A Game of Milliseconds

The landscape of Wall Street has transformed greatly over the past twenty years. Quickly disappearing are the days of alpha males with Don Draper haircuts dressed in well-fitted suits, yelling bid-ask prices; they have been all but supplanted by nerdy computer programmers and math geeks, with degrees from places like MIT and Stanford.  The advent of modern technology has helped give rise to high-frequency trading (HFT). HFT utilizes powerful computers and complex algorithmic trading strategies to rapidly transact large quantities of securities. These complex algorithms allow traders to anticipate emerging market trends in a fraction of a second and then exploit those trends for profit. Although margins may be mere pennies, the sheer volume of executed trades can make HFT highly lucrative, a fact which is reflected in the prevalence of this practice. HFT plays a significant role in the market, accounting for more than 50% of all equity trades in recent years. Gross profits from HFT are estimated to be between 1.25 and 5 billion dollars annually.

Although margins may be mere pennies, the sheer volume of executed trades can make HFT highly lucrative.

A broad range of trading strategies fall underneath the umbrella of HFT, most of which would be impossible without modern day computers and sophisticated software. The following example will illustrate the general principle behind HFT. Suppose a trader places an order to buy 50 shares of a stock and the current cheapest price is $100.00. HFT firms have made deals with stock exchanges to get information about placed orders. Given this information, an HFT firm knows your interest in buying a particular stock.  It can (in .001 seconds) then buy all shares of the stock at a price of $100.00  before you can and then immediately place a sales order for $100.01, the price you will be forced to pay. The result is nearly risk free profit.

Other HFT strategies are more complicated. Statistical arbitrage involves exploiting deviations from steady statistical relationships between securities. The idea is that if two historically correlated stocks have a relationship that temporarily weakens, a duo long/short strategy can be implemented to “guarantee profit,” assuming the spread between the two stocks will converge. For example, consider two stocks, Pepsi and Coca Cola. These two companies produce similar products and have correlated stock behavior, historically. If the price of Pepsi rises, while the price of Coca Cola remains stable, traders can buy Coca Cola and sell Pepsi. If the price of Coca Cola increases to close the price gap, the traders make money on the Coca Cola stock.  If the price of Pepsi falls, than the trader makes money off shorting the Pepsi stock. Latency arbitrage involves exploiting miniscule differences in stock prices between various exchanges. These opportunities normally exist for mere fractions of a second.  Information or news-based arbitrage relies on collecting just released financial data (unemployment data, bond rates). In milliseconds, computer software can look for certain keywords, adjust trading strategies accordingly, and execute trades with minimal human input.

Although HFT relies on clever algorithms and arbitrage strategies, much of the arms race between firms has focused on speed. Speed gives HFT one of its greatest advantages: the ability to hop to the front of the order queue for a given security. For example, if an HFT firm is unable to sell a stock at a desired price, they can simply sell to someone behind them in the order queue created for that stock, for little to no loss of profit. The speed of HFT depends on cutting down on data transfer times between exchanges and firms. Data is usually transferred using fiber optical cables or, as in more recent years, through microwave technology, which can communicate data in half the time of fiber optic cables. Firms even fight for the right of co-location, or placing their servers in the actual stock exchange to cut down on data transfer times.

Although HFT relies on clever algorithms and arbitrage strategies, much of the arms race between firms has focused on speed.

Perhaps in response to its phenomenal success, HFT has drawn many critics. Some critics argue that HFT increases market volatility and point to the Flash Crash of 2010 as evidence. During the Flash Crash, stock indexes such as the NASDAQ, S&P 500, and Dow Jones collapsed but rebounded very quickly, all within a 36 minute time frame. The Dow Jones dipped nearly 9% over this period of time. The SEC then investigated the Flash Crash and concluded that “aggressive selling” of certain futures by HFT firms helped contribute to the market plunge. Other critics see the strategies used by HFT as predatory, taking advantage of tiny gaps in the market at the expense of the other investors. Michael Lewis, author of The Big Short, makes this very argument in his book Flash Boys.  On the other hand, some believe HFT provides market liquidity and makes transactions much quicker and smoother for all investors. HFT may actually even lower trading costs for the average investor. In addition, some argue that HFT firms are competing against each other, not against long-term value investors. Although new legislation has aimed at regulating HFT more tightly, everything seems to suggest that HFT is here to stay. The world of finance will be inextricably linked to cutting edge technology, with power in the hands of those who understand it; technology has never been so important in the financial world.

About The Author

Michael Li