Seminar co-organized by the Department of Management & Technology and by the Department of Marketing

Abstract:  Suppose you ask several experts about their forecasts and use the average as your prediction. Do you expect your prediction to be more or less accurate if the expert’s predictions vary a lot? We show that the answer depends on the distribution of forecasts. Predictions based on the average of variable observations are generally more accurate when the observations are drawn from distributions with lower kurtosis than the normal distribution and less accurate for distributions with higher kurtosis. Using this result, we characterize how the accuracy of a collective prediction - based on the average of several forecasts - varies with the variability of these forecasts. More variable forecasts are less accurate when the distribution of forecasts is strongly peaked around a mode close to the truth. More variable forecasts are more accurate only when the distribution of forecasts is flat-topped or skewed, or when there is heterogeneity in how correlated forecasts are.