Testing the Performance of Multiproduct Optimal Hedging with Time-Varying Correlations in Storable and Non-storable Commodities

Recent steady growth in the volatility of commodity markets, and the increasing need for proper risk management tools in production settings that make use of inputs and outputs in futures markets, may be addressed via multiproduct hedging. This study determines and contrasts the effectiveness of multiproduct optimal hedging – that incorporate time-varying correlations – between storable and non-storable commodity settings, especially during recent periods of increased volatility. A soybean complex is considered for storable production-related commodities, and a feedlot operator is considered for non-storable production-related commodities. br br Multiproduct optimal time-varying hedge ratios are determined via a multivariate state dependent model of regime switching dynamic correlations. This model estimates time-varying correlations for multiple series in different correlation regimes (i.e., the conditional correlations matrix is not constant in this model). Two correlation regimes are estimated for the time periods considered, for both storable and non-storable production settings. More importantly, significant improvement of multiproduct hedging is determined for the storable commodity setting – soybean complex- over simple hedging strategies with time-varying correlations and the naïve strategy (1:1 hedge ratio). However, there is no significant improvement found for the nonstorable commodity setting – feedlot operator – over simple hedging strategies with time-varying corroborated using two different data sets for cash prices of feeder and live cattle.