MULTIVARIATE GARCH APPROACH TO MODELING RETURNS ON EXCHANGE RATES AND CRUDE OIL PRICES IN NIGERIA

This study applies the multivariate Generalized Autoregressive Conditional Heteroscedasticity (MGARCH) model to analyze the dynamics of returns on the Naira/Dollar exchange rate and crude oil prices in Nigerian foreign markets. Data spanning 1991 to 2017 were sourced from the Central Bank of Nigeria (CBN) online statistical database. Preliminary time series analysis revealed the presence of volatility clustering, justifying the use of MGARCH. Model estimation using EViews 10 determined the maximum lags for exchange rate and crude oil prices to be 3 and 2, respectively. Cointegration analysis confirmed the existence of two cointegrating equations, while Error Correction Model (ECM) coefficients indicated that 31.41% and 62.78% of disequilibrium in the co-movements of exchange rate and crude oil returns are corrected within one period, with the ECM for exchange rate being significant at the 10% level. MGARCH results further revealed bidirectional volatility spillovers between the two markets, highlighting the interdependence and dynamic risk transmission of exchange rate and crude oil price returns. The findings underscore the importance of understanding volatility interactions for policymakers, investors, and risk managers in Nigeria’s foreign exchange and commodity mark