Abstract

The high-frequency stochastic volatility model is fit to intraday returns assuming that the intraday volatility consists of the autoregressive process, the seasonal component of the intraday volatility pattern, and the announcement component responding to macroeconomic announcements. The high-frequency realized stochastic volatility model augments the high-frequency stochastic volatility model with the daily realized volatility by taking account of the bias in the daily realized volatility caused by microstructure noise and non-trading hours. This article extends the high-frequency realized stochastic volatility model such that the return distribution follows the generalized hyperbolic (GH) skew Student's t-distribution. A Bayesian method using an efficient Markov chain Monte Carlo is developed for the analysis of the proposed model. The application to tail risk management such as Value-at-Risk (VaR) and expected shortfall (ES) is provided using the 5-minute returns of S&P 500 E-mini futures. This is a joint work with Toshiaki Watanabe (Hitotsubashi University).