We solve the maximization problem numerically based on simulated yield and price
data for the next five years. Stochastic trends are used in the simulation of Grant County yield,
Portland cash price, and CBOT futures price, based on historical data.
We find optimal solutions for farmers in both Whitman County and Grant County vary
with model specifications, reinforcing the importance of appropriate model selection and
parameterization. Comparing the GEU model with other EU models shows that the general form
of GEU has advantages in incorporating more preference information about the decision maker.
The commonly used MA-EU model gives almost the same results when the risk aversion is
specified at the same level as in the CES-EU, indicating that these two types of models might be
interchangeable. However, these results are different than the GEU model when the preferences
parameters are set at different levels. To conclude, (1) GEU is more general and can incorporate
more flexible preference, (2) the commonly used additive EU models may yield biased results
relative to the decisions based on the true preference. The results are completely different in the
risk neutral and perfect substitution MR-GEU setting.
The optimal choice of the hedging ratios is around 30% and that of the crop insurance
purchase is always 85% in both counties. These levels are in line with the existing static one
period studies. The subsidy in crop insurance overshadows its risk management feature so that
the optimal insurance coverage is invariant with respect to the preference alternatives.
Based on GEU framework, we investigate the impacts of intertemporal preferences,
hedging and crop insurance costs, and U.S. government payment programs on the risk
management behavior of a Whitman County wheat producer.
The GEU framework has flexibility in the parameterization of the farmer’s preferences
towards risk, timing, and intertemporal substitutability of consumption. We employ this feature
30