{"paper_id":"2504.06932","title":"Maximizing Battery Storage Profits via High-Frequency Intraday Trading","authors":["David Schaurecker","David Wozabal","Nils Löhndorf","Thorsten Staake"],"paper_date":"2025-04-09","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2504.06932v3","page_url":"https://thequant.space/flowcharts/maximizing-battery-storage-profits-via-high-frequency-intrad/","scores":{"empirical_rigor":9,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes an automated high-frequency trading strategy for grid-scale battery storage using a dynamic programming formulation that is orders of magnitude faster than exact MILP.","quote":"This paper introduces and evaluates an automated high-frequency trading strategy for battery energy storage systems trading on the intraday market for power while explicitly considering the dynamics of the limit order book, market rules, and technical parameters.","summary":"Maximizing revenue for grid-scale battery energy storage systems in continuous intraday electricity markets requires strategies that are able to seize trading opportunities as soon as new information arrives. This paper introduces and evaluates an automated high-frequency trading strategy for…","strategy":"high-frequency rolling intrinsic policy"},"assumptions":{"status":"available","items":[{"assumption":"The continuous intraday market operates with a limit order book and price increments of 0.01 € and volume increments of 0.1 MWh.","type":"market","stated":"explicit","quote":"The exchange allows for price increments of 0.01 € and volume increments at 0.1 MWh, defining the minimum action size on the market as u = 0.1 MWh."},{"assumption":"Battery charging and discharging efficiencies are η+ = η− = 0.95 (round‑trip ≈0.9).","type":"model","stated":"explicit","quote":"We assume efficiency losses of η+ = η−= 0.95, i.e., a round trip efficiency of η+η−≈0.9 similar to Cole and Karmakar (2023);"},{"assumption":"Linear degradation cost of νdeg = 4 €/MWh (8 €/MWh round‑trip).","type":"model","stated":"explicit","quote":"We use linear degradation costs of νdeg = 4 €/MWh, i.e., a round‑trip degradation cost of 8 €/MWh for every MWh of electricity sold by the battery."},{"assumption":"Trading fee νtrade = 0.09 €/MWh reflects EPEX volume‑based fees.","type":"execution","stated":"explicit","quote":"Furthermore, we choose νtrade = 0.09 €/MWh, reflecting the current EPEX volume‑based trading fees per each matched order."},{"assumption":"A combined delay Δt (solve‑time + technical delay) may exceed 1 ms, limiting reaction to order‑book updates.","type":"execution","stated":"explicit","quote":"If the combined delay exceeds 1 ms (the tick duration of the EPEX Spot market)"},{"assumption":"The battery has an energy‑to‑power ratio of 1 h, with power limit ¯f = −f = 10 MW and energy limit ¯s = 10 MWh.","type":"model","stated":"explicit","quote":"Unless otherwise stated, for our case study, we use a BESS with an energy‑to‑power ratio (duration) of 1 hour, setting ¯f = −f = 10 MW and ¯s = 10 MWh."}]},"datasets":{"status":"available","items":[{"name":"2021 EPEX Spot continuous intraday order‑book data (Germany)","provider":"EPEX Spot","kind":"order-book","access":"unknown","period":"2021","universe":"German hourly intraday products","quote":"All results are generated using historical 2021 order book data from the EPEX Spot continuous intraday market for Germany, focusing on hourly products."}],"known":[],"sample":{"start":"2021","end":"2021","market":"German intraday electricity market","asset_class":"other","frequency":"intraday"},"data_resolution":"l2-book"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":true,"detail":"Variable cost ν = νtrade + νdeg = 4.09 €/MWh (νtrade = 0.09 €/MWh, νdeg = 4 €/MWh).","quote":"Furthermore, we choose νtrade = 0.09 €/MWh, reflecting the current EPEX volume‑based trading fees per each matched order. In total, the variable cost ν = νtrade + νdeg thus takes the value of ν = 4.09 €/MWh.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Variable cost ν = νtrade + νdeg = 4.09 €/MWh (νtrade = 0.09 €/MWh, νdeg = 4 €/MWh).","card_quote":"Furthermore, we choose νtrade = 0.09 €/MWh, reflecting the current EPEX volume‑based trading fees per each matched order. In total, the variable cost ν = νtrade + νdeg thus takes the value of ν = 4.09 €/MWh.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The high‑frequency policy earns 58% more revenue than re‑optimizing only once per hour.","kind":"performance","quote":"Our policy earns 58% more than when re‑optimizing only once every hour"},{"finding":"It earns 14% more than re‑optimizing once per minute, showing profits depend critically on trading speed.","kind":"performance","quote":"and 14% more than when re‑optimizing once per minute"},{"finding":"The dynamic‑programming formulation is 2–3 orders of magnitude faster than an exact MILP solution while preserving profitability.","kind":"performance","quote":"solved using a dynamic programming approximation that is two to three orders of magnitude faster than an exact mixed‑integer linear programming solution."},{"finding":"Training a parametric extension of the rolling intrinsic increases yearly revenue by 8.4% out‑of‑sample (≈€376k vs €347k).","kind":"performance","quote":"an increase of 8.4% in trading reward over the entire year 2021, accumulating €376k, compared to the €347k reward of the standard policy"},{"finding":"Solution speed improvement of up to three orders of magnitude is achieved while maintaining sufficient accuracy.","kind":"theoretical","quote":"Our method achieves a solution speed improvement of up to three orders of magnitude while maintaining a sufficient level of accuracy."}],"reported_performance":{"strategy":"high-frequency rolling intrinsic policy","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"walk-forward","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":28,"citations":{"cited_by_in_corpus":2,"by_independent_authors":2,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":2,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T17:41:30Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}