Sources of turning point forecast errors

Abstract
Little attention has been devoted to explaining the failure to predict the turning points at the beginning of recessions. This note builds on a model that showed that a turning point might not be predicted if forecasters' prior probabilities of a recession were low. Costs associated with various types of errors might also produce this result. A model is developed and then applied to a particular US business cycle indicator. The results show that costs as well as priors may contribute to this type of forecasting error.

This publication has 6 references indexed in Scilit: