Template-Type: ReDIF-Paper 1.0 Series: Tinbergen Institute Discussion Papers Creation-Date: 2004-01-27 Number: 04-015/4 Author-Name: Charles S. Bos Author-Email: email@example.com Author-Workplace-Name: Faculty of Economics and Business Administration, Vrije Universiteit Amsterdam Author-Name: Neil Shephard Author-Email: firstname.lastname@example.org Author-Workplace-Name: Nuffield College, University of Oxford Title: Inference for Adaptive Time Series Models: Stochastic Volatility and Conditionally Gaussian State Space form Abstract: This discussion paper led to a publication in 'Econometric Reviews', 2006, 25(2-3), 219-244.
In this paper we replace the Gaussian errors in the standard Gaussian, linear state space model with stochastic volatility processes. This is called a GSSF-SV model. We show that conventional MCMC algorithms for this type of model are ineffective, but that this problem can be removed by reparameterising the model. We illustrate our results on an example from financial economics and one from the nonparametric regression model. We also develop an effective particle filter for this model which is useful to assess the fit of the model. Classification-JEL: C15; C32; C51; F31 Keywords: Markov chain Monte Carlo; particle filter; cubic spline; state space form; stochastic volatility File-Url: http://papers.tinbergen.nl/04015.pdf File-Format: application/pdf File-Size: 1014738 bytes Handle: RePEc:tin:wpaper:20040015