Abstract
Accurate short-term electricity price forecasting is crucial forstrategically scheduling demand and generation bids in day-ahead markets. Whiledata-driven techniques have shown considerable prowess in achieving highforecast accuracy in recent years, they rely heavily on the quality of inputcovariates. In this paper, we investigate whether asynchronously publishedprices as a result of differing gate closure times (GCTs) in some bidding zonescan improve forecasting accuracy in other markets with later GCTs. Using astate-of-the-art ensemble of models, we show significant improvements of 22%and 9% in forecast accuracy in the Belgian (BE) and Swedish bidding zones (SE3)respectively, when including price data from interconnected markets withearlier GCT (Germany-Luxembourg, Austria, and Switzerland). This improvementholds for both general as well as extreme market conditions. Our analysis alsoyields further important insights: frequent model recalibration is necessaryfor maximum accuracy but comes at substantial additional computational costs,and using data from more markets does not always lead to better performance - afact we delve deeper into with interpretability analysis of the forecastmodels. Overall, these findings provide valuable guidance for marketparticipants and decision-makers aiming to optimize bidding strategies withinincreasingly interconnected and volatile European energy markets.