'Flash Boys' reignites debate: Is high-frequency trading a digital age menace?

'Flash Boys: A Wall Street Revolt' focuses on a particular technique in high-frequency trading, in which automated algorithmic traders use their superior speed to their advantage.

Insider traders on Wall Street, like the shadowy dons of Mafia lore, have always fascinated the American public.

From Gordon Gekko’s “greed ... is good” in Oliver Stone’s “Wall Street” to Jordan Belfort’s rags-to-decadent-riches in Martin Scorsese's “The Wolf of Wall Street,” there’s always been a mix of horror and awe at those bold but nefarious characters who learn to game the system and reap a ton of cash.

But according to a highly publicized book released this week, infamous insider traders like Ivan Boesky are being replaced by a new digital age menace: the algobot.

That is, the algorithmic and automated robot trader – proprietary software programs that are fast replacing their human counterparts and that now make up half the volume of trades in stock exchanges across America. Referred to as “high-frequency trading,” these algobots can analyze markets and, in the literal blink of an eye, buy and sell thousands of securities.

Controversy about the dangers of high-frequency trading is nothing new. But this week’s release and relentless publicity of “Flash Boys: A Wall Street Revolt,” in which bestselling author Michael Lewis claims that parasitic traders armed with lightning-quick computers have “rigged” the markets in their favor, has re-sparked an often explosive debate on Wall Street.

New York Attorney General Eric Schneiderman, as well as the Federal Bureau of Investigation and the Securities and Exchange Commission, is now investigating the high-tech wizardry of high-frequency trading, which Mr. Schneiderman has called “Insider Trading 2.0.”

Mr. Lewis’s book pays special attention to a particular technique of these savvy automated algorithmic traders, in which they use their superior speed to spot a slower trader about to make a trade. In about the time it takes to click a mouse, the algobot can snatch away the shares other traders are trying to buy, and then offer them back to the traders at a higher price.

“They’ve figured out how to front-run market orders,” says Larry Doyle, a former mortgage-backed securities trader who wrote “In Bed With Wall Street: The Conspiracy Crippling Our Global Economy,” a scathing critique of Wall Street banks and regulators. Most “front running” is illegal, especially when brokers use information about their own clients’ imminent trades to gain an unfair advantage in predicting the market.

But algobots don’t use client information. They simply “see” other traders preparing to order a stock at a certain price, and in a fraction of a second before they can act, they buy it first. They then sell it back to the original trader at a higher price and collect a transaction fee in what is basically a risk-free technique.

This is not front running, proponents contend. This is just a better and faster way to analyze the market and react. And on Wall Street now, each millisecond is money.

These higher prices are often just a penny or two higher, they say, and the transaction fee is often just a fraction of penny. High-frequency traders make money in the sheer volume of these thousands-per-second trades – doing it millions of times a day.

“In general, somebody here has just invested in a better mousetrap,” says Charles Jones, professor of finance at Columbia Business School in Manhattan and an expert on high-frequency trading. “So why shouldn’t they get a return on that better mousetrap?”

These better mousetraps may actually be making the markets more efficient, cutting trading costs and creating greater “liquidity” in the markets – that is, more cash to spend on investments.

“Bid-ask spreads are far smaller for most stocks,” says Professor Jones, who has analyzed the effects of high-frequency trading. “You won’t pay more than a penny in terms of a markup to buy a share or sell a share. It’s amazing if you think back to where we were 20 years ago. And commissions, they’re all under $10 for as many shares as you want to trade.”

He adds, “So in a sense, there’s never been a better time to be a retail trader than in this automated world with high-frequency traders.”

Even firms who do not use algobot trading have risen up to support the technique. “How do we feel about high-frequency trading? We think it helps us,” wrote Clifford Asness and Michael Mendelson, executives at AQR Capital Management in Greenwich, Conn., in The Wall Street Journal Tuesday. “It seems to have reduced our costs and may enable us to manage more investment dollars.”

But the smartest “algos” in the room get help, critics say. The big stock exchanges like Nasdaq and the New York Stock Exchange cater to them, offering incentives for high-speed trading in a kind of “pay to play” system for faster access to market information.  

“As for-profit enterprises, what they’re trying to do now is, how do we attract volume to our exchanges?” Mr. Doyle says. “So the way that they attract volume is, they cut and negotiate deals with different trading entities and these high-frequency traders.”

Such deals include letting companies place computer servers within the exchange’s trading venues and providing extra bandwidth, ultrafast cables, and high-speed switches for high-frequency traders.

“Each of these services offers clients a timing advantage – often in milliseconds – that allows high-frequency traders to make rapid and often risk-free trades before the rest of the market can react,” Schneiderman said last fall. “As a result, these traders guarantee themselves enormous revenue and force large investors to develop complicated and expensive defensive strategies to conceal their orders from parasitic traders.”

These defensive strategies in turn have led to the proliferation of “dark pools” – alternative trading venues in which financial institutions can conduct business outside the public exchanges. These defensive and behind-the-scenes trading pools, which attempt to hide from front-running algobots, are far less regulated and have far less transparency. They involve only a few traders agreeing to prices outside the broader market.

“Today I think we need to revisit some issues relating to this much-litigated and scrutinized territory because we’re seeing something far more insidious than traditional insider trading,” Schneiderman also said last fall. “Small but powerful groups within the market are able to use soon-to-be public information combined with high-frequency trading in a way that distorts our markets far more than ... Ivan Boesky or even Gordon Gekko could ever have imagined.”