{"paper_id":"2511.02016","title":"ABIDES-MARL: A Multi-Agent Reinforcement Learning Environment for Endogenous Price Formation and Execution in a Limit Order Book","authors":["Patrick Cheridito","Jean-Loup Dupret","Zhexin Wu"],"paper_date":"2025-11-03","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2511.02016v1","page_url":"https://thequant.space/flowcharts/abides-marl-a-multi-agent-reinforcement-learning-environmen/","scores":{"empirical_rigor":6,"math_complexity":7,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes ABIDES-MARL, a new multi‑agent reinforcement learning framework combined with a realistic limit‑order‑book simulation to study equilibrium behavior in financial market games.","quote":"We present ABIDES-MARL, a framework that combines a new multi-agent reinforcement learning (MARL) methodology with a new realistic limit-order-book (LOB) simulation system to study equilibrium behavior in complex financial market games.","summary":"We present ABIDES-MARL, a framework that combines a new multi-agent reinforcement learning (MARL) methodology with a new realistic limit-order-book (LOB) simulation system to study equilibrium behavior in complex financial market games. The system extends ABIDES-Gym by decoupling state collection…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Perfect competition among market makers in the classical Kyle model.","type":"model","stated":"explicit","quote":"A central assumption of the Kyle model is perfect competition among market makers."},{"assumption":"Finite set of competing market makers replaces the infinite‑maker assumption.","type":"model","stated":"explicit","quote":"First, we replace the assumption of infinitely many market makers with a finite set of competing market makers."},{"assumption":"The market clears at each step, meaning all orders are absorbed by market makers.","type":"market","stated":"explicit","quote":"We assume that the market clears at each step, meaning all orders are absorbed by market makers."},{"assumption":"Market impact λ(n) can vary over time, reflecting dynamic microstructure.","type":"execution","stated":"explicit","quote":"Our formulation allows λ(n) to vary over time, consistent with the"},{"assumption":"Risk aversion is modeled through a running penalty on inventory.","type":"model","stated":"explicit","quote":"Risk aversion is modeled through a running penalty on inventory."},{"assumption":"Agents observe either the full LOB (exchange) or a hidden LOB with VWAP inference (OTC).","type":"market","stated":"explicit","quote":"State representations. We distinguish between two scenarios. The first, denoted as exchange, assumes that all market participants can observe the full limit order book (LOB). The second, denoted as OTC, assumes the LOB is hidden; participants must instead infer its structure through the volume-weighted average price (V"}]},"datasets":{"status":"available","items":[{"name":"ABIDES-MARL simulation","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"We validate the approach by solving an extended Kyle model within the simulation system, recovering the gradual price discovery phenomenon.","known_id":"abides"}],"known":[{"id":"abides","name":"ABIDES simulator","provider":"ABIDES","kind":"synthetic","access":"public","quote":"To highlight the ﬂexibility of ABIDES-MARL for strategy development, ﬁve separate exe-cution policies are evaluated while the informed trader and market makers use the PPO policies obtained from multi-agent training: (i) PPO (multi-agent).","mentions":44,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":null,"detail":"Market impact λ(n) is used as the cost model, varying over time and depth‑weighted across market makers.","quote":"Our formulation allows λ(n) to vary over time, consistent with the","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Market impact λ(n) is used as the cost model, varying over time and depth‑weighted across market makers.","card_quote":"Our formulation allows λ(n) to vary over time, consistent with the","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"unclear","findings":[{"finding":"The ABIDES-MARL framework recovers the gradual price discovery phenomenon when solving the extended Kyle model.","kind":"theoretical","quote":"We validate the approach by solving an extended Kyle model within the simulation system, recovering the gradual price discovery phenomenon."},{"finding":"Execution strategies shape market‑maker behavior and price dynamics at equilibrium.","kind":"theoretical","quote":"show that, at equilibrium, execution strategies shape market-maker behavior and price dynamics."},{"finding":"The multi‑agent PPO policy underperformed because counterparties adapted adversarially during training.","kind":"performance","quote":"The multi-agent PPO policy underperformed because counterparties had adapted adversarially during training, highlighting that execution costs in adaptive markets are inherently endogenous to the learning dynamics of competitors."},{"finding":"Across strategies, the time‑weighted average price (TWAP) rule remained the most robust under linear market dynamics.","kind":"performance","quote":"Across strategies, the time-weighted average price (TWAP) rule remained the most robust under linear market dynamics"},{"finding":"A retrained single‑agent PPO policy outperformed others under non‑linear settings.","kind":"performance","quote":"a retrained single-agent PPO policy outperformed others under non‑linear settings"}],"reported_performance":null,"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":42,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"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":"2026-11-14T17:23:04Z"},"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/"}