From electricity prices to profits: multidimensional probabilistic forecasting for BESS trading

Battle Count: 76

Highly relevant to quantitative trading in electricity markets. The paper directly addresses probabilistic forecasting for trading decisions (BESS arbitrage), risk management (VaR, stopping rules), and economic evaluation of forecast quality. The methodology of constructing multidimensional ensemble forecasts and translating them into profit distributions is directly applicable to algorithmic trading strategies. The focus on decision-relevant quantities (profits) rather than just price accuracy is particularly valuable for practitioners.

Moderate complexity. The ARX model is straightforward to implement. The MS method requires careful coordination of random splitting across different estimation windows (the novel algorithm in Section 3.5.1). The forecast averaging procedure and the multidimensional ensemble construction add complexity. The BESS trading strategy with stopping rules is relatively simple. However, the overall pipeline (rolling window estimation, multiple splits, ensemble construction, profit forecasting, trading decisions) requires careful implementation. No code is provided.

The paper provides detailed algorithm descriptions, model specifications (Eq. 1), and evaluation procedures. Data sources are clearly identified (transparency.entsoe.eu for electricity data, Investing.com for commodity prices). The MS method is referenced to Maciejowska & Nitka (2026). However, no code repository is explicitly provided, and some implementation details (e.g., exact number of splits N, specific window sizes) are described but not fully parameterized for direct replication.