Smoothing Quantile Regression Averaging: A new approach to probabilistic forecasting of electricity prices
arXiv:2302.00411 · doi:10.1016/j.jcomm.2025.100501
Abstract
Accurate short-term price forecasting is essential for daily operations in electricity markets. This article introduces a new method, called Smoothing Quantile Regression (SQR) Averaging, that improves upon well-performing probabilistic forecasting schemes. To demonstrate its utility, a comprehensive study is conducted on two electricity markets, including recent data covering the COVID-19 pandemic and the Russian invasion of Ukraine. The performance of SQR Averaging is evaluated both in terms of reliability and sharpness measures, and economic benefits from a trading strategy. The latter utilizes battery storage and sets limit orders using selected quantiles of the predictive distribution. SQR Averaging leads to profit increases of up to 3.5\% on average compared to the benchmark strategy based solely on point forecasts. This is strong evidence for the practical value of using probabilistic forecasts in day-ahead power trading, even in the face of the COVID-19 pandemic and geopolitical disruptions.
References in corpus (7)
- Neural basis expansion analysis with exogenous variables: Forecasting electricity prices with NBEATSx
- Day-ahead electricity price forecasting with high-dimensional structures: Univariate vs. multivariate modeling frameworks
- Forecasting Electricity Spot Prices using Lasso: On Capturing the Autoregressive Intraday Structure
- Operational Research: Methods and Applications
- Conformal Prediction Interval Estimations with an Application to Day-Ahead and Intraday Power Markets
- Optimal bidding in hourly and quarter-hourly electricity price auctions: trading large volumes of power with market impact and transaction costs
- Postprocessing of point predictions for probabilistic forecasting of day-ahead electricity prices: The benefits of using isotonic distributional regression