A Heuristic for Combining Correlated Experts

It is intuitive and theoretically sound to combine experts' forecasts based on their proven skills, while accounting for correlation among their submissions. Simpler combination methods, however, that assume independence of forecasts or equal skill, have been found to be empirically robust. One explanation for this is that empirical estimation of skill and correlation introduces error, leading to worse aggregated forecasts than simpler alternatives. We offer a heuristic that accounts for skill and reduces estimation error by assuming a common correlation among the experts. Our theoretical results present an optimal form for this common correlation and we offer Bayesian estimators to be used in practice. The common correlation heuristic is shown to outperform alternative combination methods on macroeconomic and experimental forecasting data.