Estimation of Stochastic Volatility Models with Heavy Tails and Serial Dependence

Joshua Chan and Cody Hsiao (2014)
In: I. Jeliazkov and X.-S. Yang (Eds.), Bayesian Inference in the Social Sciences, 159-180, John Wiley & Sons, Hoboken, New Jersey
[ Working Paper | Code ]

This code estimates three stochastic volatility models: a standard SV model, a SV model with MA(1) Gaussian errors and a SV model with MA(1) Student's t errors.

Instead of the conventional Kalman filter, this code implements the more efficient precision sampler of Chan and Jeliazkov (2009).

These estimation methods are illustrated using empirical examples involving precious metal and foreign exchange returns.