A re-evaluation of the quasi-Bayes approach to the linear combination of forecasts

The subjective forecasts used in decision analysis should, in principle, synthesize all available evidence about the subject in analysis. In this manner, when part of the evidence consists of a variety of forecasting models or expert opinions, Decision Theory requires the decision maker to formulate a combination of these predictors. This work takes into account the Bayesian methodologies outperformance and quasi-Bayes, as well as the classical model of optimal combination, all applied to the linear combination of petroleum price forecasts, generated by experts from Petrobrás-the Brazilian oil company-for several international markets. It presents a theoretical description of the methodologies followed by a comparative analysis between performances of the best expert’s forecasts and combinations. The performances and features of these combinations are also comapred.