Quantile regression and copula-based analysis of risk and dependence in the South African stock market

Source abstract

This study examines the relationship between the South African stock market and key macroeconomic factors, namely exchange rate movements, crude oil prices, and gold prices, using monthly data from 2007 to 2025. To capture heterogeneous market behavior and downside risk, the study employs quantile regression, quantile autoregression, Conditional Autoregressive Value at Risk (CAViaR) models, and copula models. The results reveal that exchange rate movements exert a consistently negative effect on stock returns across the conditional distribution, while return persistence remains weak, suggesting a relatively efficient market. Downside risk is found to be time-varying and is driven primarily by past risk and adverse market shocks rather than contemporaneous macroeconomic variables. Copula analysis indicates strong negative overall dependence between the Johannesburg Stock Exchange (JSE) and South African Rand/United States Dollar (USD/ZAR) but no asymptotic tail dependence, while the JSE–Brent relationship exhibits moderate symmetric tail dependence. The JSE–Gold relationship is weak, with negligible lower-tail dependence. These findings demonstrate that macroeconomic influences and market risk vary across different market states, underscoring the value of quantile based approaches for modeling financial risk and informing portfolio risk management in emerging markets.