Comentario sobre “Bayesian Analysis of Stochastic Volatility models”

Abstract

There has been an increasing interest in stochastic volatility (SV) models in the last two or three years. Such models are appealing because they follow very naturally from much of finance theory and their properties are relatively easy to derive. Nevertheless, most econometric work has been carried out within the autoregressive conditional heteroscedasticity (ARCH) framework. By assuming that conditional variance is an exact function of past observations, ARCH models are formulated in such a way that the likelihood function may be obtained directly. SV models do not have this property, and this article by Jacquier, Polson, and Rossi (JPR) is an important contribution in that it provides a relatively efficient method of estimation.