Our paper is among the first to provide an empirical assessment as to whether High Frequency Trading (HFT) has a positive effect on market fairness. We find the presence of HFT has significantly mitigated the frequency and severity of price dislocation and the likelihood of manipulation, counter to recent concerns expressed in the media that HFT exacerbates market manipulation. Taken together with the existing evidence that HFT reduces transactions costs and enhances price discovery, we demonstrate not only how market fairness can be operationalized, but how it can be used in market structure decisions. New research conducted by Professor Douglas Cumming, PhD candidate Feng Zhan, and Professor Mike Aitken, directly examines the link between HFT and end-of-day price dislocation. End-of-day (EOD) price dislocations are one of the most common forms of manipulation in view of the many important functions closing prices have on computing for example: index values, prices for related securities, compensation, and fund net asset values. Consequently, a massive incentive exists for market participants to manipulate security closing prices. HFT has become commonplace in many exchanges around the world. While estimates vary due to the difficulty in ascertaining data, estimates suggest HFT accounts for 50-70% of equity trades in the U.S., 40% in Canada, and 35% in London. Some commentators have expressed concern that HFT might increase the prevalence of market manipulation. By virtue of the speed with which they can enter orders and execute transactions, HFT have the potential to facilitate manipulation more easily in a number of ways. For example, HFT can be used to enter purchase orders at successively higher prices to create the appearance of active interest in a security; this behavior is often referred to as ‘ramping/gouging’. Similarly, ‘market setting’, whereby HFT cross-orders at the short-term high or low to effect the volume weighted average price, or to set the price in one market for the purpose of a making a profit in another market (such as between option markets and the underlying equities market) is another form of manipulation. There are however at least two reasons to believe that HFT will on average curtail market manipulation. First, the introduction of exchange surveillance systems designed to pick up patterns of illegal manipulation activities, not one-off manipulation. Second, HFT has been reported to improve liquidity by reducing bid-ask spreads and facilitate price discovery. It is much more difficult for manipulators. to engage in market manipulation in the presence of greater market efficiency. By examining a monthly panel dataset of the frequency and severity of EOD dislocation cases from 22 stock exchanges around the world from January 2003 to June 2011, this study documents that in the presence of HFT, EOD price dislocation is on average less frequent in terms of the number cases, and less pronounced in terms of the average EOD trading value surrounding suspected cases. Further, this study examines specific dates when EOD price dislocation is most likely to be manipulation, including dates when options expire and end of quarter calendar dates; the data show that in the presence of HFT, EOD price dislocation is less pronounced on these dates as well. Overall, the results in our paper support the view that the price discovery and liquidity function of HFT on average significantly dominates any role that HFT may play in facilitating market manipulation, at least with respect to EOD price dislocation.
Author(s): Douglas Cumming, Feng Zhan, Michael Aitken