The Impact of High-Frequency Trading on the U.S. Equity Market

This dissertation studies the impact of high-frequency trading (HFT) on the U.S equity market. I investigate the trading behavior of high-frequency traders (HFTs) using a massive dataset that contains the NASDAQ ITCH feed messages of all S&P 500 component stocks in year 2012. I identify clusters of extremely high cancellation activity (Blocher et al., 2016) in the order book and use high cancel clusters as a proxy for high-frequency cancellation activity. I examine the change in liquidity measures from one-half second before each cancel cluster starts until after the cancel cluster closes and find that with the presence of high cancel activity, liquidity measures recover to their pre-cluster level faster than in non-cancel clusters. Furthermore, an analysis of 1-minute time intervals finds various HFT proxies to be positively related to liquidity, especially for large-cap stocks and certain sectors. Using the Li criterion (Li et al. 2018), I differentiate trades placed by HFTs versus low-frequency traders (LFTs) and compares the two types of trades under the VAR/VMA framework (Hasbrouck 1991). Evidence shows that HFT trades contribute more to the price discovery process than LFT trades and HFTs impose adverse selection costs on LFTs. This study disambiguates unreliable liquidity and faster price discovery