{"paper_id":"2308.14473","title":"Joint Calibration of Local Volatility Models with Stochastic Interest Rates using Semimartingale Optimal Transport","authors":["Benjamin Joseph","Gregoire Loeper","Jan Obloj"],"paper_date":"2023-08-28","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2308.14473v1","page_url":"https://thequant.space/flowcharts/joint-calibration-of-local-volatility-models-with-stochastic/","scores":{"empirical_rigor":7.5,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a non‑parametric method for joint exact calibration of a local volatility model and a correlated stochastic short‑rate model using semimartingale optimal transport.","quote":"We develop and implement a non-parametric method for joint exact calibration of a local volatility model and a correlated stochastic short rate model using semimartingale optimal transport.","summary":"We develop and implement a non-parametric method for joint exact calibration of a local volatility model and a correlated stochastic short rate model using semimartingale optimal transport. The method relies on the duality results established in Joseph, Loeper, and Obloj, 2023 and jointly…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"A classic parametric model is used as the initial reference model and calibrated as well as possible to market prices.","type":"model","stated":"explicit","quote":"We choose as a starting reference model a classic parametric model, calibrated as well as possible to market prices."},{"assumption":"The short‑rate dynamics follow a Hull‑White model with constant parameters a and σr > 0.","type":"model","stated":"explicit","quote":"Where a, σr(t) > 0 are constants and b(·) is a function of time, calibrated so that the dynamics of (rt)0⩽t⩽T match the market data"},{"assumption":"The correlation coefficient ρ(t, Zt, rt) is assumed to be almost constant and close to the reference model in the joint calibration.","type":"model","stated":"explicit","quote":"We expect the correlation coefficient ρS to be almost constant and close to the reference model, owing to the assumption on the correlation in (21)"},{"assumption":"The state variable Vt is known when calibrating the local‑stochastic volatility model.","type":"data","stated":"explicit","quote":"We assume that the state variable Vt is known, so the model characteristics that we want to calibrate are the volatility of the log‑stock and the correlation between the state variables."},{"assumption":"Market data are obtained from a Bloomberg terminal on a specific date.","type":"market","stated":"explicit","quote":"Data obtained from a Bloomberg terminal at the Saïd Business school in Oxford on 23/05/2022."},{"assumption":"The optimisation tolerances ε1 and ε2 are set to 1×10⁻³ and 1×10⁻⁸ respectively.","type":"other","stated":"explicit","quote":"ε1 1 × 10−3 Tolerance for the difference in model and observed IV; ε2 1 × 10−8 Tolerance in the policy iteration for φ"}]},"datasets":{"status":"available","items":[{"name":"European SPX options","provider":"Bloomberg","kind":"options","access":"commercial","period":null,"universe":null,"quote":"We demonstrate the performance of our approach on market data using European SPX options and European cap interest rate options.","known_id":"bloomberg"},{"name":"European cap interest rate options","provider":"Bloomberg","kind":"options","access":"commercial","period":null,"universe":null,"quote":"We demonstrate the performance of our approach on market data using European SPX options and European cap interest rate options.","known_id":"bloomberg"},{"name":"Simulated data (reference model)","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"We test the methods on simulated data. We use the generating model as the reference model for the short rate."}],"known":[{"id":"bloomberg","name":"Bloomberg","provider":"Bloomberg L.P.","kind":"market-prices","access":"commercial","quote":"1Data obtained from a Bloomberg terminal at the Saïd Business school in Oxford on 23/05/2022.","mentions":1,"used_in":1}],"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 joint calibration convexifies the problem, yielding a unique solution for a fixed reference and regularity of the solution.","kind":"theoretical","quote":"First, it convexifies the problem giving a unique solution (for a fixed reference). Second, this convexity is known to give regularity to the solution"},{"finding":"Joint calibration matches market implied volatility skews for both equity and short‑rate options, as shown in Figures 2 and 3.","kind":"pricing-calibration","quote":"Figure 2: Implied volatility skews of the Short Rate calibrating options under the reference model and calibrated HW‑CEV model."},{"finding":"Compared to sequential calibration, the joint approach allows perturbations in the interest‑rate dynamics and can improve calibration accuracy.","kind":"performance","quote":"the joint problem allows for perturbations in the interest rate, whereas the full sequential problem does not"},{"finding":"Imposing bounds on β11 and β22 reduces numerical instabilities but does not eliminate them entirely.","kind":"statistical","quote":"Imposing the bounds reduced these instabilities significantly, but did not remove them entirely."},{"finding":"The optimisation parameters (ε1, ε2, δ bounds) are tuned to achieve tolerance levels for model‑observed IV differences and policy iteration convergence.","kind":"other","quote":"ε1 1 × 10−3 Tolerance for the difference in model and observed IV; ε2 1 × 10−8 Tolerance in the policy iteration for φ"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":21,"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":1,"extracted_at":"2026-11-14T18:22:55Z"},"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/"}