Multivariate Analysis of the Dynamic Correlations among Global Commodity Prices, Exchange Rate and Stock Market Returns
L. E. George *
Ignatius Ajuru University of Education, Rumuolumeni, Port Harcourt, Rivers State, Nigeria.
J. B. Nwikpe
Ignatius Ajuru University of Education, Rumuolumeni, Port Harcourt, Rivers State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
This study examined the dynamic correlations among global commodity prices, the exchange rate and stock market returns in Nigeria. The study was motivated by the need to understand the evolution of the relationships among global commodity and financial markets over time, particularly during periods of economic and financial uncertainty, and the increasing integration of these systems. The study employed an ex post facto research design, using daily secondary data from the Central Bank of Nigeria, the Nigerian Exchange Group and recognised global financial databases for the period from 29 December 2009 to 7 November 2025. The study employed a multivariate t-distribution-based Dynamic Conditional Correlation-Generalized Autoregressive Conditional Heteroskedasticity (DCC-GARCH) model and a Vector Autoregressive (VAR) model. The VAR lag-selection criteria indicated a one-period lag, and the results identified significant short-run interactions involving copper and gold returns in the Brent crude oil equation, lagged natural gas returns in the natural gas equation, and lagged natural gas and stock market returns in the Nigerian stock-market equation. The DCC-GARCH results showed substantial volatility persistence across the six markets, with persistence coefficients of 0.9864 for Brent crude oil, 0.9811 for gold, 0.9990 for natural gas, 0.9922 for copper, 0.9109 for stock market returns and 0.9990 for the exchange rate. The estimated DCC parameters were statistically significant, with a correlation innovation coefficient of 0.00668 and a correlation persistence coefficient of 0.99041, indicating that the conditional correlation structure changes in response to new information but remains highly persistent over time. The results also showed that correlations varied considerably across market pairs, with average correlations ranging from -0.0972 between natural gas and exchange rate returns to 0.2164 between gold and copper returns. The correlations also moved between negative and positive values over the study period, indicating that the relationships were dynamic rather than constant. The study concludes that global commodity prices, the exchange rate and Nigerian stock market returns are interconnected through time-varying and persistent correlation structures.
Keywords: Multivariate analysis, dynamic correlations, global commodity prices