Abstract

Regime-switching GARCH (generalized autoregressive conditional heteroscedastic) model combines the idea of Markov switching and GARCH model, which can also be considered as an extension of Hidden Markov model. The statistical inference for this model is, however, rather difficult because of the dependencies to the whole regime history. In this paper, we obtain the consistency of quasi-maximum likelihood estimator for a reduced regime-switching GARCH model. Simulation studies to illustrate consistency, asymptotic normality of the proposed estimators, as well as a model misspecification problem are also presented

Keywords

Asymptotic Normality; Consistency; GARCH model; Hidden Markov Model; Regime-switching; Quasi-MLE

Published in

Research report (Centre of Biostochastics)
2005, volume: 2005, number: 2, pages: 1-15
Publisher: SLU

Authors' information