Paper: arXiv 2601.05085

Authors: Emma Hubert, Dimitrios Lolas, Ronnie Sircar

Abstract

We study the problem of forecasting and optimally trading day-ahead versus real-time (DART) price spreads in U.S. wholesale electricity markets. Building on the framework of Galarneau-Vincent et al., we extend spike prediction from a single zone to a multi-zone setting and treat both positive and negative DART spikes within a unified statistical model. To translate directional signals into economically meaningful positions, we develop a structural and market-consistent price impact model based on day-ahead bid stacks. This yields closed-form expressions for the optimal vector of zonal INC/DEC quantities, capturing asymmetric buy/sell impacts and cross-zone congestion effects. When applied to NYISO, the resulting impact-aware strategy significantly improves the risk-return profile relative to unit-size trading and highlights substantial heterogeneity across markets and seasons.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced statistical modeling (multi-zone spike forecasting, structural price impact models) and derives closed-form optimal position solutions, indicating high math complexity. It is heavily data-driven with out-of-sample validation across multiple ISO markets (NYISO, ISO-NE, ERCOT) over several years, demonstrating high empirical rigor and backtest readiness.

Research Flowchart

  flowchart TD
  A["Research Goal: Forecast & trade DART price spreads<br>Optimize zonal positions in multi-zone markets"] --> B["Data Inputs: NYISO day-ahead & real-time prices<br>Market bid stacks, zonal correlations"]
  B --> C["Methodology: Unified statistical model<br>Positive/Negative spike prediction across zones"]
  C --> D["Computational Process: Structural price impact model<br>Closed-form optimal INC/DEC quantities"]
  D --> E["Execution: Impact-aware trading strategy<br>Day-ahead commitment vs Real-time adjustment"]
  E --> F["Key Findings: Significantly improved risk-return profile<br>vs unit-size trading; Market/seasonal heterogeneity"]

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 a multi‑zone DART spike forecasting model combined with a calibrated structural price‑impact model that yields closed‑form optimal virtual trade sizes.

Source sentence
we extend spike prediction from a single zone to a multi‑zone setting and treat both positive and negative DART spikes within a unified statistical model.

Assumes 6

  • Price impact of virtual trades is linear in the aggregate system‑wide quantity and zone‑specific quantity. model
    Source sentence
    we approximate the zonal response of prices to aggregate traded quantity St by a piecewise linear function, with energy‑impact coefficients k+E and k‑E, and a separate linear impact from the trade size qt,z in the specific zone with coefficient kz
  • All predictors are constructed using only information available strictly before the day‑ahead bid deadline. data
    Source sentence
    As day‑ahead bids must be submitted well in advance of the operating day, all predictors are constructed using only data available strictly before th…
  • Spike events are extremely rare and the feature set is largely linear in its predictive structure. statistical
    Source sentence
    the spike events of interest are extremely rare and the feature set is largely linear in its predictive structure; consequently, more flexible models exhibit a tendency to overfit the noise
  • Logistic regression provides stable out‑of‑sample probability forecasts that can be reliably translated into trading signals. model
    Source sentence
    Logistic regression, by contrast, delivered the most robust and interpretable out‑of‑sample performance across all three markets, and produced stable probability estimates that translate reliably into trading signals
Show 2 more
  • Cross‑zone DART movements are synchronized in ISO‑NE and ERCOT, allowing a single representative zone to capture most variation. market
    Source sentence
    ISO‑NE and ERCOT exhibit highly synchronized zonal DART movements, so that a single representative zone captures most of the relevant variation
  • Higher forecasted load increases losses and congestion in expectation, justifying the linear‑quadratic impact model. model
    Source sentence
    Both components are consistently positively correlated with forecast load across all seasons and time buckets, supporting the modeling assumption that higher forecasted load increases losses and congestion in expectation

Data

  • NYISO, ISO‑NE, ERCOT market data · public data portals of the corresponding system operators market-prices public
    NYISO: 2015–2019; ISO‑NE and ERCOT: 2018 · hourly panel of day‑ahead and real‑time prices, system and zonal load forecasts, exogenous
    Source sentence
    All market data were obtained from the public data portals of the corresponding system operators, namely ERCOT [11], ISO New England [21], and NYISO [29].
  • GridStatus platform · GridStatus (https://www.gridstatus.io) other public
    Source sentence
    Data access and aggregation were facilitated using the GridStatus platform (https://www.gridstatus.io).

Costs

Modeled Costs are part of the reported headline results.

A linear‑quadratic price‑impact model is calibrated from historical day‑ahead bid stacks and incorporated into P&L calculations.

