Interpretation of point forecasts with unknown directive

Point forecasts can be interpreted as functionals (i.e., point summaries) of predictive distributions. We extend methodology for the identification functional based on time series and associated realizations. Focusing state-dependent quantiles expectiles, we provide a generalized method moments estimator functional, along with tests optimality under general joint hypotheses relationships information bases. Our are more flexible, in simulations better calibrated powerful than existing solutions. In empirical examples, economic growth model output precipitation indicative overstatement anticipation extreme events.