Predictability of GARCH-Type Models in Estimating Stock Returns Volatility. Evidence from Kenya

Purpose: The aim of this paper was to  evaluate which of the seven GARCH-type models, namely sGARCH, IGARCH, EGARCH, TGARCH, GJRGARCH, APARCH, and CGARCH, was suitable for predicting the Nairobi Securities Exchange-listed firms' volatility.   Theoritical framework: The Efficient Market Hypothesis is crucial in predicting market value of stocks. Therefore, this study employed the efficient market hypothesis to the the predictability of the stocks returns volatility.   Design/Methodology/Approach: In  this study, we used census approach to collect data from 49 Nairobi Securities Exchange listed firms. The data was collected from 1st January 2011 to 31st December 2020. TO evaluate the volatility, we used the GARCH-type models.   Findings: The study found that the APARCH model as the best suitable for forecasting the volatility of Nairobi Securities Exchange-listed firms.   Research, Practical & Social implications: We propose the the APARCH model as the best suitable model for predicting volatility of stock returns. The findings can be used by investors in making judicious financial decisions. For acedmic purpose, the findings are essential in supporting new knowledge of which model is best fit in predicting the NSE stocks returns volatility.   Original/ Value: The study contributes to the literature on the best suitable model in predicting the volatility of the stocks returns.