Volatility clustering and persistence in Kenyan financial markets: a GARCH approach

Abstract

This study aimed to understand volatility in emerging countries, particularly Kenya, by investigating volatility clustering and persistence in the Nairobi Securities Exchange (NSE) utilizing Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models. The research emphasized the significance of understanding market volatility patterns, considering their implications for market efficiency, risk management, and investment strategies. Using weekly time series data from the NSE 25 share index spanning January 2016 to December 2023, the study evaluated the effectiveness of various GARCH model specifications, specifically asymmetric EGARCH and TGARCH models, in capturing volatility trends. The findings revealed significant volatility clustering and persistence in the NSE, characterized by periods of high volatility following similar patterns. Notably, the TGARCH model outperformed the EGARCH model in terms of model fit and forecasting accuracy, as evidenced by lower Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) values, as well as superior Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) metrics. These results provide insights into the unique volatility dynamics of the NSE and offer actionable recommendations for policymakers and investors to enhance market stability and improve investment decisions. Overall, this research fills existing knowledge gaps related to volatility in the Kenyan financial sector and contributes to the broader literature on emerging markets.