Commodity price volatility, stock market performance and economic growth: evidence from BRICS countries

Abstract

The study investigated the nexus between commodity price volatility, stock market performance, and economic growth in the emerging economies of Brazil, Russia, India, China, and South Africa (the BRICS) predicated on two hypotheses. First, the study hypothesised that in modern integrated financial systems, commodity price volatility predisposes stock market performance to be non-linearly related to economic growth. The second hypothesis was that financial crises are an inescapable feature of modern financial systems. The study used daily data on stock indices and selected commodity prices as well as monthly data on national output proxies and stock indices. The study analysed data for non-linearities, fractality, and entropy behaviour using the spectral causality approach, univariate GARCH, EGARCH, FIGARCH, DCC-GARCH, and Markov Regime Switching (MRS) – GARCH. The four main findings were: first, spectral causality tests signalled dynamic non-linearities in the relationship between the three commodity futures prices and the BRICS stock indices. Second, the predominantly non-linear relationship between commodity prices and stock prices was reflected in the nexus between the national output proxies and the indices of the five main commodity classes. Third, spectral causality analysis revealed that the causal structures between commodity prices and national output proxies were non-linear and dynamic. Fourth, the Nyblom parameter stability tests revealed evidence of structural breaks in the data that was analysed. The DCC-GARCH model uncovered strong evidence of contagion, spillovers, and interdependence. The study added to the body of knowledge in three ways. First, micro and macro levels of commodity price changes were linked with corresponding stock market performance indicator changes. Second, unlike earlier studies on the commodity price – stock market performance – economic growth nexus, the study employed spectral causality analysis, single - regime GARCH analysis, Dynamic Conditional Correlation (DCC) – GARCH and a two-step Markov – Regime – Switching – GARCH as a unified analytical approach. Third, spectral causality graphs depicting relationships between stock indices and national output proxies revealed benign business cycle effects, thus, contributing to broadening the scope of business cycle theory