Evaluating multi-step system forecasts with relatively few forecast-error observations

Working paper

This paper develops a new approach for evaluating multi-step system forecasts with relatively few forecast-error observations. It extends Clements and Hendry (1993a) using Abadir et al.(2014) to generate "design-free" estimates of the general matrix of the forecast-error second-moment when there are relatively few forecast-error observations. Simulations show that the usefulness of alternative methods deteriorates when their assumptions are violated. The new approach compares well against these methods and provides correct forecast rankings.

Authors: Hendry, D; Martinez, A

• Publication status: Published
• Publisher: University of Oxford
• Journal: International Journal of Forecasting
• Series: Department of Economics Discussion Paper Series
• Publication date: 2016-03-08
• DOI: 10.1016/j.ijforecast.2016.08.007
• ISSN: 0169-2070
• Paper number: 784
• Keywords: GFESM; MSFE; forecast evaluation; moment matrices; forecast error; invariance