{"paper_id":"2312.15950","title":"The implied volatility surface (also) is path-dependent","authors":["Hervé Andrès","Alexandre Boumezoued","Benjamin Jourdain"],"paper_date":"2023-12-26","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2312.15950v3","page_url":"https://thequant.space/flowcharts/the-implied-volatility-surface-also-is-path-dependent/","scores":{"empirical_rigor":6,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a new joint model for forecasting both the implied volatility surface and the underlying asset price, using a parsimonious SSVI parameterization with path‑dependent dynamics.","quote":"We propose a new model for the forecasting of both the implied volatility surfaces and the underlying asset price.","summary":"We propose a new model for the forecasting of both the implied volatility surfaces and the underlying asset price. In the spirit of Guyon and Lekeufack (2023) who are interested in the dependence of volatility indices (e.g. the VIX) on the paths of the associated equity indices (e.g. the S&P 500)…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The implied volatility surface can be represented by the SSVI parameterization with only four parameters (a, p, ρ, η).","type":"model","stated":"explicit","quote":"we fit to historical data a parsimonious version of the SSVI parameterization (Gatheral and Jacquier, 2014) of the implied volatility surface relying on only four parameters"},{"assumption":"The two parameters a and p governing ATM‑forward implied volatility follow a path‑dependent evolution driven by past returns and squared returns.","type":"model","stated":"explicit","quote":"the two parameters ruling the at-the-money-forward implied volatility as a function of the time-to-maturity exhibit a path-dependent behavior with respect to the underlying asset price"},{"assumption":"Residuals of a and p follow Ornstein‑Uhlenbeck processes; ρ and η follow Jacobi processes to ensure no static arbitrage.","type":"model","stated":"explicit","quote":"the residuals of these two parameters and Jacobi processes for the two other parameters"},{"assumption":"Underlying asset price dynamics follow the path‑dependent volatility model of Guyon and Lekeufack (2023).","type":"model","stated":"explicit","quote":"the underlying price using the path-dependent volatility framework of Guyon and Lekeufack (2023)"},{"assumption":"Only up to 12 years of past returns are used to predict implied volatility.","type":"data","stated":"explicit","quote":"Note that since we will use at most 12 years of past returns to predict the implied volatility, the whole time series are not used in the following study."},{"assumption":"Static arbitrage is absent if the SSVI parameters satisfy the known mild conditions.","type":"model","stated":"implied","quote":"guaranteeing the absence of static arbitrage in the simulated surfaces"}]},"datasets":{"status":"available","items":[{"name":"S&P 500 implied volatility surface","provider":null,"kind":"options","access":"unknown","period":"2012-03-08 to 2022-12-30","universe":"options on the S&P 500 index","quote":"The first one corresponds to options on the S&P 500 index"},{"name":"Euro Stoxx 50 implied volatility surface","provider":null,"kind":"options","access":"unknown","period":"2012-03-08 to 2022-12-30","universe":"options on the Euro Stoxx 50 index","quote":"The second one corresponds to options on the Euro Stoxx 50"},{"name":"Euro Stoxx 50 index price series","provider":null,"kind":"market-prices","access":"unknown","period":"1986-12-31 to 2022-12-30","universe":"Euro Stoxx 50 index","quote":"the Euro Stoxx 50 time series starts on December 31, 1986 and both end on December 30, 2022"}],"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":"unclear","findings":[{"finding":"A large part of ATM‑forward implied volatility movements (up to two‑year maturities) can be explained by past returns and their squares, with the effect weakening for longer maturities.","kind":"predictive-accuracy","quote":"Our empirical study reveals that a large part of the movements of the at-the-money-forward implied volatility for up to two years time-to-maturities can be explained using the past returns and their squares. Moreover, we show that this feedback effect gets weaker when the time-to-maturity increases."},{"finding":"The parsimonious four‑parameter SSVI fits market surfaces reasonably, with mean relative errors increasing from 0.80% (S&P 500) to 1.29% and from 1.18% to 1.60% for Euro Stoxx 50.","kind":"pricing-calibration","quote":"The mean of the average relative errors across the whole S&P 500 (resp. Euro Stoxx 50) data set increases from 0.80% (resp. 1.18%) to 1.29% (resp. 1.60%)."},{"finding":"Price errors also increase modestly (from 4‑6 bps to 7‑8 bps) when using the parsimonious SSVI versus the richer SSVI‑MPL model.","kind":"pricing-calibration","quote":"the mean of the average price errors accross the whole S&P 500 (resp. Euro Stoxx 50) data set increases from 4 bps (resp. 6 bps) to 7 bps (resp. 8 bps)."},{"finding":"The proposed joint model can simulate realistic implied volatility surface paths that are free from static arbitrage.","kind":"theoretical","quote":"Thanks to this model, we are able to simulate highly realistic paths of implied volatility surfaces that are free from static arbitrage."},{"finding":"Out‑of‑sample quantile envelopes of ATM implied volatility show good consistency between simulated and historical dynamics across multiple maturities.","kind":"statistical","quote":"Figure 14: Out-of-sample quantiles envelopes of the ATM implied volatility for several time-to-maturities."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":30,"citations":{"cited_by_in_corpus":1,"by_independent_authors":1,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":1,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/Jordylek/VolatilityIsMostlyPathDependent","note":null},"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":"2026-11-14T18:17:29Z"},"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/"}