Stealing Accuracy: Predicting Day-ahead Electricity Prices with Temporal Hierarchy Forecasting (THieF)

2025-08-15

AI Summary

To address the inconsistency and lack of joint optimization across multiple time scales—hourly, block-level (2–12 hours), and baseload contract-level—in day-ahead electricity price forecasting, this paper introduces, for the first time, the Temporal Hierarchies Forecasting (THieF) framework to electricity market price prediction. We propose a cross-granularity coordinated modeling approach that integrates diverse base models—including linear regression, shallow neural networks, gradient-boosted trees, and Transformers—and enforces hierarchical consistency via a unified reconciliation algorithm. Evaluated on four years of German day-ahead market data, our method reduces mean absolute forecasting error by up to 13% compared to single-scale benchmarks, while maintaining computational overhead comparable to conventional uniscale approaches. This work not only validates THieF’s efficacy in electricity price forecasting but also establishes a scalable, model-agnostic paradigm for multi-scale temporal forecasting in energy markets.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Time-Series/Data Streams
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

We introduce the concept of temporal hierarchy forecasting (THieF) in predicting day-ahead electricity prices and show that reconciling forecasts for hourly products, 2- to 12-hour blocks, and baseload contracts significantly (up to 13%) improves accuracy at all levels. These results remain consistent throughout a challenging 4-year test period (2021-2024) in the German power market and across model architectures, including linear regression, a shallow neural network, gradient boosting, and a state-of-the-art transformer. Given that (i) trading of block products is becoming more common and (ii) the computational cost of reconciliation is comparable to that of predicting hourly prices alone, we recommend using it in daily forecasting practice.

Problem

• Improving day-ahead electricity price prediction accuracy
• Reconciling forecasts for multiple temporal product levels
• Validating results across diverse models and market conditions

Innovation

• Temporal hierarchy forecasting (THieF) method
• Reconciles multi-level electricity price forecasts
• Consistent accuracy across diverse model architectures