Regime-Dependent and Time-Varying Dependence in Naira Exchange Rate Volatility: A Copula-Based Inverse Gamma Approach

Abstract

This study investigates the distributional characteristics, regime behavior, and dependence structure of monthly Naira exchange rate volatility against the United States Dollar (USD), Euro (EUR), and British Pound (GBP) using data spanning January 2005 to December 2024. Exchange rate volatility is constructed from logarithmic returns and analyzed within a unified framework combining heavy-tailed marginal modeling, volatility-based regime classification, copula-based dependence measures, and rolling-window dynamics. Empirical results reveal that volatility across all currency pairs exhibits extreme right-skewness and excess kurtosis, providing strong evidence of heavy-tailed behavior inconsistent with Gaussian assumptions. The inverse gamma distribution is shown to adequately characterize the marginal distribution of volatility, particularly in capturing extreme realizations. Regime-specific analysis indicates that volatility dynamics are state-dependent, with high-volatility regimes associated with substantially larger volatility magnitudes and heavier tails. Dependence analysis based on Kendall’s tau and copula models reveals weak co-movement during low and medium volatility regimes but a sharp intensification of dependence during high-volatility regimes, suggesting crisis-driven contagion effects. Rolling-window estimates further confirm that dependence among exchange rate volatilities is time-varying and closely linked to periods of market stress. Overall, the findings highlight the importance of heavy-tailed, regime-aware, and dependence-sensitive frameworks for modeling exchange rate volatility in emerging market economies.