Iterated Multi-Step Forecasting with Model Coefficients Changing Across Iterations

Iterated multi-step forecasts are usually constructed assuming the same model in each forecasting iteration. In this paper, the model coefficients are allowed to change across forecasting iterations according to the in-sample prediction performance at a particular forecasting horizon. The technique can thus be viewed as a combination of iterated and direct forecasting. The superior point and density forecasting performance of this approach is demonstrated on a standard medium-scale vector autoregression employing variables used in the Smets and Wouters (2007) model of the US economy. The estimation of the model and forecasting are carried out in a Bayesian way on data covering the period 1959Q1-2016Q1.