High-Frequency Liquidity in the Chinese Stock Market : Measurements, Patterns, and Determinants

We explore a broad range of high-frequency liquidity measures for the Chinese stock market, based on a comprehensive tick-level dataset for stocks on the Shenzhen Stock Exchange (SZSE) with approximately 16.7 billion events in 2020. We integrate the raw event-level data into a granular and continuous limit order book for each stock for the entire year. We summarize their liquidity levels and key distributional properties. Hypothesis tests show that order interarrival times follow Weibull---not exponential---distributions, implying that Poisson flow is not an appropriate model for order flow in the Chinese stock market. We analyze the intraday and cross-sectional patterns of liquidity, and find novel intraday periodicities in liquidity at whole-minute frequencies such as 1-minute, 5-minute, and 10-minute. Finally, we propose the aggressive-passive imbalance (API), analogous to the order flow imbalance of Cont, Kukanov, and Stoikov (2014), and develop an order-based model of the change in bid-ask spread that sheds light on the universal mechanism of spread formation with respect to order flows. To the best of our knowledge, this is by far the most comprehensive study of market liquidity for the Chinese stock market in the literature