Abstract
This study compares various techniques for computing the optimal ratio for hedging cotton and soybean price risk using futures contracts. Using a ratio similar to Sharpe's measure to evaluate the performance of various hedging methods, we find that these different methods give different results depending on the duration of the hedge. There is no clear-cut winner as far as hedging techniques are concerned. For soybeans, based on our time period, the use of regression on price levels produces the best results. On a risk-adjusted basis, the no-hedge option performs the worst in every scenario, indicating that hedging of some sort--and preferably not the full hedge (naive hedge ratio of 1)--is essential. The more complicated modified regression model and the error-correction method do not improve the effectiveness of the hedge for either commodity, and a hedger is better off with the traditional regression approaches.