Modelling Volatility of Asset Returns in Nigerian Stock Market: Applications of Random Level Shifts Models

Abstract

This study examines the impact of structural breaks on conditional variance and mean reversion in symmetric and asymmetric GARCH models. A multiple breakpoint testing procedure was used to identify structural break points in conditional variance of daily stock returns of 8 commercial banks in Nigerian stock market for the period 17th February, 2003 to 31st September, 2016. Standard GARCH, EGARCH and TGARCH models with and without break points were applied to evaluate variance persistence, mean reversion rates and leverage effects while estimating conditional volatility. Results showed high persistence in conditional volatility for the banking stocks, but when the random level shifts were incorporated into the models, there was reduction in the conditional volatility of these models. The half-lives of volatility shocks also reduce in the presence of these regime shifts. TGARCH was found to be the best fitting model among the standard GARCH and EGARCH models. The study recommends estimation of volatility models to incorporate structural breaks in order to avoid over estimation of shock persistence in the conditional variance.