Finite Sample Forecast Properties and Window Length Under Breaks in Cointegrated Systems
Essays in Honor of M. Hashem Pesaran: Prediction and Macro Modeling
ISBN: 978-1-80262-062-7, eISBN: 978-1-80262-061-0
Publication date: 18 January 2022
Abstract
The author shows that extending the estimation window prior to structural breaks in cointegrated systems can be beneficial for forecasting performance and highlights under which conditions. In doing so, the author generalizes the Pesaran and Timmermann (2005)’s forecast error decomposition and shows that it depends on four terms: (1) a period ahead risk; (2) a bias due to a conditional mean shift; (3) a bias due to a variance mismatch; (4) a gap term valid only conditionally. The author also derives new expressions for the estimators of the adjustment matrix and a constant, which are auxiliary to the decomposition. Finally, the author introduces new simulation-based estimators for the finite sample forecast properties which are based on the derived decomposition. The author’s finding points out that, in some cases, parameter instability can be neglected by extending the window backward and forecasters can be insured against higher forecast risk under this model class as well, generalizing Pesaran and Timmermann (2005)’s result. The author’s result gives renewed importance to break tests, in order to distinguish cases when break-neglection is (not) appropriate.
Keywords
Acknowledgements
Acknowledgments
This chapter was developed during my PhD under the supervision of Uwe Hassler at Goethe University and at the Deutsche Bundesbank (Research Centre). I thank Uwe Hassler, Hashem Pesaran, Allan Timmermann (the editor), a referee, Daniel Gutknecht, Kim-Kurz Jeong-Ryeol, Alessandra Luati, Paolo Santucci de Magistris, Xu Han, Mehdi Hosseinkouchack, Ying Lun Cheung, Esteban Prieto, Christoph Meinerding, and Benny Hartwig for very insightful comments that have significantly improved the quality of the paper. Also, I thank the participants of the 28th SNDE symposium 2020, the NBER-NSF time series conference 2019, the 4th Vienna time series workshop 2019 and the workshop “Hot topics in econometrics” at Goethe University for useful discussions. An earlier version of this article was published as discussion paper at the Granger Centre for Time Series Econometrics. The views expressed in this chapter are my own and do not necessarily represent the views of the European Central Bank or the Deutsche Bundesbank.
Citation
Nocciola, L. (2022), "Finite Sample Forecast Properties and Window Length Under Breaks in Cointegrated Systems", Chudik, A., Hsiao, C. and Timmermann, A. (Ed.) Essays in Honor of M. Hashem Pesaran: Prediction and Macro Modeling (Advances in Econometrics, Vol. 43A), Emerald Publishing Limited, Leeds, pp. 167-196. https://doi.org/10.1108/S0731-90532021000043A009
Publisher
:Emerald Publishing Limited
Copyright © 2022 Luca Nocciola