PREDICTING TURNING-POINTS THROUGH THE INTEGRATION OF MULTIPLE MODELS

A new method for forming composite turning-point (or other qualitative ) forecasts is proposed. Rather than forming composite forecasts by th e standard Bayesian approach with weights proportional to each model's posterior odds, weights are assigned to the individual models in prop ortion to the probability of each model's having the correct turning-p oint prediction. These probabilities are generated by legit models est imated with data on the models' past turning-point forecasts. An empir ical application to gross national product/gross domestic product fore casting of 18 Organization for Economic Cooperation and Development co untries demonstrates the potential benefits of the procedure.