Volatility Modeling of Commodity Markets in India: Application of Selected GARCH Models

The study focused on volatility modeling of commodity market in India based on the closing returns of indices of multi commodity exchange, that is, MCX AGRI, MCX METAL, MCX ENERGY, and MCX COMDEX during the period from April 1, 2013 to March 31, 2018. The study used symmetric and asymmetric models of auto regressive conditional heteroskedasticity (ARCH) family models. The study found significant high volatility persistence in all commodity indices of MCX. The asymmetric models of GARCH revealed a presence of leverage effect only in MCX ENERGY index and not in any other indices of MCX. Finally, AIC and SIC criteria were used to identify the best models that better described the volatility of the commodity market as the AIC and SIC values were the lowest in the most appropriate model. It was found that EGARCH (1, 1) model best fitted among all other models for MCX AGRI and MCX ENERGY; for MCX Metal, TARCH (1, 1) model was found to be the best fit; and for MCX COMDEX, GARCH(1, 1) model best described the volatility.