Was the Babson Prize deserved? An enquiry into an early forecasting model

In this paper, an early price-forecasting model (Sarle, American Economic Review; 1925, 15, no. 3, Supplement 2, 1-22) is evaluated. The model was one of the first attempts at the systematic specification of a multivariate price-forecasting model. The model received the Babson Prize offered by the American Economic Association in 1924 and was published in a special supplement of the American Economic Review. Despite the profile given to the model, it was virtually ignored by the economic profession. We evaluate the contribution of the model, arguing that it was well specified, using a step-wise correlation procedure to define the regression variables, that it provided accurate price forecasts compared with a contemporary AR model and that it represented the frontier of price forecasting in 1924. We discuss the model in an historical context and in the development of commodity-price-forecasting models.