Abstract/Summary

The volume of trading activity relating to China's commodity futures has grown rapidly over the course of the last decade. To improve risk management in China's commodity futures markets, this paper employs a regular vine (R-vine) copula model to study the dependence structure of commodity futures and to enhance Value-at-Risk (VaR) forecast. In doing so, we find that China’s commodity futures market is not centered on one category of commodity futures and the tail dependence between different categories of commodity futures varies significantly. Based on the dependence structure analyzed using the R-vine copula model, we forecast the VaR of individual indices, which are formed of several commodity futures, as well as forecasting the VaR of an equally-weighted portfolio. Our method can outperform the standard GARCH-VaR method in terms of VaR backtesting. The tool developed within this study will enable those involved in commodity futures markets to fundamentally improve their risk management.