Ask Brad Katsuyama if the stock market is rigged and he jumps from his chair, grabs a black marker and sketches on a whiteboard with the intensity of a head coach whipping up a Hail Mary. He draws squares to represent stock exchange servers and scribbles lines to show how high-frequency trading firms are threading cables through miles of rock, building microwave towers atop mountains and raising armies of computer-science PhDs to outrun the rest of the financial community.
“Basically, someone is trading with tomorrow’s newspaper,” he says from his new offices at 4 World Trade Center, 44 stories above Wall Street. “Not only do they get market data 10 times faster, but because their computers are located inside the exchanges, they can trade hundreds of times before anyone else knows what’s happening.”
Wall Street is in the midst of a billion-dollar drag race. Trading firms are investing huge amounts in advanced algorithms, direct data feeds and ultrafast fibre connections to sprint ahead of the pack via “latency arbitrage”. The goal: Profiting on stock swings before the other players in the market—the funds, banks, exchanges and retail investors—even know prices have changed.
This game—high-frequency bots account for 50 percent of stock trading volume—was played in the shadows until author Michael Lewis tackled the subject with Flash Boys (Norton, 2014), which cast Katsuyama as the good guy with a fix for a broken system. A simultaneous 60 Minutes piece echoed the same themes: Predatory high-speed traders outrunning powerless or unwilling regulators.
But back then the fix was little more than a theory. Fast-forward 18 months and Katsuyama, 37, is actively doing something about it. His company, the IEX Group, has devised a platform to undercut high-speed trading. Backed by $101 million from the likes of big-shot money managers Bill Ackman, David Einhorn and the Capital Group, IEX has quickly captured 1.5 percent of US market volume, with about 95 million shares trading each day.
“Instead of complaining about what was wrong,” says Katsuyama, “we decided to show people there was a different way.’’
That different way still faces entrenched competition. The stock exchanges not only tolerate high-frequency trading—they’ve also come around to profiting from it. Both the New York Stock Exchange (NYSE) and Nasdaq collect big fees through what they euphemistically call “colocation”, which allows high-frequency trading shops to connect their computers directly to exchange servers to get information before the rest of the public.
“The exchanges now determine who has certain technology and data,” says Katsuyama. “That’s not the role of an exchange.” The exchanges counter that their data and connections are available to anyone who wants it, but as with airlines, there’s coach and there’s first-class. “Everyone, from trading firms to retail broker-dealers to mutual funds, covets better services to enhance their ability to trade and make money—we provide that service to customers,” says Walt Smith, the head of Nasdaq’s equity business. (NYSE declined to comment.) Smith argues that fast trading helps investors more than it hurts them. “The data says spreads have never been tighter due to access, liquidity, SEC actions, technology and arbitrage opportunities. Transaction costs have plummeted. Mutual fund trading costs have never been lower. Katsuyama’s rhetoric is not supported by the data.”
Katsuyama’s way of winning the speed game is to slow it down. His secret weapon: A metal box with 38 miles of fibre optics coiled inside that stalls incoming trades in a 350-microsecond (a microsecond is a millionth of a second) holding pattern. This delay neutralises the speed edge needed to game the market the same way speed bumps and humps kill the advantage a Ferrari has over a Ford.
“IEX is a safer and cleaner venue than others,” says David L Brooks, the global head of equity trading at the Boston Co, a $44 billion asset manager. “He’s quickly gained the trust of the buy side.” Big banks, often referred to on Wall Street as the sell side, like it, too. “We’ve done the analysis, and IEX is a lot better” for short-term trading, says Brian Levine, co-head of Goldman Sachs’ global equity trading group. “The whole shoe-box phenomenon has eliminated latency arb opportunities.”
Plaudits result in profits: From a standing start IEX generated $27 million in revenue last year. Katsuyama says he’s already breaking even. But if he reaches his goal of a 10 percent market share by 2020, it’s a good guess that profit margins could be in the range of 30 percent. To hit that market share, he needs to move from a private platform to a fully registered exchange. Katsuyama says the paperwork is nearly ready to file, and he’s hoping for approval from the Securities & Exchange Commission (SEC) by year-end. (With IEX’s speed bump, it could be tricky—current SEC rules ban interfering with an order when it hits an exchange. So IEX has hired John Ramsay, the former SEC head of trading and markets, to aid the effort.)
“Katsuyama is offering something substantially better—more efficient trading,” says Steve Wynn, the casino billionaire who invested in IEX after reading Flash Boys. “He’ll either get a tonne of business from the other guys, or they’re going to copy him. Either way he’s going to make the exchanges change whether they like it or not.”
Changing how Wall Street operates was far from Katsuyama’s mind when he launched his finance career. A mild-mannered Japanese-Canadian, he worked up the ranks at the Royal Bank of Canada (RBC), moving to its New York office to trade energy and tech stocks. Eventually he was earning seven figures running RBC’s New York trading desk. But in 2007 the markets he thought he knew began to change. It was hard to trade the stock he saw on his screens—bids and offers ran away from him the moment he entered an order. It used to be that buying, say, 100,000 shares of Apple at $100 was straightforward. Now he’d get only part of the order filled at the $100 that appeared on his screen, before the price jumped to $100.01.
