{"paper_id":"2412.15959","title":"Battery valuation on electricity intraday markets with liquidity costs","authors":["Enzo Cognéville","Thomas Deschatre","Xavier Warin"],"paper_date":"2024-12-20","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2412.15959v1","page_url":"https://thequant.space/flowcharts/battery-valuation-on-electricity-intraday-markets-with-liqui/","scores":{"empirical_rigor":8.5,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a complete modelling framework to value several batteries in the electricity intraday market at the trading session scale.","quote":"we propose a complete modelling framework to value several batteries in the electricity intraday market at the trading session scale.","summary":"In this paper, we propose a complete modelling framework to value several batteries in the electricity intraday market at the trading session scale. The model consists of a stochastic model for the 24 mid-prices (one price per delivery hour) combined with a deterministic model for the liquidity…","strategy":"stochastic optimisation with liquidity cost model"},"assumptions":{"status":"available","items":[{"assumption":"Battery efficiency is set to ρ = 0.92.","type":"model","stated":"explicit","quote":"The battery eﬃciency is assumed to be ρ = 0.92, so:"},{"assumption":"Decisions must be taken Δ between 1 and a few hours before the delivery hour.","type":"execution","stated":"explicit","quote":"the decision must be taken at the time H −∆, where ∆should be between 1 and a few hours, since the market is illiquid."},{"assumption":"Liquidity parameters are estimated using the last 28 days of data and updated weekly.","type":"model","stated":"explicit","quote":"we estimate the price model and the liquidity parameters using the last 28 days of data and parameters are updated every Monday of each week."},{"assumption":"Battery starts each day with zero inventory.","type":"model","stated":"explicit","quote":"The battery is managed assuming zero inventory at T1 = 0 hour each day."},{"assumption":"Injection and withdrawal capacities are 1 MWh per hour.","type":"model","stated":"explicit","quote":"the injection and withdrawal capacities are 1MWh per hour."},{"assumption":"Liquidity curve is modeled by an asymmetric additive form (2).","type":"model","stated":"explicit","quote":"We fit a model close to (2) for the liquidity curve, which is more appropriate than (1) and shows evidence of market illiquidity."}]},"datasets":{"status":"available","items":[{"name":"EPEX order book data","provider":"EPEX","kind":"order-book","access":"commercial","period":"2021-2023","universe":"French and German intraday electricity markets","quote":"The dataset used consists of all orders for each trading session during 2021, 2022 and 2023 on the French and German intraday electricity markets, provided by the market platform EPEX."}],"known":[],"sample":{"start":null,"end":null,"market":"German and French electricity intraday markets","asset_class":"other","frequency":"intraday"},"data_resolution":"l2-book"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":true,"detail":"Liquidity costs are modeled as a price impact function P(m,Δ,V)=f_m,Tm-Δ + [A_m,−(Tm-Δ)V−B_m,−(Tm-Δ)]1{V<0}+ [A_m,+(Tm-Δ)V+B_m,+(Tm-Δ)]1{V>0}.","quote":"the purchase of a volume V (positively counted if purchase) at date Tm −∆for delivery at date Tm, costs a price P per unit of volume: P(m, ∆, V ) =fm,Tm−∆+ [Am,−(Tm −∆)V −Bm,−(Tm −∆)]1{V <0}+ [Am,+(Tm −∆)V + Bm,+(Tm −∆)]1{V >0}.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Liquidity costs are modeled as a price impact function P(m,Δ,V)=f_m,Tm-Δ + [A_m,−(Tm-Δ)V−B_m,−(Tm-Δ)]1{V<0}+ [A_m,+(Tm-Δ)V+B_m,+(Tm-Δ)]1{V>0}.","card_quote":"the purchase of a volume V (positively counted if purchase) at date Tm −∆for delivery at date Tm, costs a price P per unit of volume: P(m, ∆, V ) =fm,Tm−∆+ [Am,−(Tm −∆)V −Bm,−(Tm −∆)]1{V <0}+ [Am,+(Tm −∆)V + Bm,+(Tm −∆)]1{V >0}.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"Including liquidity costs is essential, especially with many batteries, leading to much higher profits and avoiding large losses.","kind":"performance","quote":"We show that it is essential to take liquidity into account, especially when the number of batteries is large: it allows much higher profits and avoids high losses using our liquidity model."},{"finding":"Using a stochastic model for mid‑prices significantly improves battery valuation compared with a deterministic spot‑price forecast.","kind":"performance","quote":"The use of our stochastic model for the mid-price also significantly improves the results (compared to a deterministic framework where the mid-price forecast is the spot price)."},{"finding":"Backtest for 2021 shows the depth (liquidity) model yields higher profit per battery than the no‑depth model (e.g., 23 994 € vs 21 264 € for 10 batteries).","kind":"performance","quote":"Table 5. Backtest gain (euros per battery) for the optimisation of ˆN nh batteries for the year 2021"},{"finding":"In the French market, liquidity is low; even a single battery gains only a small profit from accounting for price impact.","kind":"performance","quote":"The liquidity in France is rather low and even with a single battery, there is a small gain when taking into account the price impact as shown on Table 5."},{"finding":"In the more liquid German market, the price‑impact model does not improve back‑testing profits for a single battery and adds little for up to 20 batteries.","kind":"performance","quote":"the price impact model does not improve backtesting profits. Up to 20 batteries, the price impact model does not bring much more profit."}],"reported_performance":{"strategy":"stochastic optimisation with liquidity cost model","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":15,"citations":{"cited_by_in_corpus":1,"by_independent_authors":1,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":1,"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":"2026-11-14T17:49:24Z"},"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/"}