High Frequency Trading: What we have learned and where we’re headed
Think back to 2009 when high frequency trading (HFT) first became a topic of interest. At the time very little research existed on the issue -- essentially all we knew was that high frequency traders were fast and technologically sophisticated. They spent heavily to achieve low-latency, by purchasing co-located servers and by employing programmers who could optimize data analysis techniques. Since then researchers, including myself, have started studying HFT in greater depth.[i]
My work specifically looks at the activities and impact of HFT on U.S. equity markets, and my research has uncovered several key characteristics of high frequency traders. The following points summarize my work to date:
High frequency traders (1) tend to follow a price reversal strategy driven by order imbalances, (2) earn gross trading profits of approximately $2.8 billion annually, (3) do not seem to systematically engage in a non-high frequency trader anticipatory trading strategy, (4) have trading strategies that are more correlated with each other than do non-high frequency traders, (5) change their fraction of trading involvement only moderately as volatility increases, (6) add substantially to the price discovery process, (7) provide the best bid and offer quotes for a significant portion of the trading day and do so strategically so as to avoid informed traders, but provide only one-fourth as much book depth as non-high frequency traders, and (8) may dampen intraday volatility.[ii]
Overall, along the measures that I look at, high frequency traders are rather dull. In fact, staying around in volatile times, aiding price discovery, competing for liquidity, and potentially dampening volatility, may be positive benefits that HFT firms are bringing to the equity markets. My previous work has explored these results and their limitations in a paper entitled “High Frequency Trading and Its Impact on Market Quality,” but next I want to examine areas of study for future research.
From my perspective, future research should focus on at least two concerns regarding HFT. The main concern I raise in my current work is that HFT firms generally do not provide much liquidity depth. That is, they do not make many shares available through limit orders near the best bid and offer. As noted above, high frequency traders provide about one-fourth as much depth on the order book of NASDAQ as do non-high frequency traders. If these traders are taking the role of traditional market makers in being major providers of liquidity, then it would be preferable to see them adding depth beyond a couple hundred shares at the best bid and offer. On the other hand, from the perspective of a market maker, when one provides more shares at competitive prices, he has a greater probability of finding himself being a significant liquidity provider to an informed trader as, at least historically, intermediate and large trades were associated with such traders. A behavior that I have not analyzed in my study but that may invalidate this lack-of-depth-contribution concern is that high frequency traders might replace their limit orders immediately after one of their previous orders has been hit.
If high frequency traders do not provide much liquidity depth this is indeed an issue, but there are several ways to mitigate its significance. I’ll give two examples of incentive-based techniques that could be implemented. First, instead of the current rule of the limit order execution queue being based on price and time, it could also take into consideration the size of the limit order posted so that larger liquidity providing orders may gain priority over smaller ones. Second, exchanges could adjust liquidity rebates based on how many shares a liquidity provider was willing to make available for a trade. This is an extension of the Joint Commodity Futures Trading Commission - Securities and Exchange Commission (SEC) Advisory Committee’s suggestion to alter liquidity rebates based on market conditions.[iii]
Additional future research should also address whether HFT firms have been engaging in manipulative or otherwise illegal trading activities, as some have purported. In my work I examine whether, in aggregate, high frequency traders are engaging in an anticipatory trading strategy. I do not find evidence consistent with them doing so, but the data aggregation does make it more difficult to statistically detect such behavior. There are other activities of concern as well. For instance, Trillium Trading was fined by the Financial Industry Regulatory Authority (FINRA) for engaging in a price manipulation scheme called “layering,” which involves the use of hidden and displayed limit orders to improve market execution. Are others doing the same? That has yet to be seen but is worth exploring.
In addition, HFT firms have been accused of quote stuffing. Quote stuffing refers to the placement and almost-immediate cancelation of large volumes of limit orders. The claim is that a firm might flood the market with limit orders and cancellations -- not with the intention to trade but with the aim of slowing down other HFT firms’ information processing systems or creating delays in an exchange’s data feeds. I am unaware of any formal cases regarding this behavior, but if it is occurring, the SEC and FINRA may be interested.
While more is known about HFT today than in the past, it is still a clandestine industry. With the amount of media and regulatory attention HFT is receiving it is likely we will continue to hear and care about it in to the future. Hopefully we will quickly gain an even better understanding of how it impacts markets, and what regulatory and market changes are necessary to minimize any associated negative side-effects.
[i] Other work includes “Low-Latency Trading” by Joel Hasbrouck and Gideon Saar, “The Flash Crash: The Impact of High Frequency Trading on an Electronic Market” by Andrei Kirilenko, Albert Kyle, Mehrdad Samadi, and Tugkan Tuzun, and “High Frequency Trading and The New-Market Makers” by Albert Menkveld, among others.
[ii] Brogaard, Jonathan. High Frequency Trading and Its Impact on Market Quality. November 22, 2010. SSRN.
[iii] Joint Commodity Futures Trading Commission - Securities and Exchange Commission Advisory Committee on Emerging Regulatory Issues. February 18, 2011. Recommendations Regarding Regulatory Responses to the Market Events of May 6, 2010.
About Jonathan Brogaard
Jonathan is a Finance PhD candidate at the Kellogg School of Management, Northwestern University and concurrently is completing his JD at the Northwestern University School of Law. Prior to joining Kellogg and Northwestern Law, he was an undergraduate student at Occidental College. Jonathan's research interests include high frequency trading, law and finance, investments, and the taxation of financial products. In the fall he will be joining the faculty of the Foster School of Business at the University of Washington.