Estimation of Stochastic Volatility Models with Markov Chain Monte Carlo Methods

Download or read book Estimation of Stochastic Volatility Models with Markov Chain Monte Carlo Methods written by Maximilian Richter and published by . This book was released on 2016 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Markov Chain Monte Carlo (MCMC) methods are a Bayesian approach to tackle one of the main obstacles encountered in the estimation of modern-day stochastic volatility models: the curse of dimensionality induced by the increasing number of latent variables. This thesis strives to study the performance of affine jump-diffusion models in comparison to state-of-the-art Lévy-based return dynamics. Thus MCMC methods are applied to a novel dataset of S & P500 returns that comprises different periods of economic turmoil, such as the subprime crisis. The subordinate research goal is to address difficulties in the implementation of the MCMC methodology. In line with previous studies, the results indicate that jump components are indeed crucial for capturing complex patterns like skewness and excess kurtosis of the return distributions. Moreover, infinite-activity Lévy jumps prove to be superior to discrete compound Poisson jumps.