Multivariate Generalizations of the Markov-Switching Model

We present a multivariate generalization of the simple markov-switching model. We allow for the introduction of several latent processes that have a simple parametric distribution. The matrix-variate bernoulli distribution yields a flexible yet parsimonious pattern of dependence between the different latent processes while preserving the markovian property. We also show how to estimate the model in the bayesian framework and give several examples

MoreLess

Year of publication:	2006-07-04
Authors:	Khaled, Mohamad
Institutions:	Society for Computational Economics - SCE
Subject:	Bayesian statistics \| markov-switching \| matrix-variate bernoulli distribution \| multivariate generalizations

More details

Series:	Computing in Economics and Finance 2006.
Type of publication:	Book / Working Paper
Notes:	The text is part of a series Computing in Economics and Finance 2006 Number 297
Classification:	C11 - Bayesian Analysis ; C32 - Time-Series Models ; C51 - Model Construction and Estimation
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005342985