Information Criterion and Joint Determination of the Numbers of Regimes and Variables in Markov Switching Model: Simulation and Empirical Application

Abstract

This study has three purposes. The first one is to examine the performance of four information criteria in the context of joint determination of the numbers of regimes and variables in Markov switching model (hereafter called the MS model). These criteria are Akaike (1974) information criterion (hereafter called AIC), Schwarz (1978) information criterion (hereafter called SIC), HQC (Hannan & Quinn 1979), and Markov switching criterion: MSC (Smith, Naik & Tsai 2006). The second purpose is to investigate further whether the numbers of regimes and variables in aggregate time series are similar to those in individual time series. Third, to verify the simulation results from the second objective, this study applies the MS model to both aggregate and individual time series in reality.