Heavy-tailed behavior of commodity price distribution and optimal hedging demand

Abstract

This study explores the importance of imposing a correct distributional hypothesis in a risk management strategy, by comparing hedge ratios under the restrictive normality assumption to those under the generalized stable distribution. Concepts are illustrated for the case of a representative Pennsylvania dairy farm manager who purchases corn as a feed input. The results show that time processes of corn prices and basis risk in five Pennsylvania regions do not correspond to the normal distribution, and they more correctly correspond to one of the stable distribution set. The estimated hedge ratios under the stable distribution are typically larger than those under the normal distribution. The difference would be a bias from imposing a wrong distributional assumption.