Under the cover of night, the firm Spread Networks was secretly laying an 827-mile fiber optic cable connecting Chicago and Northern New Jersey in the straightest line possible. The workers, who were told to keep quiet and avoid asking questions of their employer, ran the line under rivers and even through miles of hard rock in the Allegheny Mountains of Western Pennsylvania.
This $300 million project was designed to connect the financial markets of Chicago and New York City, the largest financial derivatives and stock exchanges in the country. This enormous and massively risky project was intended to reduce the journey of electronic information from 17 milliseconds to 13 milliseconds.
The four-millisecond advantage of the new line, just one tenth of the time it takes to blink an eye, was completely revolutionary for a new breed of Wall Street people known as high-frequency traders (HFTs). When informed of the existence of this new line, the large HFT firms unanimously shelled out $14 million apiece to gain access to the line. It was a financial arms race and the stakes were high. Arriving just four milliseconds late would literally cost billions of dollars in profits.
The Flash Crash
Michael Lewis’s journey to understanding this new and exceptionally secretive variety of trader inspired him to write Flash Boys. The popular financial journalist and best-selling author of Moneyball, Liar’s Poker, and The Blind Side wrote Flash Boys in what has become the first exposé on HFT. The book has inspired massive pushback from Wall Street; Lewis himself described the reaction in financial circles as a “shitstorm.”
All the way back to its inception in 1999, high-frequency trading remained fairly unknown until the Flash Crash of 2010. As Bloomberg describes, May 6 of that year was relatively quiet in the American stock market until a sizeable mutual fund put in a large sell order on a series of stock market index futures contracts. Within minutes, the Dow Jones Industrial Average had plunged about 1000 points (approximately 9 percent of its total value) and just as quickly regained almost all the losses that had occurred. It was the largest single-day point decline in the history of Dow.
High-frequency firms would usually help abate the sell pressure from the mutual fund order. However, on this day, high market volatility led the HFT firms to sell their stake in the same index futures, thereby magnifying the effect of the initial sell order from the mutual fund. The algorithms used by both the HFT firms and the mutual fund responded to this negative stimulus with further pressure, effectively acting as a positive feedback loop. Eventually, rational investors realized what was occurring and returned the market to the more natural equilibrium.
What is High-Frequency Trading?
High-frequency traders, as defined by The New York Times, make money in a variety of ways but are all united by a reliance on algorithmic electronic trading, the process of initiating trades using programmed computers. These firms are very influential and have grown to account for approximately 73 percent of total daily market volume on U.S. exchanges. This large volume share leads critics to argue that HFT is responsible for events like the Flash Crash and overall greater market volatility. While supporters deny this claim, there is evidence supporting a strong correlation between high volatility and HFT market activity.
HFT firms rely mostly on the strategy of statistical arbitrage—riskless profit—opportunities in liquid securities including equities, bonds, futures, and foreign exchange contracts. Sometimes this can simply be a price discrepancy between one exchange and another; other times, more complex computer algorithms are used to “front-run” market orders. As Lewis states, HFT strategies have produced unfair market environment that he describes as “rigged.”
In Flash Boys, Lewis describes how the strategy of front-running orders arose from well-intentioned but ineffective policy decisions by the federal government. The SEC deregulated electronic trading in 1999 ending the oligopoly that a small number of exchanges had on American equity trading. Instead of a few distinct exchanges, there were now thirteen exchanges that look nothing like the antiquated vision of the stock market that persists in the popular imagination. As Lewis explains in Flash Boys, “Over the past decade, the financial markets have changed too rapidly for our mental picture of them to remain true to life.” Instead of busy men in jackets running around with slips of paper, the stock exchanges today are simply stacks of computer servers in Northern New Jersey.
