{"paper_id":"2311.06476","title":"Relative entropy-regularized robust optimal order execution","authors":["Meng Wang","Tai-Ho Wang"],"paper_date":"2023-11-11","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2311.06476v2","page_url":"https://thequant.space/flowcharts/relative-entropy-regularized-robust-optimal-order-execution/","scores":{"empirical_rigor":3,"math_complexity":9,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a relative entropy‑regularized robust optimal control framework for order execution, including a market impact model that incorporates a permanent impact component from the averaged market trading rate.","quote":"The problem of order execution is cast as a relative entropy-regularized robust optimal control problem in this article. To take into account the liquidity and uncertainty risk from market trading rate when executing orders, we propose a price impact model as follows.","summary":"The problem of order execution is cast as a relative entropy-regularized robust optimal control problem in this article. The order execution agent’s goal is to maximize an objective functional associated with his profit-and-loss of trading and simultaneously minimize the execution risk and the…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The prior distribution of the market trading rate is Gaussian with mean mt and precision st.","type":"model","stated":"explicit","quote":"The “prior” distribution π0 t is Gaussian with mean mt and precision st.","support":"full"},{"assumption":"The execution problem is linear‑quadratic with Gaussian prior, leading to Riccati and linear ODEs.","type":"model","stated":"explicit","quote":"Under the assumptions of linear-quadratic model with Gaussian prior, the rHJI equation reduces to a system of Riccati and linear differential equations. The following assumptions are imposed in order for the problem (3.2) to remain in a Gaussian and LQ structure though other choices are by all means possible. A1.","support":"full"},{"assumption":"The coefficient Raa satisfies st + βRaa > 0 for all t.","type":"model","stated":"explicit","quote":"Moreover, the coefficient Raa is further assumed satisfying st +βRaa > 0 for all t, where recall that st is the precision for the prior Gaussian distribution π0 t given in Assumption A1.","support":"full"},{"assumption":"The market liquidity and uncertainty risk are captured by the average market trading rate ⟨a⟩, factored out of the diffusion term.","type":"market","stated":"explicit","quote":"In this paper, we factor it out from the diffusion term so as to assess the market’s liquidity and uncertainty risk in order execution.","support":"full"},{"assumption":"Coefficients are constant over time to obtain closed‑form solutions.","type":"model","stated":"explicit","quote":"Further imposing constancy of the corresponding coefficients, the system of differential equations can be solved in closed form, resulting in analytical expressions for optimal strategy and trajectory as well as the posterior distribution of market volume. Further assuming time independent coefficients, the solution to the aforementioned system of ODEs can be obtained in closed form, yield-ing closed form","support":"full"},{"assumption":"The efficient price is a martingale when buy and sell orders even out on average.","type":"market","stated":"explicit","quote":"Also if buy and sell orders even out in average, apart from the possible permanent impact incurred from the agent’s trading, the efficient price is a martingale.","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":"Price impact model includes permanent impact from market trading rate and stochastic components; costs are embedded in the objective via the R term.","quote":"To take into account the liquidity and uncertainty risk from market trading rate when executing orders, we propose a price impact model as follows. The efficient price St is assumed governed by the stochastic differential equation 4 M. WANG AND T.-H. WANG (SDE) dSt = γdXt + γM⟨a⟩tdt + σS t dW S t , (2.1) where ⟨·⟩t denotes averaging over πt, i.e., ⟨a⟩t = R aπt(a)da and σS t = qR ˜σ2(a)πt(a)da for some ˜σ.","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Price impact model includes permanent impact from market trading rate and stochastic components; costs are embedded in the objective via the R term.","card_quote":"To take into account the liquidity and uncertainty risk from market trading rate when executing orders, we propose a price impact model as follows. The efficient price St is assumed governed by the stochastic differential equation 4 M. WANG AND T.-H. WANG (SDE) dSt = γdXt + γM⟨a⟩tdt + σS t dW S t , (2.1) where ⟨·⟩t denotes averaging over πt, i.e., ⟨a⟩t = R aπt(a)da and σS t = qR ˜σ2(a)πt(a)da for some ˜σ.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Under constant coefficients the rHJI equation reduces to closed‑form analytical expressions for the optimal trading strategy and posterior distribution of market volume.","kind":"theoretical","quote":"Further imposing constancy of the corresponding coefficients, the system of differential equations can be solved in closed form, resulting in analytical expressions for optimal strategy and trajectory as well as the posterior distribution of market volume. Further assuming time independent coefficients, the solution to the aforementioned system of ODEs can be obtained in closed form, yield-ing closed form","support":"full"},{"finding":"Numerical examples show that the optimal strategies outperform conventional TWAP (and VWAP) strategies, achieving lower variance and better performance.","kind":"performance","quote":"Heuristics can be that, as mentioned earlier that TWAP implementation does not take dynamic of market environment into account, when exposed to extreme market conditions, it performs much more wildly compared to the optimal. Overall, the optimal strategies outperform, in some cases much better, and incurring lower variance as opposed to their TWAP counterparties as shown in the figures. 8.","support":"full"},{"finding":"The proposed model yields a robust control solution that accounts for execution risk, market liquidity, and uncertainty risk.","kind":"theoretical","quote":"The order execution agent's goal is to maximize an objective functional associated with his profit-and-loss of trading and simultaneously minimize the execution risk and the market's liquidity and uncertainty. In addition to execution risk, the model takes into account market’s liquidity and uncertainty risk.","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":20,"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":"2026-11-14T20:00:23Z"},"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/"}