{"paper_id":"2412.12199","title":"Stochastic Gradient Descent in the Optimal Control of Execution Costs","authors":["Simeon Kolev"],"paper_date":"2024-12-14","paper_type":"methodological","paper_url":"https://arxiv.org/pdf/2412.12199v1","page_url":"https://thequant.space/flowcharts/stochastic-gradient-descent-in-the-optimal-control-of-execut/","scores":{"empirical_rigor":6,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes using stochastic gradient descent (SGD) to derive alternative execution cost policies in dynamic markets where explicit optimal solutions may not exist.","quote":"This study aims to bridge this gap by proposing an approach that leverages stochastic gradient descent (SGD) to derive alternative solutions for optimizing execution cost policies in dynamic markets where explicit mathematical solutions may not yet exist.","summary":"Bertsimas and Lo’s seminal work laid the groundwork for addressing the implementation shortfall dilemma in institutional investing, emphasizing the significance of market microstructure and price dynamics in minimizing execution costs. However, the ability to derive a theoretical Optimum market…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The existence of a mathematically derived optimal solution that is a function of the underlying market dynamics is assumed.","type":"model","stated":"explicit","quote":"The proposed methodology assumes the existence of a mathematically derived optimal solution that is a function of the underlying market dynamics.","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":"available","modeled":"yes","headline_net_of_costs":null,"detail":null,"quote":"After properly tuning the SGD algorithms, I will compare each model’s predicted optimum purchasing order size, the total accumulated execution cost, and the per period variance in the SGD’s strategy vs the Optimum strategy.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":null,"card_quote":"After properly tuning the SGD algorithms, I will compare each model’s predicted optimum purchasing order size, the total accumulated execution cost, and the per period variance in the SGD’s strategy vs the Optimum strategy.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"unclear","findings":[{"finding":"SGD strategies displayed significantly increased variance in order size (Bt) but produced relatively similar total accumulated execution costs compared to the optimal strategy.","kind":"performance","quote":"Though the SGD strategies displayed a significant increased variance in Bt, the algorithms still produced relatively similar total accumulated execution costs.","support":"full"},{"finding":"The SGD-derived optimal order sizes and total execution costs are negligibly suboptimal relative to the true mathematical optimum.","kind":"performance","quote":"Comparative analysis shows that the unique mathematical optimal is the true minimum execution cost, however, the SGD derived optimal Bt and total execution costs are negligibly suboptimal.","support":"full"},{"finding":"None of the SGD algorithms were able to beat the optimum strategy, yet they converged to a relative optimum in practice.","kind":"performance","quote":"14 Conclusion In conclusion, even though none of the SGD algorithms were not able to beat the Optimum strategy derived by Bertsimas and Lo in the “Optimal Control of Execution Costs” (1998), the algorithms were able to derive alternative strategies that achiever relative converge to the true optimum solution, in practice.","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":23,"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":2,"extracted_at":"2026-11-14T19:53:20Z"},"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/"}