Abstract

In high-frequency trading, accurately modeling multivariate volatility is essential for effective risk management and portfolio optimization. This paper tackles the challenges of intraday data, which is both irregular and high-frequency, by introducing two hierarchical models: the irregular basic multivariate stochastic volatility autoregressive conditional duration (IR-BMSV-ACD) model and the irregular dynamic multivariate stochastic volatility autoregressive conditional duration (IR-DMSV-ACD) model. These models incorporate the autoregressive conditional duration (ACD) model to account for irregular gaps between transactions, improving the understanding of market microstructure and future dynamics. To optimize model complexity and accuracy, we use lasso regularization for variable selection, effectively reducing non-significant parameters to zero. The analysis is conducted within a Bayesian framework using the Hamiltonian Monte Carlo (HMC) algorithm with No-U-turn sampler (NUTS) in R via the cmdstanr package. We demonstrate the efficacy of our methodology through simulation studies and real-data analysis of intra-day prices of health stocks traded on the New York Stock Exchange (NYSE) at the microsecond level. Utilizing the refresh time sampling technique, we synchronize transactions, compute synchronized log-returns and gaps, and use these for modeling purposes. This is joint work with Sreeram Anantharaman (PhD UConn, PostDoc Brown University) and Sumanta Basu (Cornell University).