In this paper, we present a Bayesian analysis for stochastic volatility models (SV) and a generalized form of this model, with the aim to estimate the volatilities of financial time series. Considering same special cases of the SV models, we use Markov Chain Monte Carlo methods and the software WinBugs to get the posterior summaries of interest for the different forms of SV models. We also introduce some Bayesian discrimination methods to choose the best model to be used to estimate the volatilities and to get predictions of the financial time series. An example of application is introduced with the Brazilian financial series IBOVESPA.
Stochastic volatility models; Bayesian analysis; MCMC methods; financial time series; IBOVESPA
