Forecasting Volatility in Commodity Market: Application of Select GARCH Models

Abstract

Commodity prices are volatile and volatility varies over time. Investing in commodities has generated heavy returns and has become increasingly popular, in spite of the high risks associated. The issue of volatility in commodity markets has attracted discussions at international forums too. This study is conducted to assess the volatility related issues of spot and future indices (metal, energy and agriculture) of MCX of India and to gauge the existence of leverage effect in the select commodity indices. It explores various measures of volatility spanning from November 1, 2005 to March 31, 2014. It has applied various ARCH family models such as Generalized Autoregressive Conditional Heteroscedasticity (GARCH), Exponential GARCH (EGARCH) and Component GARCH (CGARCH) models.

The results of GARCH indicate high persistence of volatility and news sensitivity. For metal spot, energy spot and metal future persistence is extremely high but for agriculture spot and energy future indices do not portray meaningful result. EGARCH model indicates that there is leverage effect in energy spot, agriculture spot and metal future but other indices do not verify presence of leverage effect. The results of CGARCH model indicate towards subsistence of trend and transitory component of volatility in all the indices except energy future.

Keywords: commodity, MCX, GARCH