{"paper_id":"2311.11913","title":"Deep Calibration of Market Simulations using Neural Density Estimators and Embedding Networks","authors":["Namid R. Stillman","Rory Baggott","Justin Lyon","Jianfei Zhang","Dingqiu Zhu","Tao Chen","Perukrishnen Vytelingum"],"paper_date":"2023-11-20","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2311.11913v2","page_url":"https://thequant.space/flowcharts/deep-calibration-of-market-simulations-using-neural-density/","scores":{"empirical_rigor":6,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a novel calibration approach for market simulators using neural density estimators and embedding networks.","quote":"we develop a novel approach to the calibration of market simulators by leveraging recent advances in deep learning, specifically using neural density estimators and embedding networks.","summary":"The ability to construct a realistic simulator of financial exchanges, including reproducing the dynamics of the limit order book, can give insight into many counterfactual scenarios, such as a flash crash, a margin call, or changes in macroeconomic outlook. In recent years, agent-based models have…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Uniform prior distribution over model parameters.","type":"model","stated":"explicit","quote":"We assume a uniform prior distribution, with lower bound of [2, 1, −4, −2] and upper bound of [3, 2, −2, 0] in log-space."},{"assumption":"Uniform prior distribution for the extended Chiarella model.","type":"model","stated":"explicit","quote":"We again assume a uniform prior distribution, with lower bound [−2, −1, −1, −1, 4, −2] and upper bound [0, 1, 1, 0, 6, 0] in log-space."},{"assumption":"The posterior is valid only if the prior is correctly constrained and the model can describe the data generation process.","type":"statistical","stated":"explicit","quote":"the posterior is true under the assumption of a correctly constrained prior and that the model is able to describe the underlying data generation process."},{"assumption":"Market observations are accurate (if partial).","type":"data","stated":"explicit","quote":"we assume that our observation of the market is accurate (if partial)"},{"assumption":"Fixed simulation budget of 10^4 simulations.","type":"other","stated":"explicit","quote":"we sample from a model-specific prior distribution with a fixed simulation budget of 104 simulations, where 1,000 are held back for validation and 1,000 for testing."}]},"datasets":{"status":"available","items":[{"name":"Synthetic data","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"both when applied to synthetic and historical data"},{"name":"Historical futures data on HKEX","provider":"HKEX","kind":"market-prices","access":"unknown","period":"one day","universe":"futures contract","quote":"For both models, we compare our results to historical data from one days trading of a futures contract on the HKEX exchange."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"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":"pricing-calibration","findings":[{"finding":"The ZI trader model parameters are recovered with RMSE 1.85 ± 0.52 using mid‑price and total volume.","kind":"performance","quote":"we are able to identify the four parameters with RMSE of 1.85 (+/- 0.52), when using the mid-price and total volume"},{"finding":"Using VWAP data improves ZI parameter recovery to RMSE 1.21 ± 0.48.","kind":"performance","quote":"RMSE of 1.21 (+/-0.48) when using the VWAP at the fist level of the LOB"},{"finding":"Extended Chiarella model parameters are calibrated with RMSE 0.77 ± 0.26 using mid‑price and total volume.","kind":"performance","quote":"we are able to calibrate the 6 parameters with a RMSE of 0.77 (+/- 0.26) when using the midprice and total volume data"},{"finding":"Using VWAP for the extended Chiarella model yields RMSE 0.82 ± 0.29.","kind":"performance","quote":"RMSE of 0.82 (+/- 0.29) when using the VWAP at the first level"},{"finding":"The simulator reproduces most stylised facts, but volatility clustering is weaker than in historical data.","kind":"statistical","quote":"The only stylised fact that is observed less strongly in our simulator is the volatility clustering, shown in Figure 2(f), which decreases with increasing lag but is not consistently positive."}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[],"contradicts":[],"qualifies":[],"compared_with":42,"citations":{"cited_by_in_corpus":3,"by_independent_authors":2,"how_cited":{"used_as_baseline":0,"uses_method_or_data":1,"extends":0,"discusses_findings":0,"mentions":1,"unclassified":1},"independent_papers":[{"paper":{"paper_id":"2507.09863","title":"Towards Realistic and Interpretable Market Simulations: Factorizing Financial Power Law using Optimal Transport","paper_date":"2025-07-14","authors":["Ryuji Hashimoto","Kiyoshi Izumi"],"empirical_rigor":6,"page_url":"https://thequant.space/flowcharts/towards-realistic-and-interpretable-market-simulations-fact/"},"how":"uses_method_or_data","quote":null}],"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":"2026-11-14T18:19:17Z"},"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/"}