Investor Sentiment and Stock Market Volatility in South Africa

Abstract This study examines the relationship between investor sentiment and stock market volatility in South Africa over the period 2007–2024. Using a composite sentiment index, the South African Investor Sentiment Index (SAISI), constructed from six behavioural and market-based proxies, the study finds no contemporaneous relationship between sentiment and volatility. However, SAISI significantly predicts volatility at 3–6 month horizons, suggesting delayed market adjustment to sentiment-driven expectations. GARCH-family models reveal high volatility persistence and asymmetric responses to negative shocks, consistent with loss aversion theories. Out-of-sample forecasts show sentiment-augmented models achieve 16.4% root mean squared error (RMSE) improvements over autoregressive benchmarks. The findings demonstrate that investor sentiment carries forward-looking information complementing traditional volatility models, with critical implications for investors, regulators and policymakers managing emerging market risk.