Computationally Efficient Estimation of Factor Multivariate Stochastic Volatility Models - Statistics > ComputationReportar como inadecuado




Computationally Efficient Estimation of Factor Multivariate Stochastic Volatility Models - Statistics > Computation - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Abstract: An MCMC simulation method based on a two stage delayed rejectionMetropolis-Hastings algorithm is proposed to estimate a factor multivariatestochastic volatility model. The first stage uses kstep iteration towards themode, with k small, and the second stage uses an adaptive random walk proposaldensity. The marginal likelihood approach of Chib 1995 is used to choose thenumber of factors, with the posterior density ordinates approximated byGaussian copula. Simulation and real data applications suggest that theproposed simulation method is computationally much more efficient than theapproach of Chib. Nardari and Shephard 2006}. This increase in computationalefficiency is particularly important in calculating marginal likelihoodsbecause it is necessary to carry out the simulation a number of times toestimate the posterior ordinates for a given marginal likelihood. In additionto the MCMC method, the paper also proposes a fast approximate EM method toestimate the factor multivariate stochastic volatility model. The estimatesfrom the approximate EM method are of interest in their own right, but areespecially useful as initial inputs to MCMC methods, making them more efficientcomputationally. The methodology is illustrated using simulated and realexamples.



Autor: Weijun Xu, Li Yang, Robert Kohn

Fuente: https://arxiv.org/







Documentos relacionados