The USDA WASDE (World Agricultural Supply and Demand Estimates) price forecasts are published in the form of an interval, but typically analyzed as point estimates. Thus, all information about uncertainty imbedded in the forecast is ignored. The purpose of this paper is to evaluate the accuracy of WASDE price forecasts using methodology suitable for testing judgmental interval forecasts. Empirical analysis includes traditional statistical tests as well as an alternative behavioral evaluation (accuracy-informativeness tradeoff model). The results of the traditional analysis indicate overconfidence of WASDE price interval forecasts, while the results of the behavioral approach suggest rational inaccuracy.


Downloads Statistics

Download Full History