Facebook LinkedIn Youtube Twitter

IAEE Members and subscribers to The Energy Journal: Please log in to access the full text article or receive discounted pricing for this article.

Prepress Content: The following article is a preprint of a scientific paper that has completed the peer-review process and been accepted for publication within The Energy Journal.

While the International Association for Energy Economics (IAEE) makes every effort to ensure the veracity of the material and the accuracy of the data therein, IAEE is not responsible for the citing of this content until the article is actually printed in a final version of The Energy Journal. For example, preprinted articles are often moved from issue to issue affecting page numbers, and actual volume and issue numbers. Care should be given when citing Energy Journal preprint articles.

Size Matters: Estimation Sample Length and Electricity Price Forecasting Accuracy

Short-term electricity price forecasting models are typically estimated via rolling windows, i.e. by using only the most recent observations. Nonetheless, the literature does not provide guidelines on how to select the optimal size of such windows. This paper shows that determining the appropriate window prior to estimation dramatically improves forecasting performances. In addition, it proposes a simple two-step approach to choose the best performing models and window sizes. The value of this methodology is illustrated by analyzing hourly datasets from two large power markets (Nord Pool and IPEX) with a selection of eleven different forecasting models. Incidentally, our empirical application reveals that simple models, such as a simple linear regression (SLR) with only two parameters, can perform unexpectedly well if estimated on extremely short samples. Surprisingly, in the Nord Pool, such SLR is the best performing model in 13 out 24 trading periods.

Download Executive Summary Purchase ( $25 )

Keywords: Electricity price forecasting, Day-ahead market, Parameter instability, Bandwidth selection, Statistical models, Artificial neural networks.

DOI: 10.5547/01956574.41.4.cfez

References: Reference information is available for this article. Join IAEE, log in, or purchase the article to view reference data.

Published in Volume 41, Number 4 of the bi-monthly journal of the IAEE's Energy Economics Education Foundation.