Discrete-Time Stochastic Volatility Models and MCMC-Based Statistical Inference

In this paper, we review the most common specifications of discrete-time stochasticvolatility (SV) models and illustrate the major principles of corresponding MarkovChain Monte Carlo (MCMC) based statistical inference. We provide a hands-on approachwhich is easily implemented in empirical applications and financial practice andcan be straightforwardly extended in various directions. We illustrate empirical resultsbased on different SV specifications using returns on stock indices and foreign exchangerates.