Source sentence
we estimate system‑wide energy impact coefficients and zone‑specific congestion sensitivities, yielding a linear–quadratic impact model for virtual load

Finds 5

  • The impact‑aware strategy yields realized price‑impact slopes between roughly $8 and $41 per MWh per 1,000 MWh, matching calibrated coefficients. performance
    Source sentence
    The results in Table 17 show that the realized slopes lie between roughly 8 and 41$/MWh per 1000 MWh. These magnitudes are fully consistent with the calibrated Summer–Peak coefficients k+E ∈[34.64, 46.48]/1000MWh
  • Logistic regression outperforms more flexible models in out‑of‑sample trading P&L despite similar in‑sample accuracy. predictive-accuracy
    Source sentence
    While some nonlinear models achieved marginally higher in‑sample classification accuracy, they offered no consistent improvement in out‑of‑sample trading P&L
  • NYISO exhibits heterogeneous zone‑level DART distributions, making multi‑zone modeling essential; Long Island is the most profitable zone. statistical
    Source sentence
    Long Island emerges as the most profitable load pocket, combining frequent extreme spreads with high average load
  • Restricting trades to the model‑consistent direction (side‑clipping) improves total P&L and reduces contradictory positions. performance
    Source sentence
    Figure 13 shows that this clipped strategy performs better on the 2022–2025 test set. The total P&L increases relative to the unconstrained optimizer
  • Cross‑zone interactions via the common energy component can make offsetting positions optimal to concentrate risk where marginal price impact is lowest. theoretical
    Source sentence
    Closed‑form expressions for the optimal zonal quantities show that portfolio‑level decisions are shaped not only by local expected revenues, but also by cross‑zone interactions through the common energy component

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 36 related papers). Coverage is partial, so this is neither replication nor refutation.

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.

Cited References (47)

  1. Spillovers Into the German Electricity Market From the Gas, Coal, and CO2 Emissions Markets (2025) — link
  2. Financial Arbitrage and Efficient Dispatch in Wholesale Electricity Markets (2015)
  3. Electricity pricing using a periodic GARCH model with conditional skewness and kurtosis components (2021) — link
  4. Foreseeing the Worst: Forecasting Electricity DART Spikes (2023) — link
  5. An optimal trading problem in intraday electricity markets (2016)
  6. Virtual Trading (2025) — link
  7. Virtual Bid Financial Assurance Methodology (2008) — link
  8. Locational Based Marginal Pricing (2025) — link
and 39 more
  1. FAQ s: Locational Marginal Pricing (2025) — link
  2. NYISO Energy Marketplace: E-Learning Module (2021) — link
  3. Day-Ahead Energy Markets (NEWEM) — link
  4. DAM Overview and Timeline — link
  5. Congestion Price Component (2025) — link
  6. Locational Minimum Installed Capacity Requirements Study for the 2024–2025 Capability Year (2024) — link
  7. Qualified Scheduling Entities (2025)
  8. Registration and Qualification (2025)
  9. Optimal execution of portfolio transactions (2001)
  10. No-dynamic-arbitrage and market impact (2010)
  11. Optimal trading strategy and supply/demand dynamics (2013)
  12. The price impact of order book events (2014)
  13. Price impact (2010)
  14. Electricity forward prices: A high-frequency empirical analysis (2004)
  15. It never rains but it pours: modeling the persistence of spikes in electricity prices (2009)
  16. Modeling and forecasting electricity spot prices: A functional data perspective (2013)
  17. Forecasting spikes in electricity prices (2012)
  18. An empirical comparison of alternative schemes for combining electricity spot price forecasts (2014) — link
  19. Forecasting day-ahead electricity prices: A review of state-of-the-art algorithms, best practices and an open-access benchmark (2021) — link
  20. Forecasting day-ahead price spikes for the Ontario electricity market (2016) — link
  21. Measuring Market Inefficiencies in California ’s Restructured Wholesale Electricity Market (2002) — link
  22. The Elements of Statistical Learning: Data Mining, Inference, and Prediction (2009)
  23. Divergence measures based on the S hannon entropy (1991)
  24. ERCOT Market Data (2025)
  25. ISO N ew E ngland Markets and Operations Data (2025)
  26. NYISO Market Data (2025)
  27. GridStatus.io
  28. Deep Learning-Based Electricity Price Forecast for Virtual Bidding in Wholesale Electricity Market — → 2412.00062
  29. Electricity intraday price modelling with marked H awkes processes (2022)
  30. Modelling interregional links in electricity price spikes (2015)
  31. Self-exciting jump processes with applications to energy markets (2018)
  32. The modeling and forecasting of extreme events in electricity spot markets (2014)
  33. Estimating fast mean-reverting jumps in electricity market models (2020)
  34. Hedging local volume risk using forward markets: Nordic case (2017)
  35. Learning in S tackelberg Games with Application to Strategic Bidding in the Electricity Market (2024)
  36. Stochastic behaviour of the electricity bid stack: From fundamental drivers to power prices (2009)
  37. A model for hedging load and price risk in the T exas electricity market (2013) — link
  38. Distributional forecasting of electricity DART spreads with a covariate-dependent mixture model (2025) — link
  39. Virtual Bidding and Electricity Market Design (2016) — link

1 of 47 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.