This is Part 2 of our debate. Read Part 1.
High-frequency trading (HFT) provides four main benefits to markets. It improves the efficiency of the market by providing liquidity and lowering trading costs. It improves price discovery (the process of finding where prices equate demand and supply). It reduces systematic risk in markets and reduces the likelihood of flash crashes. And it reduces market manipulation, thereby improving market integrity. These benefits are based on empirical studies involving massive data sets carried out by academics without any preconceived bias or stake in their outcomes.
All available data indicate market efficiency, in terms of trading costs (such as bid-ask spreads), has gone down with the presence of HFT, while trading volume has gone up. The effect has been large. For instance, a 2011 study by T. Hendershott, C. Jones and A. Menkveld shows a narrowing of bid-ask spreads by roughly 50%. This means it’s cheaper for all traders, whether human or machine, to trade on exchanges with high-frequency traders.
Read: Read: IIROC team to study HFT impact
Empirical evidence has also shown HFT reduces market fragmentation and enhances price discovery. For example, a 2010 study by J. Brogaard shows HFT quotes add substantially to price discovery and contribute more to price discovery than non-HFT quotes. A further study by Brogaard, Hendershott and R. Riordan in 2013 finds high-frequency traders trade in the direction of permanent price changes and against transitory price changes; incorporate public information more quickly; and incorporate limit order imbalances in trades. These effects are beneficial to both institutional and retail traders, reducing trading costs, reducing temporary price pressure and limiting intraday volatility.
Consistent with evidence of improved market efficiency, evidence shows high-frequency traders reduce systematic risk in markets. For example, a number of studies show high-frequency traders reduce the likelihood of flash crashes and do not increase or contribute to them. In a 2013 study, J. MacIntosh explains this was even the case with the May 2010 Flash Crash, for example, which is often misattributed to high-frequency traders. Further, S. Groth’s 2011 study finds high-frequency traders do not withdraw from markets during periods of high volatility, and this finding is consistent with studies from Brogaard (2010) and Brogaard et al. (2013). Also, A. Golub et al. in 2012 show 68% of mini flash crashes were caused by intermarket sweep orders—not high-frequency traders.
Read: ETFs and HFT: Need for speed
Market quality, in terms of the absence of manipulation, has gone up with the presence of HFT. For example, a recent study by M. Aitken et al. in 2013 shows end-of-day prices were substantially less distorted in stock markets around the world after HFT entered those exchanges. Aitken et al. examined closing price manipulation in 22 stock exchanges around the world from January 2003 to June 2011. They found high-frequency traders reduced the frequency and severity of end-of-day dislocation or mispricing. The data indicate the probability of end-of-day dislocation declined by at least 20%, and trading value around dislocation events went down by at least 40%. On days where the end-of-day price distortion was more likely to be attributable to manipulation— for example, on days when options expire, or at the end of quarters and the end of the year—the impact of HFT was even stronger: the probability of end-of-day dislocation went down by at least 70%. Interestingly, the evidence from Aitken et al. considers many other factors curtailing end-of-day manipulation, such as better trading rules occurring on some of the exchanges in the period considered, as well as surveillance to detect such manipulation. The data indicate HFT was more important than trading rules or surveillance in mitigating end-of-day distortions.
This finding is important, since exchanges and regulators spend significant resources on rule design and enforcement. Enforcement involves computer algorithms to detect unusual trading patterns, as well as a human element. In Canada, Investment Industry Regulatory Organization of Canada (IIROC) staffers carry out surveillance. Enforcement is a vital aspect of ensuring market quality, but evidence from Aitken et al. shows HFT reduces cases—and, therefore, the need to carry out enforcement—because end-of-day manipulation is less common with HFT than without it.
Read: Institutional traders concerned by HFT
Rogue traders or market manipulators can use HFT to engage in forms of manipulation other than end-of-day, such as front-running (trading in advance of a client’s orders). Data from large-scale systematic studies of such manipulation, however, indicate no such evidence of front-running from high-frequency traders.
Michael Lewis’s recent book, Flash Boys, discusses some examples of front-running by high-frequency traders, but the evidence is not systematic.
Non-high-frequency traders can engage in front-running just like high-frequency traders can—there is no systematic pronounced evidence of front-running unique to high-frequency traders.
Overall, there is much evidence showing high-frequency traders provide significant benefits to exchanges and their investors in terms of enhanced market efficiency and integrity. It’s easy to focus on the negative elements of HFT, but it’s important not to forget the advantages new technologies such as HFT can provide. One just needs to be able to curtail the human element of rogue traders programming HFT algorithms in ways to manipulate markets and engage in insider trading.
Regulators must keep pace with new technology and trading patterns in the markets. With the high-quality staff and technology at the Ontario Securities Commission, IIROC and TSX, Canadian markets will continue to lead the world in ensuring an efficient and fair venue for investors.
Douglas Cumming is a professor and Ontario research chair with the Schulich School of Business, York University.
Get a PDF of this article.