Improved Tests for Forecast Comparisons in the Presence of Instabilities

Luis Martins and Pierre Perron

No 2014-003, Boston University - Department of Economics - Working Papers Series from Boston University - Department of Economics

Abstract: Giacomini and Rossi (2010) proposed a aÌ uctuations test and a one-time reversal test for comparing the out-of-sample forecasting performance of two competing models in the presence of possible instabilities. In the simulations and empirical applications, they use a version of their test based on the sample variance of the loss di§erences instead of the relevant, and advocated, long-run variance (HAC estimate). We repli- cate the power properties of their tests with the appropriate HAC correction using exactly the same design they used. We show that the power functions of the tests are substantially lower than what they report. More importantly the power functions are non-monotonic. To alleviate the power problems of their tests in the presence of instabilities in the di§erences of the loss functions, we suggest using simple tests (sup- Wald and UDmax) for changes in the mean of the loss-di§erences. These are shown to have higher monotonic power, especially the UDmax version. We use their empirical examples to show that the practical relevance of the issue raised.

Keywords: non-monotonic power; structural change; forecasts; long-run variance. (search for similar items in EconPapers)
Pages: 16
Date: 2014-03
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
http://www.bu.edu/econ/files/2009/07/Perron-Improv ... arisons-Jan-2014.pdf First version, 2014 (application/pdf)
Our link check indicates that this URL is bad, the error code is: 403 Forbidden (http://www.bu.edu/econ/files/2009/07/Perron-Improved-Tests-for-Forecast-Comparisons-Jan-2014.pdf [301 Moved Permanently]--> https://www.bu.edu/econ/files/2009/07/Perron-Improved-Tests-for-Forecast-Comparisons-Jan-2014.pdf)

Related works:
Journal Article: Improved Tests for Forecast Comparisons in the Presence of Instabilities (2016)
Working Paper: Improved Tests for Forecast Comparisons in the Presence of Instabilities (2015)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:bos:wpaper:wp2014-003

Access Statistics for this paper

More papers in Boston University - Department of Economics - Working Papers Series from Boston University - Department of Economics Contact information at EDIRC.
Bibliographic data for series maintained by Program Coordinator ().