Abstract:
This paper develops a new approach for evaluating multi-step system forecasts with relatively few forecast-error observations. It extends the work of Clements and Hendry (1993) by using that of 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 with these methods and provides correct forecast rankings.
Citation:
Hendry, D. & Martinez, A. (2016. 'Evaluating multi-step system forecasts with relatively few forecast-error observations'. International Journal of Forecasting, 33, 359-372.