Hyndman and Koehler (2006) recommend that the Mean Absolute Scaled Error (MASE) should become the standard when comparing forecast accuracies. This note supports their claim by showing that the MASE fits nicely within the standard statistical procedures initiated by Diebold and Mariano (1995) for testing equal forecast accuracies. Various other criteria do not fit, as they do not imply the relevant moment properties, and this is illustrated in some simulation experiments.

Additional Metadata
Keywords forecast accuracy, forecast error measures, statistical testing
Persistent URL dx.doi.org/10.1016/j.ijforecast.2015.03.008, hdl.handle.net/1765/78815
Series Econometric Institute Reprint Series
Journal International Journal of Forecasting
Franses, Ph.H.B.F. (2015). A note on the Mean Absolute Scaled Error. International Journal of Forecasting, 32, 20–22. doi:10.1016/j.ijforecast.2015.03.008