The second change that enabled the existence of HFT on a larger scale was the SEC’s implementation of Reg NMS in 2007, which legally required stock brokers to seek the best price available for their clients regardless of which exchange it was found on. If a large order were being placed, some shares would be purchased on the cheapest exchange before moving onto the next exchange and then the next until the entire order could be filled. In the case of a large institutional investor like a pension fund, a market order for 100,000 shares would likely be filled on a half-dozen or more exchanges, each trade occurring milliseconds after the last
As an example, let’s assume the California Public Employees’ Retirement System (CALPERS), a large pension fund, is interested in buying 100,000 shares of General Electric at $25.00 a share. The fund then instructs its broker to make the purchase. The broker plugs the trade into the computer, and a matching algorithm is used to purchase these shares at the cheapest price it can find on the market. Once 15,000 shares are purchased on the NYSE, the HFT firms identify the large trade that has hit the first exchange and rush ahead to the other exchanges to purchase shares and drive the price up before selling the shares a fraction of a second later. To the human eye, the stock market seems to move at a frantic, almost constant pace. However, to a computer operating at a millisecond pace, Lewis says the stock market landscape at the time of the CALPERS purchase looks like a massive “obelisk rising from a desert.” In fact, in 98.22 percent of all microseconds (one millionth of a second) there is no market activity whatsoever.
Once CALPERS arrives at the other exchanges searching for 85,000 shares, it may be forced to buy at the new market price of $25.02, at which point HFT firms sell their recently acquired shares pocketing the spread of two cents per share. This front-running strategy is the bread and butter of HFT and is responsible for billions in profits.
The Knightmare
Throughout Flash Boys, Lewis refers to HFT as a riskless practice that will consistently produce large profits. He provides the example of the HFT firms that never had a single losing day—ever. Lewis’s assertion is mostly right. However, HFT is extremely complex and thus presents a good amount of institutional risk that does not appear on the day-to-day balance sheet.
This is most clear in the case of Knight Capital, one of the giants of both HFT and the stock market in general. Knight accounted for 17 percent of all volume on the NYSE before a computer malfunction caused Knight to unintentionally start an aggressive stock buy. Within minutes, Knight had accidentally purchased $7 billion in stock. Each purchase raised the price of the security slightly, which inspired other HFT firms to jump in and sell to their struggling competitor. The company’s trading system was shut down and the company lost $400 million trying to liquidate its accidental positions. In the wake of this disaster, Knight Capital was acquired by another HFT giant in Chicago-based Getco. This incident indicates that, while HFT presents little risk in the short term, Lewis fails to acknowledge the large-scale risk that does exist.
However profitable it has been in the past, HFT is becoming a saturated market as more and more participants are nipping at profits that used to only be experienced by a few firms. Traditional HFT strategies have become very difficult to use profitably in a crowded industry. In fact, Frederi Viens of Purdue University estimates that traditional HFT profits have declined from a peak of $5 billion in 2009 to about $1.25 billion today. The top HFT firms are adapting to this change by decreasing their activity in the equity market while transitioning into the futures and derivatives markets.
Other HFT firms are adapting in eccentric ways, as Bloomberg has noted; indeed, some are abandoning statistical arbitrage altogether. Many of these firms have moved onto risky momentum trading using technical indicators to try and predict security prices by observing short-term trends. Still other firms are using sophisticated algorithms to analyze news headlines and Twitter feeds, hoping to get ahead of the human reaction to market-altering news.
Who’s Going to Prison?
The front-running strategy seems to occupy a legal grey area; few if any laws that govern these practices. However, it seems the HFT firms have all acted within the law up to this point. Indeed, HFT firms are less likely than exchanges and banks to be punished for facilitating HFT practices.
In Flash Boys, Lewis’s criticism isn’t just reserved for the public exchanges, but also for the major banks offering an equity-trading forum. He describes how, a few years ago, many institutional investors making large trades on the market started looking for a confidential way to trade securities. Soon, banks began providing an alternative route for trading securities without requiring a public exchange. These private exchanges became known as “dark pools”—private forums for security trading often facilitated by a major bank. Since their recent inception, up to 40 percent of all stock market transactions now take place in a dark pool, a far cry from 16 percent only six years ago. The trade secrecy that dark pools provided was supposed to allow firms to transact large blocks of securities without the general investing public knowing, thereby avoiding a market impact.
