Modelling of PX Stock Returns during Calm and Crisis Periods: A Markov Switching Approach

This paper deals with the modelling of the Czech stock market characterized by the weekly PX stock returns based on the Markov switching (MSW) approach. The analysed period spanning from April 8, 2007 to February 7, 2021 comprises both the "normal" calm and "turbulent" crisis periods. The two-regime MSW model thus enables to capture and specify the periods of bull and bear markets characterized by positive mean return-low volatility and negative mean return-high volatility periods, respectively. The analysis was enriched by consideration of the price/return - trading volume relationship, but it led neither to recognizable changes in values of estimated parameters nor to substantial differences in transition matrices. The presented results clearly confirmed the existence of several turbulent periods reflecting the worldwide financial and economic situation indicating the occurrence of crisis period with the probability of almost 0.14. However, the "normal" calm behaviour of returns occurred much more often and was proved to be more persistent in comparison to "turbulent" crisis period.