Modelling returns on stock indices for western and central european stock exchanges-a Markov switching approach

Abstract

In this paper a Markov switching mixture of normal distributions is applied to the monthly returns on the main stock indices for emerging financial markets in central Europe (BUX, PX50 and WIG). Additionally the results are compared to those obtained for western Europe (DAX, CAC40 and FTSE100). The results of model comparison suggest that the Markov switching mixture of normal distributions has substantially more descriptive validity than a single normal distribution. Nevertheless, there is no clear indication that, in modelling monthly returns, the Markov switching mixture of three normal distributions is superior to the mixture of two. Finally, the ability of the models to describe returns during international financial crises is evaluated.