This paper proposes an approach for estimating the uncertainty associated with model-based macroeconomic forecasts. We argue that estimated forecast intervals should account for the uncertainty arising from selecting the specification of an empirical forecasting model from the sample data. To allow this uncertainty to be considered systematically, we formalize a model selection procedure that specifies the lag structure of a model and accounts for aberrant observations. The procedure can be used to bootstrap the complete model selection process when estimating forecast intervals. We apply the procedure to assess the risk of deflationary developments occurring in Germany over the next four years.
forecasting prediction intervals bootstrapping deflation model selection