Paper: arXiv 2308.15443
Authors: Weronika Nitka, Rafał Weron
Abstract
Probabilistic price forecasting has recently gained attention in power trading because decisions based on such predictions can yield significantly higher profits than those made with point forecasts alone. At the same time, methods are being developed to combine predictive distributions, since no model is perfect and averaging generally improves forecasting performance. In this article we address the question of whether using CRPS learning, a novel weighting technique minimizing the continuous ranked probability score (CRPS), leads to optimal decisions in day-ahead bidding. To this end, we conduct an empirical study using hourly day-ahead electricity prices from the German EPEX market. We find that increasing the diversity of an ensemble can have a positive impact on accuracy. At the same time, the higher computational cost of using CRPS learning compared to an equal-weighted aggregation of distributions is not offset by higher profits, despite significantly more accurate predictions.
Complexity vs Empirical Score
- Math Complexity: 7.0/10
- Empirical Rigor: 6.0/10 (capped from 8.5: transaction costs not modelled; see penalty flags)
- Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 8.5)
Why this score: The paper involves advanced statistical concepts like proper scoring rules (CRPS) and complex ensemble weighting techniques, indicating high math complexity. It also conducts a rigorous empirical study with six years of real market data, state-of-the-art neural network models, and direct profit calculations for a trading strategy, showing high empirical rigor.
Research Flowchart
flowchart TD A["Research Goal<br>Does CRPS-minimizing ensemble<br>lead to optimal day-ahead bidding?"] B["Data Source<br>Hourly German EPEX<br>day-ahead prices"] C["Methodology<br>Ensemble prob. forecasting<br>CRPS vs. Equal-weight aggregation"] D["Computational Process<br>Hourly bidding decisions<br>based on prob. forecasts"] E["Outcome 1<br>CRPS yields<br>more accurate predictions"] F["Outcome 2<br>Higher accuracy does NOT<br>translate to higher profits"] G["Key Finding<br>Computational cost of CRPS<br>not justified by profit increase"] A --> B B --> C C --> D D --> E D --> F E --> G F --> G
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes using CRPS learning, a weighting technique that minimizes the continuous ranked probability score, to combine probabilistic electricity price forecasts and assess its impact on day-ahead bidding decisions.
Source sentence
In this article we address the question of whether using CRPS learning, a novel weighting technique minimizing the continuous ranked probability score (CRPS), leads to optimal decisions in day-ahead bidding.
Assumes 5
- A standard short-term forecast horizon of 1 day, performed in a rolling window scheme. model
Source sentence
We assume a standard short-term forecast horizon of 1 day, performed in a rolling window scheme
- Forecasts of all 24 hourly prices on day d are calculated in the morning of day d‑1, before the day‑ahead market closes. execution
Source sentence
We assume that forecasts of all 24 hourly prices on day d are calculated at the same time in the morning of day d −1, i.e., before the day-ahead market for day d closes
- Model parameters are estimated using a calibration sample of the D most recent past observations. model
Source sentence
the model parameters are estimated using a calibration sample of D most recent past observations
- Battery charging and discharging efficiency is 90%. execution
Source sentence
We assume that the efficiency of charging as well as discharging the battery is 90%
Show 1 more
- Total usable battery capacity is 2 MWh and it starts half‑charged at the beginning of the simulation. execution
Source sentence
we assume that the total usable capacity of the battery is B = 2 MWh and that at the beginning of the simulation period the battery starts halfway charged (B = 1)
Data
- German EPEX day‑ahead electricity prices market-prices public
Source sentence
hourly day-ahead electricity prices from the German EPEX market spanning the period from 1 January 2015 to 31 December 2020
- Load and RES generation forecasts fundamentals public
Source sentence
prices as well as the day-ahead predictions of the loads and RES generation are publicly available from the ENTSO‑E Transparency platform
Costs
Not modeled
Finds 5
- Increasing the diversity of an ensemble positively impacts forecast accuracy. predictive-accuracy
Source sentence
We find that increasing the diversity of an ensemble can have a positive impact on accuracy.
- CRPS learning yields slightly better predictions on average, but does not lead to higher profits despite higher accuracy. negative
Source sentence
higher computational cost of using CRPS learning compared to an equal-weighted aggregation of distributions is not offset by higher profits, despite significantly more accurate predictions.
- The DDNN_JSU_CRPS_LEAR ensemble achieves the lowest CRPS score among all models and ensembles. predictive-accuracy
Source sentence
The DDNN_JSU_CRPS_LEAR ensemble achieves the lowest CRPS score.
- CRPS learning yields slightly better predictions on average compared to equal weighting. predictive-accuracy
Source sentence
Nevertheless, CRPS learning yields slightly better predictions on average.
- Introducing diversity to a pool of forecasts is highly beneficial for both forecast accuracy (CRPS) and trading profits in the German day‑ahead market. performance
Source sentence
introducing diversity to a pool of forecasts is highly beneficial, both in terms of forecast accuracy measured by the CRPS and profits from a trading strategy implemented in the German day-ahead power market
Checked by others
Supported by
- From On-chain to Macro: Assessing the Importance of Data Source Diversity in Cryptocurrency Market Forecasting
Compare the findings
This paper: Increasing the diversity of an ensemble positively impacts forecast accuracy.
That paper: Data source diversity significantly enhances forecasting model performance across time horizons.
Contradicted by
- Improving Algorithmic Trading for BESS on Intraday Electricity Markets through Distributional Price Forecasts
Compare the findings
This paper: CRPS learning yields slightly better predictions on average, but does not lead to higher profits despite higher accuracy.
That paper: Forecast errors and instability are strongly negatively correlated with daily profits, indicating predictive accuracy drives returns.
5 papers in the archive : Probabilistic Forecasting for Day-ahead Electricity Prices, … ; Isotonic Quantile Regression Averaging for uncertainty …
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.