Reanalysing Price Asymmetries in the Nordic Intraday Market

Abstract

This paper investigates whether positive forecast errors influence the intraday price differently than negative forecast errors. The regression analysis focuses on the Nordic intraday market. With this paper we show that no price asymmetries exist, and that previous conclusions are non-robust to reanalysis with a different model specification. The model specification in this paper solves the issue with autocorrelation, which have troubled previous studies. Specifically, the order of autoregressive and moving average parts depend on the time of day, since the underlying autocorrelation structure of the price data differs noticeably with the time of day.