Exploring Macroeconomic Dynamics on Stock Returns in African Countries: Garch-Midas and Regularization Regression Approach

Abstract

Measuring stock market volatility and its determinants is critical for stock market participants as volatility spillover effects alters corporate performance. This thesis adopted two different approaches where the analysis was distinctly implemented using GARCH-MIDAS and regularization regression methods. The classic GARCH as a benchmark and the univariate GARCH-MIDAS framework are the first two GARCH family models whose forecasting outcomes are examined in this thesis. The second analysis was a shrinkage approach that adopted two techniques: LASSO and ridge regularization approaches, which were used to determine the most influential regressors for stock index returns and volatility in the three markets. The outcome of GARCH MIDASanalyses suggests that inflation, interest rate, currency exchange rate, and price of oil are significant determinants of the volatility of the Johannesburg Stock Market All Share Index. While for Nigeria the volatility reacts significantly to exchange rate and price of oil. Furthermore, Ghanaian equity volatility is significantly influenced by inflation, the exchange rate, the interest rate, and the price of oil, especially for the long term volatility component. The significant shock of the oil price and exchange rate to volatility are present in all three markets using the GARCH-MIDAS framework. As an alternative, the machine learning algorithms selected the money supply, oil price, interest rate, and exchange rate as the most critical indicators for predicting South African stock returns. In the Nigeria scenario, the regularization algorithm produced conclusion where oil price and money supply were specified as the utmost relevant variables in predicting asset returns. Additionally, currency exchange rate, interest rate, and price of crude oil were the ultimate significant indicators that determined stock returns in Ghana.