The Forecast Combination Puzzle: A Simple Theoretical Explanation

18 Pages Posted: 23 Oct 2013 Last revised: 1 Sep 2014

Gerda Claeskens

KU Leuven - Department of Economics

J.R. Magnus

VU University Amsterdam - Faculty of Economics and Business Administration

Andrey L. Vasnev

University of Sydney

Wendun Wang

Erasmus University Rotterdam (EUR)

Multiple version iconThere are 3 versions of this paper

Date Written: August 31, 2014

Abstract

This papers offers a theoretical explanation for the stylized fact that forecast combinations with estimated optimal weights often perform poorly in applications. The properties of the forecast combination are typically derived under the assumption that the weights are fixed, while in practice they need to be estimated. If the fact that the weights are random rather than fixed is taken into account during the optimality derivation, then the forecast combination will be biased (even when the original forecasts are unbiased) and its variance is larger than in the fixed-weights case. In particular, there is no guarantee that the ‘optimal’ forecast combination will be better than the equal-weights case or even improve on the original forecasts. We provide the underlying theory, some special cases and an application in the context of model selection.

Keywords: forecast combination, optimal weights, model selection

JEL Classification: C53, C52

Suggested Citation

Claeskens, Gerda and Magnus, J.R. and Vasnev, Andrey L. and Wang, Wendun, The Forecast Combination Puzzle: A Simple Theoretical Explanation (August 31, 2014). Available at SSRN: https://ssrn.com/abstract=2342841 or http://dx.doi.org/10.2139/ssrn.2342841

Gerda Claeskens

KU Leuven - Department of Economics ( email )

Leuven, B-3000
Belgium

Jan R. Magnus

VU University Amsterdam - Faculty of Economics and Business Administration ( email )

De Boelelaan 1105
Amsterdam, 1081HV
Netherlands

Andrey L. Vasnev (Contact Author)

University of Sydney ( email )

Sydney, NSW 2006
Australia

HOME PAGE: http://www.econ.usyd.edu.au/staff/andreyv

Wendun Wang

Erasmus University Rotterdam (EUR) ( email )

Burgemeester Oudlaan 50
3000 DR Rotterdam, Zuid-Holland 3062PA
Netherlands

Register to save articles to
your library

Register

Paper statistics

Downloads
67
rank
225,643
Abstract Views
534
PlumX