Efficient Bayesian estimation of a multivariate stochastic volatility model with cross leverage and heavy-tailed errors (Revised version of CARF-F-191(2009) )

Abstract

The efficient Bayesian estimation method using Markov chain Monte Carlo is proposed for a multivariate stochastic volatility model that is a natural extension of the univariate stochastic volatility model with leverage and heavy-tailed errors, where we further incorporate cross leverage effects among stock returns. Our method is based on a multi-move sampler which samples a block of latent volatility vectors and is described first in the literature for a multivariate stochastic volatility model with cross leverage and heavy-tailed errors. Its high sampling efficiency is shown using numerical examples in comparison with a single-move sampler which samples one latent volatility vector at a time given other latent vectors and parameters. The empirical studies are given using five dimensional stock return indices in Tokyo Stock Exchange.