{"paper_id":"2310.16703","title":"No-Arbitrage Deep Calibration for Volatility Smile and Skewness","authors":["Kentaro Hoshisashi","Carolyn E. Phelan","Paolo Barucca"],"paper_date":"2023-10-25","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2310.16703v3","page_url":"https://thequant.space/flowcharts/no-arbitrage-deep-calibration-for-volatility-smile-and-skewn/","scores":{"empirical_rigor":5.5,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a Derivative-Constrained Neural Network (DCNN) that extends a multilayer perceptron with derivative terms in the loss to calibrate the implied volatility surface while enforcing no‑arbitrage constraints.","quote":"we have introduced a Derivative-Constrained Neural Network (DCNN), which is an enhancement of a multilayer perceptron (MLP) that incorporates derivatives in the objective function.","summary":"Volatility smile and skewness are two key properties of option prices that are represented by the implied volatility (IV) surface. However, IV surface calibration through nonlinear interpolation is a complex problem due to several factors, including limited input data, low liquidity, and noise.…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Option price data are sparse, requiring interpolation or additional information.","type":"data","stated":"explicit","quote":"The challenge of this inverse problem is the scarcity of options price data."},{"assumption":"The SABR model parameters (α=0.2, β=1.0, ν=0.6, ρ=−0.4) generate synthetic option prices for training.","type":"model","stated":"explicit","quote":"... α = 0.2, β = 1.0, q = 0, and r = 0.04 as the parameters of the SABR model."},{"assumption":"No‑arbitrage conditions are represented by sign constraints on first and second derivatives of the call price.","type":"model","stated":"explicit","quote":"The necessary and sufficient conditions for no‑arbitrage are represented as non‑strict inequalities for several first and second derivatives, −e^{−rτ} ≤∂C/∂K ≤0"},{"assumption":"Random seeds are fixed across experiments for statistical consistency.","type":"statistical","stated":"explicit","quote":"We consistently used the same random seeds across different conditions for the statistical analysis and changed these seed values ten times."},{"assumption":"Automatic differentiation provides exact derivative values for the penalty terms.","type":"other","stated":"implied","quote":"DCNN allows us to generate a smooth surface and incorporate the no‑arbitrage condition thanks to the derivative terms in the loss function."}]},"datasets":{"status":"available","items":[{"name":"S&P 500 call options intraday dataset","provider":"Yahoo finance library","kind":"options","access":"public","period":"10th to 14th July 2023","universe":"S&P 500 call options","quote":"The empirical experiments were conducted using intraday prices for S&P 500 call options from 10th to 14th July 2023; we obtained about two thousand points on a daily basis via the Yahoo finance library as Table 1.","known_id":"yahoo-finance"},{"name":"Synthetic SABR‑generated option prices","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"In numerical experiments, we train the model using prices generated with the SABR model to produce smile and skewness parameters."}],"known":[{"id":"yahoo-finance","name":"Yahoo Finance","provider":"Yahoo","kind":"market-prices","access":"public","quote":"5.3 The S&P 500 dataset The empirical experiments were conducted using intraday prices for S&P 500 call options from 10th to 14th July 2023; we obtained about two thousand points on a daily basis via the Yahoo finance library as Table 1.","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":"DCNN improves interpolation of the premium surface for volatility smile and skewness compared with cubic spline and standard MLP.","kind":"pricing-calibration","quote":"DCNN significantly improved the interpolation of the premium surface for volatility smile and skewness due to deep learning with no‑arbitrage constraints."},{"finding":"DCNN produces fewer violations of the no‑arbitrage derivative constraints, as shown by reduced red‑coloured areas in derivative plots.","kind":"theoretical","quote":"In Figure 9, we see that the plots for DCNN have fewer areas in red, denoting the regions where the no‑arbitrage conditions on the derivatives are broken."},{"finding":"During training, the penalty associated with derivative terms decreases, indicating a balance between price accuracy and no‑arbitrage compliance.","kind":"performance","quote":"The findings demonstrate that as the learning process proceeds, there’s a decrease in the penalty associated with the derivative term, reflecting a balance between accuracy and penalty from derivative terms."},{"finding":"DCNN effectively enforces the sign constraints on first and second derivatives of the call price, satisfying the necessary no‑arbitrage conditions.","kind":"theoretical","quote":"DCNN effectively ensures that the results do not infringe on the conditions for derivative values, each of which aligns with an inequality constraint in Eq. (3)."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":23,"citations":{"cited_by_in_corpus":4,"by_independent_authors":3,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":3,"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":"2026-11-14T18:20:21Z"},"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/"}