Long-run expectations in a Learning-to-Forecast Experiment

We conduct a Learning to Forecast Experiment (LtFE) using a novel setting in which we elicit subjects' short and long-run expectations on the future price of an asset. We find that: (i) the rational expectations equilibrium (REE) is not a meaningful description for subjects' expectations, (ii) which are, instead, better described by an adaptive learning scheme. (iii) Subjects exhibit a higher degree of inertia when revising long-run expectations vis-a-vis short-run expectations.