A TIME SERIES VOLATILITY MODEL FOR ALL SHARE INDEX RETURNS USING MARKOV SWITCHING AUTOREGRESSION

This study explored the Autoregressive models incorporating Regime switching or Markov switching non-linear predictive models. This stemmed from the complexities observed in economic phenomena the understanding of which will help in recommending better model fits. The study collected data on all share index returns (Jan, 1985 - December 2019) from the Nigeria Stock Exchange, fit an appropriate MS-AR model and estimate its parameters. The parameters of the model were obtained while their properties like the expected duration, autocorrelation measure and the goodness of fit were equally computed in testing the applicability of the model. The result shows that the MS (3)-AR (3) as a predictive model was appropriate, efficient and robust enough for forecasting the returns of the all-share index of the Nigerian stock exchange over the sampled period. The study is therefore relevant for modelling all share index returns by investors, policy makers, researchers and the general public.