Modelling the dependency structure among the BRICS market returns using Copula ARMA-EGARCH approach

This study investigates the dependency structure among BRICS market returns using a Copula-based ARMA-EGARCH modelling framework. The stationarity tests using ADF and PP confirmed that the closing prices are non-stationary at level but stationary after first differencing. Pearson correlation and scatter plots indicate strong positive linear relationships among the markets, suggesting co-movement. Given the evidence of tail dependence, copula models are employed to capture both upper and lower tail dependencies. ARMA models were selected based on AIC, and diagnostic tests confirmed their adequacy. The hybrid ARMA-EGARCH(1,1) model was used to model volatility, capturing both positive and negative leverage effects across different markets. Residuals from these models were transformed into uniform distributions and used to fit four copulas, namely, Student-t, Frank, Clayton, and Gumbel. Results showed that the Student-t copula best captures the dependence structure, with significant dependence parameters for all pairs. Lower and upper tail dependencies were captured by Clayton and Gumbel copulas respectively, indicating asymmetric behaviour during market extremes. The findings suggest strong financial integration among BRICS markets, limiting diversification benefits and increasing contagion risk. The study recommends the use of flexible distributions and more advanced copula models like Joe-Clayton in future research to enhance Value-at-Risk estimation and risk management practices.