The value of selecting the best forecasting model as the basis for empirical economic policy analysis is questioned. When no model coincides with the data generation process, non-causal statistical devices may provide the best available forecasts: examples from recent work include intercept...

Since forecast failure is due to unanticipated large shifts in deterministic factors,'sensible' agents should adopt 'robust forecasting rules'. Unless the model coincides with the generating mechanism, one cannot even prove that causal variables will dominate non-causal in forecasting. In such a...

We consider forecasting using a combination, when no model coincides with a non-constant data generation process (DGP). Practical experience suggests that combining forecasts adds value, and can even dominate the best individual device. We show why this can occur when forecasting models are...

Identification is an essential attribute of any model's parameters, so we consider its three aspects of 'uniqueness', 'correspondence to reality' and 'interpretability'. Observationally-equivalent over-identified models can co-exist, and are mutually encompassing in the population;...

This paper describes some recent advances and contributions to our understanding of economic forecasting. The framework we develop helps explain the findings of forecasting competitions and the prevalence of forecast failure. It constitutes a general theoretical background against which recent...

We examine the properties of automatic model selection, as embodied in PcGets, and evaluate its performance across different (unknown) states of nature. After describing the basic algorithm and some recent changes, we discuss the consistency of its selection procedures, then examine the extent...