Works in Progress

The Cancel-to-Trade Ratio and Market Quality
(with Robert Connolly)
Abstract:   A salient feature of high frequency trading (HFT) is the submission and almost immediate cancelation of numerous orders.   Using the SEC’s new MIDAS data system, we characterize the daily cancel-to-trade ratio, a common proxy for HFT.   We find that the unconditional distribution of the daily cancel-to-trade ratio exhibits extreme positive skewness, driven primarily by small and infrequently traded stocks.   Large and frequently traded stocks, however, are not immune to extreme cancelation.   We extend our analysis of the cancel-to-trade ratio by using it as a proxy for HFT in testing several theories linking HFT to various measures of market quality.   The results suggest that, on average, more HFT (a high cancel-to-trade ratio) is associated with deteriorating market quality.   We also provide insights into differences in cancelation activity between common stocks and ETFs.