{"paper_id":"2407.16566","title":"Alleviating Non-identifiability: a High-fidelity Calibration Objective for Financial Market Simulation with Multivariate Time Series Data","authors":["Chenkai Wang","Junji Ren","Peng Yang"],"paper_date":"2024-07-23","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2407.16566v7","page_url":"https://thequant.space/flowcharts/alleviating-non-identifiability-a-high-fidelity-calibration/","scores":{"empirical_rigor":4,"math_complexity":7.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a maximization-based aggregation function to create a new calibration objective that combines multiple time‑series features.","quote":"To implement this theoretical finding, a maximization-based aggregation function is proposed based on existing discrepancy metrics to form a new calibration objective function.","summary":"The non-identifiability issue has been frequently reported in social simulation works, where different parameters of an agent-based simulation model yield indistinguishable simulated time series data under certain discrepancy metrics. This issue largely undermines the simulation fidelity yet lacks…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The optimal parameter ω* lies within the defined parameter space ΩD,k.","type":"model","stated":"explicit","quote":"Without loss of generality, it is assumed that the optimal parameter ω* lies in the defined parameter space, i.e., ω*∈ΩD,k, where D( ˆXkT , M(ω*)) < ϵ."},{"assumption":"Selected features are not linearly correlated, ensuring each adds independent constraints.","type":"data","stated":"explicit","quote":"as long as the features are selected and not linearly correlated, they can contribute to alleviation"},{"assumption":"Synthetic time series are generated by the PGPS model with parameters randomly sampled within specified ranges.","type":"data","stated":"implied","quote":"The 10 synthetic time series are generated using the PGPS model by randomly sampling parameters within the ranges specified in table I"},{"assumption":"Each individual calibration task corresponds to a distinct univariate time series derived from a single feature.","type":"model","stated":"explicit","quote":"the task of calibrating to a multivariate time series is naturally divided into multiple individual tasks, each of which calibrates the model to one distinct univariate time series."},{"assumption":"Features are handcrafted (price, volume, bid/ask directions, order arrival time, etc.) and can be represented as 1‑dimensional series.","type":"data","stated":"explicit","quote":"this feature can be either price, volume, bid/ask directions, order arrival time, or any other handcrafted features"},{"assumption":"The non‑identifiable set shrinks exponentially as the number of independent features increases.","type":"other","stated":"explicit","quote":"it is rigorously analyzed that this probability can be reduced with more time series features in an exponential manner."}]},"datasets":{"status":"available","items":[{"name":"Synthetic PGPS datasets","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"The 10 synthetic time series are generated using the PGPS model by randomly sampling parameters within the ranges specified in table I"},{"name":"000001.sz (Shenzhen Stock Exchange)","provider":"Shenzhen Stock Exchange","kind":"market-prices","access":"public","period":"1200 time steps at a frequency of 3 seco","universe":null,"quote":"The real data is chosen as 000001.sz from Shenzhen Stock Exchange of China, consisting of 1200 time steps at a frequency of 3 seconds from 9:30 a.m. to 10:30 a.m. of a day in 2019"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"intraday-bars"},"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":"Incorporating multiple time‑series features exponentially alleviates non‑identifiability.","kind":"theoretical","quote":"This paper theoretically demonstrates that incorporating multiple time series data features during the model calibration phase can exponentially alleviate non-identifiability as the number of features increases."},{"finding":"Empirical studies show the proposed method significantly reduces the non‑identifiable set.","kind":"statistical","quote":"Extensive empirical studies have successfully shown that the proposed method can significantly alleviate the"},{"finding":"Calibrating with more features improves simulation fidelity; e.g., F3 reduces MSE by 37.9% versus F1 on real data.","kind":"performance","quote":"the simulated mid-price of F3 is more similar to the real observed data than F1 by an improvement of 37.9% in MSE"},{"finding":"The max‑W aggregation outperforms other aggregation and discrepancy measures, achieving the lowest rank (1.50).","kind":"performance","quote":"... Within the max group, max‑W achieves the lowest rank (1.50), significantly outperforming KL (p = 4.88 × 10−3), KS (p = 6.84×10−3), and MSE (p = 0.0527)"},{"finding":"Adding linearly dependent features does not reduce non‑identifiability, confirming theoretical predictions.","kind":"theoretical","quote":"adding linearly dependent features does not reduce non-identifiability, aligning with the theoretical findings"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":26,"citations":{"cited_by_in_corpus":1,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":1},"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":"2026-11-14T18:05:45Z"},"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/"}