Forecast evaluation tests and negative long-run variance estimates in small samples

In this paper, we show that when computing standard Diebold-Mariano-type tests for equal forecast accuracy and forecast encompassing, the long-run variance can frequently be negative when dealing with multi-step-ahead predictions in small, but empirically relevant, sample sizes. We subsequently cons...

Full description

Bibliographic Details
Main Authors: Harvey, David I., Leybourne, Stephen J., Whitehouse, Emily J.
Format: Article
Published: Elsevier 2017
Subjects:
Online Access:https://eprints.nottingham.ac.uk/43017/
_version_ 1848796623875866624
author Harvey, David I.
Leybourne, Stephen J.
Whitehouse, Emily J.
author_facet Harvey, David I.
Leybourne, Stephen J.
Whitehouse, Emily J.
author_sort Harvey, David I.
building Nottingham Research Data Repository
collection Online Access
description In this paper, we show that when computing standard Diebold-Mariano-type tests for equal forecast accuracy and forecast encompassing, the long-run variance can frequently be negative when dealing with multi-step-ahead predictions in small, but empirically relevant, sample sizes. We subsequently consider a number of alternative approaches to dealing with this problem, including direct inference in the problem cases and use of long-run variance estimators that guarantee positivity. The finite sample size and power of the different approaches are evaluated using extensive Monte Carlo simulation exercises. Overall, for multi-step-ahead forecasts, we find that the recently proposed Coroneo and Iacone (2016) test, which is based on a weighted periodogram long-run variance estimator, offers the best finite sample size and power performance.
first_indexed 2025-11-14T19:50:56Z
format Article
id nottingham-43017
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:50:56Z
publishDate 2017
publisher Elsevier
recordtype eprints
repository_type Digital Repository
spelling nottingham-430172020-05-04T19:15:40Z https://eprints.nottingham.ac.uk/43017/ Forecast evaluation tests and negative long-run variance estimates in small samples Harvey, David I. Leybourne, Stephen J. Whitehouse, Emily J. In this paper, we show that when computing standard Diebold-Mariano-type tests for equal forecast accuracy and forecast encompassing, the long-run variance can frequently be negative when dealing with multi-step-ahead predictions in small, but empirically relevant, sample sizes. We subsequently consider a number of alternative approaches to dealing with this problem, including direct inference in the problem cases and use of long-run variance estimators that guarantee positivity. The finite sample size and power of the different approaches are evaluated using extensive Monte Carlo simulation exercises. Overall, for multi-step-ahead forecasts, we find that the recently proposed Coroneo and Iacone (2016) test, which is based on a weighted periodogram long-run variance estimator, offers the best finite sample size and power performance. Elsevier 2017-11-01 Article PeerReviewed Harvey, David I., Leybourne, Stephen J. and Whitehouse, Emily J. (2017) Forecast evaluation tests and negative long-run variance estimates in small samples. International Journal of Forecasting, 33 (4). pp. 833-847. ISSN 0169-2070 Forecast evaluation; Long-run variance estimation; Simulation; Diebold-Mariano test; Forecasting http://www.sciencedirect.com/science/article/pii/S0169207017300559 doi:10.1016/j.ijforecast.2017.05.001 doi:10.1016/j.ijforecast.2017.05.001
spellingShingle Forecast evaluation; Long-run variance estimation; Simulation; Diebold-Mariano test; Forecasting
Harvey, David I.
Leybourne, Stephen J.
Whitehouse, Emily J.
Forecast evaluation tests and negative long-run variance estimates in small samples
title Forecast evaluation tests and negative long-run variance estimates in small samples
title_full Forecast evaluation tests and negative long-run variance estimates in small samples
title_fullStr Forecast evaluation tests and negative long-run variance estimates in small samples
title_full_unstemmed Forecast evaluation tests and negative long-run variance estimates in small samples
title_short Forecast evaluation tests and negative long-run variance estimates in small samples
title_sort forecast evaluation tests and negative long-run variance estimates in small samples
topic Forecast evaluation; Long-run variance estimation; Simulation; Diebold-Mariano test; Forecasting
url https://eprints.nottingham.ac.uk/43017/
https://eprints.nottingham.ac.uk/43017/
https://eprints.nottingham.ac.uk/43017/