SLCG Economic Consulting's Logo


High-Frequency Trading and Market Volatility

The "flash crash" of May 6, 2010 -- when the Dow Jones Industrial Average dropped by 9% in a few minutes and quickly regained ground -- has naturally drawn wide attention. Although the sharp drop was not directly triggered by high-frequency traders (traders who execute trades based on complex algorithms and rarely hold a position more than a day), they have been blamed for fueling the selling after a mutual fund complex initialized a program to sell a large amount of E-Mini S&P 500 contracts. Not surprisingly, heated debates concerning the effect of high frequency trading on stock market volatility followed.

In September, 2011, X.Frank Zhang -- an associate professor of accounting at Yale -- published a paper arguing that stock market volatility is exacerbated by an increase in high-frequency trading. Using firm-level data from 1985 to 2009, Zhang found that the positive correlation between volatility and high-frequency trading was especially strong for firms with large market capitalization. As to the liquidity added to the market, which some believe is the major contribution of high-frequency trading, Zhang found that once the share of high-frequency trading exceeds the threshold of 50%, a "hot potato" volume effect with traders passing the same positions between themselves, starts to dominate.

Around the same time, Professor Alex Frino -- affiliated with the University of Sydney Business School and CEO of Capital Markets Co-operative Research Centre -- argued that high frequency trading reduces price volatility in equity markets rather than increasing it. His research is based on several equity markets around the world and rather than identifying a positive relationship between an increase in high frequency trading and in increase in price volatility, he found a negative relationship.

So who's right? Perhaps no conclusive answer could be drawn at this stage. In fact, it is very hard to disentangle the overall effect of high-frequency trading on volatility for at least a few reasons. First, there are different types of traders involved in high frequency trading: momentum driven traders, arbitrage-seeking traders, and liquidity-providing traders. Although the arbitrage-seeking traders are more likely to pursue trades which reduce mispricing, the momentum driven traders could add to market correlation and volatility. It will be very difficult to find data which separate different types of trading activities. Second, as the research by economists Josh Lerner and Peter Tufano has shown, it is very hard in general to estimate the effects of financial innovation on society due to the externality generated by such innovation.

The debate may still be ongoing, but regulators should consider monitoring high-frequency trading and reinforce surveillance in asset markets.