Symmetric and asymmetric GARCH estimations of the impact of macroeconomic uncertainties on stock market dynamics in Tanzania

Abstract This study examines the impact of macroeconomic uncertainties on stock market dynamics in Tanzania, motivated by the need to understand volatility patterns in a frontier market characterized by limited liquidity and high sensitivity to economic shocks. We employ symmetric and asymmetric generalized autoregressive conditional heteroskedasticity (GARCH) models, including the nonlinear asymmetric GARCH (NAGARCH) model, to capture complex volatility dynamics. The selection of macroeconomic variables is theoretically motivated: the inflation rate represents domestic monetary policy uncertainty; GDP captures economic performance and growth expectations; and gold prices serve as a proxy for global commodity conditions and safe-haven demand, which is particularly relevant given Tanzania’s mining sector. Using monthly data on the Tanzania share index, inflation rate, GDP, and gold prices from February 2007 to December 2023, we examine how macroeconomic variations influence stock market volatility. While symmetric GARCH models assume equal impacts from positive and negative shocks, asymmetric models including EGARCH, GJR-GARCH, and NAGARCH, capture differential responses. The NAGARCH model allows for nonlinearity in the relationship between past returns and volatility, offering a nuanced understanding of market reactions. Empirical findings suggest that the asymmetric NAGARCH model provides a more accurate representation of Tanzanian stock market behavior, capturing heightened sensitivity to negative shocks typical in emerging markets. The results reveal that inflation and gold prices have significant, long-term influences on stock market volatility. This study contributes to the emerging market finance literature and offers valuable insights for investors and policymakers in understanding risk patterns and designing appropriate strategies for frontier markets.