Abstract: We study the inter-trade durations in the Nasdaq limit order market and find that inter-trade durations in high frequency have two modes, one is around 10^−4 second and the other is of order 10^0 second. To interpret this phenomenon and find how the modes depend on information in the limit order book (LOB), we propose a two-state multi-factor regime switching model for inter-trade durations, in which the time-varying probability transition matrices of inter-trade durations depend on some LOB variables. We study the properties of the model, such as ergodicity, over-dispersion, and exogenity, and use the Expectation- Maximization algorithm to make inference on model parameters. Simulation studies show that our model does generate the bimodal distributions for inter-trade durations. We then use the model to analyze the Nasdaq stock data, and show that both the in-sample and the out-of-sample performances of our model are better than those of some benchmark duration models. Our empirical study also shows that two modes of inter-trade durations are dependent on some LOB factors. The talk is based on a joint work with Zhicheng Li and Haipeng Xing.

Bio: Dr. Xinyun Chen is currently an Assistant Professor in the Institute for Data and Decision Analytics, The Chinese University of Hong Kong, Shenzhen. She received her Ph.D in Operations Research from Columbia University in 2014. Her research interests include applied probability, Monte Carlo method and their applications in financial markets. She has published papers in journals including Annals of Applied Probability, Mathematics of Operations Research and Accounting and Finance. Before joining iDDA, she was an Assistant Professor at Stony Brook University and Wuhan University.