At first he thought the RBC technology must be slow. But when he asked around Wall Street, he learned the problem was rampant. “If the biggest banks can’t buy or sell what they see on their screens, then something is fundamentally wrong with the market,” he thought. Determined to solve the problem for RBC, he interviewed dozens of people who had worked in high-frequency trading and hired Ronan Ryan, a telecom expert who helped firms shave microseconds off their transaction times. The pair began experimenting.
Two decades ago most trades were executed on the NYSE or Nasdaq. By the mid-2000s the market had fragmented into over a dozen registered exchanges and 30-plus dark pools—opaque alternative markets often run by banks (IEX is technically a dark pool right now, too). So in 2007 the SEC implemented Reg NMS (National Market System) to protect investors by requiring brokers to get their clients best prices across all registered exchanges. Instead of trading stock in one block, the new regulation whacked orders into small pieces that ricocheted around the exchanges. As regulations so often do, NMS created a new problem. Soon enough computer programs were catching order signals and using their ultrafast connections and secret algorithms to jump ahead to the exchanges to make small but risk-free profits.
The RBC team (which included IEX co-founders Rob Park and John Schwall) built software, dubbed Thor, that could hide RBC’s hand by making its trades arrive simultaneously at each exchange. That kept the high-frequency traders’ software from detecting trades and racing ahead of them. Katsuyama had come up with a fix—but couldn’t do much with it. Housed in RBC, Thor was off limits to big banks like Goldman and Bank of America. And since money managers—the buy side—used stock orders as a way to pay banks commissions for services like research and investment banking, the funds wanted to spread their orders across Wall Street rather than send them all to RBC. The only way to scale was to become an independent platform.
Using $1.5 million in seed money from their friends and their own savings, Katsuyama and Ryan bolted RBC in March 2012, huddling in a closet of an office in One Liberty Plaza to create a solution that wouldn’t infringe on RBC’s ownership of Thor. Since distance was the biggest killer of speed, they developed the magic shoe box as a way of creating remoteness while staying in the financial epicentre of Manhattan.
“It’s opposite what exchanges were doing,” says Katsuyama. “Instead of bringing people as close as you can to your [stock order] matching engine, we were pushing everyone as far away as we could.” In March 2013, with their solution in hand, they raised $24.4 million from asset managers as frustrated with the markets as Katsuyama had been at RBC, including Belfer Management, Capital Group and hedgies like Ackman and Einhorn.
A year later they hit marketing gold with Michael Lewis’ Flash Boys and the 60 Minutes episode. Katsuyama watched the show in his New York apartment with his wife. As the segment ended, his phone exploded. “Guys were screaming, calling me a traitor, promising I’d never work on The Street again,’’ he recalls. It wasn’t all haters. Five hundred résumés hit his email that night. Then guys like Wynn and Spark Capital’s Alex Finkelstein called to invest. By the fall Katsuyama had $100 million in ammo to take on the predators.
It will remain a tough fight. Katsuyama’s whole pitch is that the current system is unfair—yet if he’s to succeed on a large scale, he must win over people who currently run the system. “He’s got this delicate PR problem, because he risks insulting his customers,” says Michael Lewis. “After setting off this stink bomb in the industry, he has to clean up some of the smell.” The cleanup involves bringing the sell side—the Goldmans, BofAs and Citis—into his camp. “If you want to build a scalable solution and really make a change in the market, you can’t leave out the brokers,” says Ronan Ryan. “You need to bring together the buy side, the brokers and the exchange.”
IEX markets itself as a transparent, fair exchange. The revenue model is simple—nine-hundredths of a penny for every share traded. IEX bans colocation and special connections. “We’re not going to charge you for every port and cable and data feed,” says Ryan. “We will charge you for buying and selling stock. The brokers respect that.”
And then there are the high-frequency clients. The debate following Flash Boys was whether high-speed trading was good or bad. Of course it’s more complicated than that. Many such firms act as market-makers, adding valuable liquidity to the system, in many ways making it better for investors than ever. Virtu—a high-frequency firm whose IPO was delayed a year because of the Flash Boys furore—is one of IEX’s biggest clients. Running an exchange is expensive, and IEX can’t compete with the big US exchanges without attracting more high-frequency trading volume. From IEX’s perspective, the trick is to deter the abusive bots while attracting the beneficial ones. “Some love us because we keep away the predatory guys. Other firms hate us,” says Katsuyama. “If you’re just making markets, you shouldn’t care about what we’ve built.”
Ultimately, the success of IEX hangs on winning that coveted exchange status. Then brokers wouldn’t be able to bypass it. (Reg NMS requires that traders would then have to use it if it has shares priced better or equal to the other exchanges.)
This alone will greatly increase volume. Second, as an exchange, IEX could list company stock. That opens up the potential for high-margin listing fees (the NYSE gets about 10 percent of its revenue this way) and for capturing the heavy volume that trades when the market opens and closes. Steve Wynn says he’ll move Wynn Resorts shares over and plans on spreading the word to other casino owners like Las Vegas Sands’ Sheldon Adelson. Katsuyama even thinks IEX’s mission to use technology to promote openness will help woo Silicon Valley companies like Apple, Google and Facebook. A move from any of those companies could spark a major migration to IEX.
In that effort Katsuyama could get a big boost from Hollywood. Screenwriter Aaron Sorkin is in talks to write a Flash Boys script for the big screen. “Guys tell me Jackie Chan should play me,” says Katsuyama. “Dude, Jackie Chan’s like 60 years old and knows karate.” He’s also made a tonne of money.