Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/152648
Authors: 
Inoue, Atsushi
Kilian, Lutz
Year of Publication: 
2003
Series/Report no.: 
ECB Working Paper 214
Abstract: 
It is standard in applied work to select forecasting models by ranking candidate models by their prediction mean square error (PMSE) in simulated ou-of-sample (SOOS) forecasts. Alternatively, forecast models may be selected using information criteria (IC). We compare the asymptotic and finite-sample properties of these methods in terms of their ability to minimize the true out-of-sample PMSE, allowing for possible misspecification of the forecast models under consideration. We first study a covariance stationary environment. We show that under suitable conditions the IC method will be consistent for the best approximating models among the candidate models. In contrast, under standard assumptions the SOOS method will select overparameterized models with positive probability, resulting in excessive finite-sample PMSEs. We also show that in the presence of unmodelled structural change both methods will be inadmissible in the sense that they may select a model with strictly higher PMSE than the best approximating models among the candidate models.
Subjects: 
Forecast accuracy
Information criteria
Model Selection
Simulated out-of-sample method
Structural change
JEL: 
C22
C52
C53
Document Type: 
Working Paper

Files in This Item:
File
Size
670.61 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.