In theory, dark pools are a well-constructed solution to the problem of high frequency trading. However dark pools are just as corruptible as the public exchanges when it comes to the distribution of investor information. Dark pools and exchanges both sell the right to receive confidential trading information just milliseconds before anyone else for hundreds of millions of dollars annually. This process of preferred exchange access is self-evidently unfair, yet not explicitly illegal.
Although most HFT practices exist within the confines of the law, some practices are drawing attention from regulators. For example, popular and active New York State Attorney General Eric Schneiderman has subpoenaed many security exchanges and HFT firms looking for wrongdoing. Additionally the NYSE agreed to pay a $4.5 million fine to the SEC to settle charges that it violated regulations on co-location—the practice of allowing HFT firms to locate computers on the same premises as the exchange’s servers so as to receive and act on trading information even faster.
Lewis takes a very pessimistic approach toward regulation of high-frequency trading, assuming that any governmental action will only distort incentives and create further opportunities for HFT profit. Instead, the majority of Flash Boys details the journey of Brad Katsuyama, the Canadian moral champion determined to bring fairness back into the stock market. Katsuyama quit his million-dollar job at the Royal Bank of Canada to start his own fully transparent stock exchange, IEX. This is the first equity-trading venue to try to combat the effect of HFT distortion by creating a 350-microsecond delay that would prevent front-running.
In addition to private solutions, there are proposed governmental answers to HFT that Lewis doesn’t acknowledge in the book. For example, U.S. Senator Charles Schumer (D-N.Y.) urged the Securities and Exchange Commission in July 2009 to ban preferred exchange access, indirectly banning front-running. Other countries, such as Italy, are taxing HFTs at a miniscule rate, in this case a .002 percent levy on transactions lasting less than half a second.
How to Move Forward
The release of Flash Boys has drawn enormous attention to HFT and has greatly shifted the conversation. Today, most informed people hold a negative view of HFT even though it has lesser-known positive effects. The supporting argument for HFT, enumerated by The New York Times, boils down to a single word: liquidity. Supporters argue that HFT provides a necessary grade of financial intermediation in capital markets by more easily matching buyers and sellers. According to this view, HFT moves supply and demand more quickly and efficiently. Additionally, the large volume of HFT trades has put downward pressure on the bid-ask spread. By minimizing its market-altering effect, HFT allows more retail investors to participate in the market at a lower cost.
At the end of the day, high-frequency trading is controversial for a few practical reasons. It is very clearly a zero-sum game in which profits are taken from the slower firm by the faster firm. Unlike traditional investing in which capital is invested in worthy firms in order to spur their growth, HFT make no value judgment on the securities they invest in. HFT is simply a mechanism of wealth distribution: from one hedge fund to another, or from one pension fund to another, albeit indirectly.
Many critics of HFT are quick to point out that some firms and individuals have been enriched by preying on the pension plans and mutual funds of regular people as in the CALPERS example. Ken Griffin, the manager of a large Chicago hedge fund, and historic Harvard donor, is the best proof of this argument. Mr. Griffin’s firm Citadel Capital has made billions of dollars front-running the purchases of large institutional investors, many of which are investing on behalf of middle-class clients.
Another poignant criticism of HFT has nothing to do with equality and everything to do with efficiency. The greatest tragedy of high-frequency trading may simply be the wasted capital, both physical and human, in the quest for arbitrage profit. The $300 million cable from Chicago to New York added no tangible societal benefit despite its price tag. Wall Street firms have accelerated their recruiting of the best academic and technological talent in the country in order to run HFT groups, often siphoning these employees from universities and productive businesses.
Perhaps the incentive system needs to be realigned so that skilled computer engineers spend time developing electronic medical records systems instead of algorithms to beat the market at the millisecond level. Perhaps fiber optic cables should be used to bring Internet access to disadvantaged places around the globe and further the spread of information.
The real question may not be practical, but rather philosophical. The American stock market is the most conspicuous symbol of the free market in the world. If such an institution can be so systemically corrupt, what does that say about the rest of our capitalist system?