Accounting for Heterogeneity in Hedging Behavior: Comparing & Evaluating Grouping Methods

Heterogeneity, i.e., the notion that individuals respond differently to economic stimuli, can have profound consequences for the interpretation of behavior and the formulation of agricultural policy. This paper compares and evaluates three grouping techniques that can be used to account for heterogeneity in financial behavior. Two are well established: company-type grouping and cluster analysis. A third, the generalized mixture regression model, has recently been developed and is worth considering as market participants are grouped such that their response to the determinants of economic behavior is similar. We evaluate the grouping methods in a hedging framework by assessing their ability to reflect relationships consistent with theory. The empirical findings show that the economic relationships are more consistent with theory within the groups identified by the mixture model, and suggest that researchers interested in identifying segments of the population in which participants behave in a similar manner may consider using of mixture model in the presence of heterogeneity in financial behavior.