This book has analysed the situation in which an expert receives a model-based forecast that he or she may decide to modify, driven by domain-specific knowledge but perhaps also by unobservable and non-replicable intuition. In various chapters it was assumed that an analyst had access to both forecasts, while in Chapters 3 and 5 the situation was also considered where an analyst would only observe the expert-adjusted forecasts and not the model that may have been used by the expert. In that case, the analyst could revert to his or her own econometric skills to create model-based forecasts in order to be able to assess the accuracy and merits of the expert-adjusted forecasts.

This chapter first summarizes the main findings in this book, based on the empirical evidence reported in the relevant literature. Then the limitations are discussed, together with some potential implications for model-builders, experts and analysts. Next, it will provide an outlook for future research, which is mainly driven by the observation that most available forecasts in macroeconomics and business are not created only by an econometric model and many if not all of these forecasts amount to expert-adjusted forecasts.