All Rights Reserved
AccessEcon LLC 2006, 2008.
Powered by MinhViet JSC

Dirk Ulbricht
''It is not structural breaks that earn average forecasts their fame''
( 2016, Vol. 36 No.2 )
Structural change is a major challenge to the applied forecaster and a potential source of large forecast errors. Large forecasting competitions demonstrate the success of combined forecasts of simple linear models over forecasting devices that endogenously model structural change. Thereby, most studies look at the average performance over time, not at or around structural changes. However, is it really reliability in the presence of structural breaks that gives average forecasts an edge over their competitors? An analysis of real-time forecasts of UK inflation indicates that it is not their break performance that earns combined forecasts their fame.
Keywords: real-time experiment, forecast combination, breaks
JEL: C5 - Econometric Modeling: General
C2 - Single Equation Models; Single Variables: General
Manuscript Received : Nov 04 2014 Manuscript Accepted : Jun 22 2016

  This abstract has been downloaded 1430 times                The Full PDF of this paper has been downloaded 153560 times