P. Galeano San Miguel, C. Garcia de la Fuente, M. P. Wiper
Financial data sets often present skewness and high kurtosis. As a consequence, it is natural to look for a model that is flexible enough to capture these characteristics. The proposal is to perform Bayesian inference and prediction for a generalized autoregressive conditional heteroskedastic (GARCH) model, where the innovations are assumed to follow a Skew Slash distribution. Gibbs sampling is used for parameter estimation and volatility prediction, and the method is illustrated using real financial data.
Palabras clave: financial time series, Bayesian analysis, Skew Slash distribution, GARCH model, Skewness, Kurtosis
Programado
XA3 Métodos bayesianos 3
18 de abril de 2012 09:00
Sala Londres
