{"paper_id":"2409.12453","title":"Theoretical and Empirical Validation of Heston Model","authors":["Zheng Cao","Xinhao Lin"],"paper_date":"2024-09-19","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2409.12453v2","page_url":"https://thequant.space/flowcharts/theoretical-and-empirical-validation-of-heston-model/","scores":{"empirical_rigor":6,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes applying the Heston model to WTI crude options and using machine‑learning (gradient descent) to calibrate its five parameters, including fixing two parameters to improve fit.","quote":"Machine learning models such as Gradient Descent are utilized to help optimize the loss function and learn the optimal parameters on select dates.","summary":"This study focuses on the application of the Heston model to option pricing, employing both theoretical derivations and empirical validations. The Heston model, known for its ability to incorporate stochastic volatility, is derived and analyzed to evaluate its effectiveness in pricing options. For…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The collected WTI data are treated as European‑style contracts for simplicity.","type":"model","stated":"explicit","quote":"In addition, we treat the collected WTI data like European-style contracts for simplicity."},{"assumption":"The number of days in a year is set to T = 365.","type":"model","stated":"explicit","quote":"we first determine the values of number of days of a year T = 365"},{"assumption":"The risk‑free interest rate is assumed constant at r = 0.036.","type":"market","stated":"explicit","quote":"and the risk free interest rate r = 0.036."},{"assumption":"Initial λ is fixed at 0.45 and does not change much during calibration.","type":"model","stated":"explicit","quote":"Since we used the same λ value (which is 0.45) as initial parameters, the output λ does not change much."},{"assumption":"Parameters η and ρ are assumed constant in the Heston model (a limitation noted by the authors).","type":"model","stated":"implied","quote":"constant correlation between underlying price and volatility"},{"assumption":"Data are assumed to be accurate except for occasional zero‑IV records, which are treated as errors.","type":"data","stated":"implied","quote":"Some records from Barchart display an implied volatility (IV) of 0.00%, which likely indicates a system error."}]},"datasets":{"status":"available","items":[{"name":"WTI futures and options data","provider":"Barchart.com","kind":"options","access":"public","period":"04‑24‑2024 to 05‑10‑2024 (six days)","universe":"Crude Oil WTI futures and option contracts","quote":"We obtain the WTI futures and options data from Barchart.com."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"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":"Calibration loss on 04‑26‑2024 is 3.1686 × 10⁻⁴, while on 05‑10‑2024 it is 9.5305 × 10⁻³, indicating poorer fit on the latter date.","kind":"pricing-calibration","quote":"fitting loss of 04-26 is 3.1686 × 10−4 and that of 05-10 is 9.5305 × 10−3."},{"finding":"Parameters v0, η, and ρ do not remain constant across options with different maturities or on different days, contradicting the Heston model’s constant‑parameter assumption.","kind":"statistical","quote":"Our results show that neither v0, η, or ρ remains constant among options with different time to maturities or on different days."},{"finding":"Fixing two parameters (λ = 0.45 and ¯v = 0.0763) and calibrating the remaining three improves the fit for multiple contracts and dates.","kind":"performance","quote":"Since the effect of ¯v and λ on IV curves is much smaller than and can be offset by v0 and η, we now fix ¯v and λ constant and regard the remaining 3 parameters as variables."},{"finding":"The effect of λ on IV curve curvature is much smaller than that of other parameters, and λ remains relatively stable during calibration.","kind":"theoretical","quote":"the effect is much smaller than that of λ. Since we used the same λ value (which is 0.45) as initial parameters, the output λ does not change much."},{"finding":"The volatility‑of‑volatility (η) and correlation (ρ) parameters are unstable across days, suggesting the Heston model’s assumption of constant η and ρ is a limitation.","kind":"negative","quote":"the parameter ρ is unstable and does not have any pattern among any option contract on different days."}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":19,"citations":{"cited_by_in_corpus":2,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":2},"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":"2026-11-14T18:01:55Z"},"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/"}