The value of commodity price forecasts in the presence of futures under uncertainty

This thesis examines the value obtained from the implementation of energy commodity forecasts in the context of a hedging decision. A decision-analytic model to value commodity price forecasts in the presence of futures is developed and subsequently applied to a data set of crude oil and natural gas prices. The findings include a method for evaluating the added value of such forecasts as well as appropriate improvements in forecast attributes to increase forecast value. Furthermore, it is found that for forecasts to be valuable, they must be accurate at predicting both gains and losses, and that there are positive and diminishing marginal returns to forecast value from improvement in key measure of accuracy in most cases. Forecast value is specific to user class, and that value is unique to specific users within each class. Lastly, the inclusion of an exponential utility function into the decision-analytic model enhances the reality of the decision characteristics of actual producers and consumers of crude oil. It is found that a producer or consumer of the commodity must be highly risk averse for price forecasts to have value when considering profit maximization alone among price forecast benefits.