Social media and financial markets: The impact of Twitter sentiment on the Johannesburg Stock Exchange

This study examines the effect of Twitter-derived investor sentiment on stock market volatility in South Africa using daily data for the JSE All Share Index from 2016 to 2023. Using GARCH-M, E-GARCH-M, and GJR-GARCH-M, the results show that the GJR-GARCH-M specification provides the best fit both before and after incorporating sentiment. Twitter sentiment significantly amplifies market volatility, with negative sentiment exerting a more substantial impact than positive sentiment, consistent with asymmetric volatility dynamics and the leverage effect. Overall, the findings demonstrate that Twitter-derived sentiment contains valuable information for modelling and understanding volatility in emerging equity markets such as South Africa.