought I was completely incapable of learning languages.” His problem with Russia was that its government wouldn’t allow him to study what he wanted to study. He wasn’t religious in any conventional sense, but he’d been born a Jew, which had been noted on his Russian passport to remind everyone of the fact. As a Jew he expected to be given especially difficult entrance exams to university, which, if he passed them, would grant him access to just one of two Moscow universities that were more accepting of Jews, where he would study whatever the authorities permitted Jews to study. Math, in Serge’s case. He’d been willing to tolerate this state of affairs; however, as it happened, he’d also been born to program computers. He hadn’t laid hands on a computer until 1986, when he was already sixteen. The first thing he’d done was to write a program: He instructed the computer to draw a picture of a sine wave. When the computer actually followed his instructions, he was hooked. What hooked him, he said, was “its detailed orientation. The way it requires an ability to see the problem and tackle it from different angles. It’s not just like chess, but like solving a particular problem in chess. The more challenging problem is not to play chess but to write the code that will play chess.” He found that coding engaged him not just intellectually but also emotionally. “Writing a program is like giving birth to a child,” he said. “It is a creation. Even though it is technical, it is a work of art. You get this level of satisfaction.”