Newest papers

Stealing profits: Spread-based temporal hierarchy forecasting for day-ahead electricity markets

Arkadiusz Lipiecki, Nikolaos Kourentzes, Rafal Weron · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2609.23223 · cited by 0 works (OpenAlex) · open-access PDF

Indonesian

Prediksi harga listrik sehari-hari mendukung keputusan perdagangan dan penyimpanan, tetapi untuk arbitrase baterai memprediksi spread harga intraday lebih relevan daripada memprediksi harga jam individu. Di sini kami menunjukkan bahwa kerangka prediksi hierarki waktu (THieF) yang bersama-sama menyesuaikan prediksi harga listrik jam dan semua spread harga intraday secara konsisten meningkatkan kinerja di dua pasar listrik Eropa utama dan tiga arsitektur prediksi yang berbeda. Menggunakan lima tahun data sampel dari Jerman dan Spanyol, kami mendapatkan peningkatan akurasi hingga 19,7% dan keuntungan keuntungan hingga 10,4% sehubungan dengan prediksi harga jam yang tidak disesuaikan. Hasilnya berlanjut bahkan untuk model dasar TabPFN yang sangat akurat. Hasil kami menunjukkan bahwa memanfaatkan hubungan koheren antara target prediksi yang relevan secara ekonomi dapat meningkatkan akurasi prediksi dan nilai keputusan, dan yang lebih baik untuk statistik tidak perlu membuat keputusan ekonomi

Machine translation (Meta M2M100); the English original follows.

Abstract

Day-ahead electricity price forecasts support trading and storage decisions, but for battery arbitrage predicting intraday price spreads is more relevant than predicting individual hourly prices. Here we show that a temporal hierarchy forecasting (THieF) framework that jointly reconciles forecasts of hourly electricity prices and all intraday price spreads consistently improves performance across two major European electricity markets and three different forecasting architectures. Using five years of out-of-sample data from Germany and Spain, we obtain accuracy improvements of up to 19.7% and profit gains of up to 10.4% relative to unreconciled hourly price forecasts. The gains persist even for a highly accurate pretrained TabPFN foundation model. Our results demonstrate that exploiting coherent relationships between economically relevant forecasting targets can improve both predictive accuracy and decision value, and that better statistical forecasts do not necessarily imply better economic decisions.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Arbitrage, Electricity, Electricity price forecasting, Econometrics, Electricity market, Economics

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange