{"paper_id":"2409.02025","title":"Logarithmic regret in the ergodic Avellaneda-Stoikov market making model","authors":["Jialun Cao","David Šiška","Lukasz Szpruch","Tanut Treetanthiploet"],"paper_date":"2024-09-03","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2409.02025v2","page_url":"https://thequant.space/flowcharts/logarithmic-regret-in-the-ergodic-avellaneda-stoikov-market/","scores":{"empirical_rigor":3.5,"math_complexity":9.2,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a regularised maximum‑likelihood learning algorithm for the price‑sensitivity parameter κ in the ergodic Avellaneda‑Stoikov market‑making model, achieving an expected regret upper bound of order ln²T.","quote":"We analyse the regret arising from learning the price sensitivity parameter $κ$ of liquidity takers in the ergodic version of the Avellaneda-Stoikov market making model. We show that a learning algorithm based on a maximum-likelihood estimator for the parameter achieves the regret upper bound of order $\\ln^2 T$ in expectation.","summary":"We analyse the regret arising from learning the price sensitivity parameter κ of liquidity takers in the ergodic version of the Avellaneda-Stoikov market making model. We show that a learning algorithm based on a maximum-likelihood estimator for the parameter achieves the regret upper bound of…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The price‑sensitivity parameter κ is positive (κ>0).","type":"model","stated":"explicit","quote":"Theorem 6. Assume κ± = κ > 0.","support":"full"},{"assumption":"The ergodic formulation admits a unique invariant distribution for the inventory and convergence to it is exponential.","type":"model","stated":"explicit","quote":"First, we prove existence of and convergence to an invariant measure in the ergodic Avellaneda–Stoikov market making model. While the well-posedness of the ergodic problem follows mostly from the analysis carried out in Gu´eant and Manziuk [26] the result on existence of and convergence to the invariant measure is new and relies on newly established explicit solution to the ergodic HJB corresponding to our problem.","support":"full"},{"assumption":"The ergodic constant γ(κ) is twice differentiable with respect to κ under misspecification.","type":"statistical","stated":"explicit","quote":"The first is the twice differentiability of the ergodic constant under the misspecified parameter in the Hamilton-Jacobi-Bellman (HJB) equation with respect to $κ$, which leads to a second--order performance gap.","support":"full"},{"assumption":"The regularised maximum‑likelihood estimator satisfies concentration inequalities for Bernoulli signals.","type":"statistical","stated":"explicit","quote":"The second is the learning rate of the regularised maximum-likelihood estimator which is obtained from concentration inequalities for Bernoulli signals.","support":"full"}]},"datasets":{"status":"none_found","items":[],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"The expected regret of the proposed algorithm is bounded by O(ln²T).","kind":"theoretical","quote":"We show that a learning algorithm based on a maximum-likelihood estimator for the parameter achieves the regret upper bound of order $\\ln^2 T$ in expectation. We will establish an upper bound on regret of order ln2 T arising from having to learn the key unknown parameter online (while executing a strategy) in the ergodic market making model.","support":"full"},{"finding":"Numerical experiments confirm the convergence of the algorithm and the robustness of the regret bound.","kind":"performance","quote":"Numerical experiments confirm the convergence and the robustness of the proposed algorithm. Numerical experiments confirm the convergence and the robustness of the proposed algorithm. 1.","support":"full"},{"finding":"The ergodic HJB equation admits explicit solutions and a unique optimal ergodic control.","kind":"theoretical","quote":"4. Conclusion In this paper, we introduced and analysed the ergodic formulation of the Avellaneda– Stoikov market making model. We established explicit solutions to the ergodic Hamilton– Jacobi–Bellman (HJB) equation and thus derived the optimal ergodic Markov controls.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":16,"citations":{"cited_by_in_corpus":3,"by_independent_authors":2,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":2,"unclassified":1},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/Galen-Cao/MM_parmater_learning","note":null},"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":2,"extracted_at":"2026-11-14T19:55:36Z"},"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/"}