{"paper_id":"2611.20746","title":"European derivatives pricing via constrained Gaussian process regressions","authors":["Arnaud Servais","Donatien Hainaut"],"paper_date":"2026-11-16","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2611.20746","page_url":"https://thequant.space/flowcharts/european-derivatives-pricing-via-constrained-gaussian-process-regressions/","scores":{"empirical_rigor":7,"math_complexity":9,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a constrained Gaussian process regression method for pricing European derivatives under multivariate jump‑diffusion dynamics, conditioning the GP prior on the Feynman–Kac PIDE at collocation points.","quote":"This article develops a Gaussian process method for pricing European derivatives when the underlyings, and the risk factors driving them, follow a multivariate jump-diffusion.","summary":"This article develops a Gaussian process method for pricing European derivatives when the underlyings, and the risk factors driving them, follow a multivariate jump-diffusion. A Gaussian process prior on the pricing function is conditioned on the Feynman-Kac partial integro-differential equation at…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"An arbitrage‑free market is assumed and pricing is performed under a risk‑neutral measure Q.","type":"market","stated":"explicit","quote":"As this paper aims to price and hedge derivatives, we assume an arbitrage-free market and work under a risk-neutral measure Q. The coefficients µ, Σ, M and the processes W t, Lt are assumed to be specified so that the discounted prices of the tradable securities in yt are (Q, F)-martingales.","support":"full"},{"assumption":"The underlying assets and risk factors follow a multivariate jump‑diffusion process.","type":"model","stated":"explicit","quote":"This article develops a Gaussian process method for pricing European derivatives when the underlyings, and the risk factors driving them, follow a multivariate jump-diffusion.","support":"full"},{"assumption":"The kernel used in the Gaussian process must be at least twice differentiable in each argument.","type":"statistical","stated":"explicit","quote":"The kernel must therefore be at least twice differentiable in each argument.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Monte Carlo simulation benchmark","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"The method is illustrated with derivatives of two underlyings under the Heston model with Kou double-exponential jumps, a five-dimensional problem, and benchmarked against Monte Carlo simulation.","support":"full"},{"name":"Heston–Kou model synthetic paths","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"9 Numerical Application We show the generality of the method on a stochastic volatility model with jumps, the Heston– Kou model. It follows the model of Bates (1996), which adds log-normal jumps to the Heston (1993) stochastic variance, with the jumps replaced by the double-exponential jumps of Kou (2002).","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"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 method returns a closed‑form, differentiable price function, allowing Greeks to be obtained without simulation or finite differences.","kind":"theoretical","quote":"The price is returned as a closed-form and differentiable function, allowing the Greeks to be obtained without simulation or finite differences.","support":"full"},{"finding":"A convergence theorem shows the iterative scheme converges from any starting point in the supremum norm, with a worst‑case iteration bound.","kind":"theoretical","quote":"Our main theoretical contribution is a convergence result for this iteration: using Weissinger's fixed-point theorem in the supremum norm, we show that it converges from any starting point, and provide a worst-case number of iterations for a given precision.","support":"full"},{"finding":"On a five‑dimensional Heston–Kou basket problem, the CGPR prices and sensitivities agree with Monte Carlo benchmarks.","kind":"performance","quote":"Both their prices and sensitivities agreed with the Monte Carlo benchmark.","support":"full"},{"finding":"Error maps show low pricing errors over the domain, and the posterior mean surface tracks Monte Carlo prices with 95% credible bands containing the benchmarks.","kind":"performance","quote":"The profiles of Figure 1, obtained by averaging the grid over all but one coordinate, show that the posterior mean follows the benchmark along the price, the variance and time, and that the 95% band contains the Monte Carlo prices. The error maps of Figure 3 show low errors over the domain.","support":"full"},{"finding":"Greeks derived from the explicit price surface match Monte Carlo estimates without additional fitting or simulation.","kind":"performance","quote":"Finally, since the surface is an explicit function of the state, its Greeks follow directly from Section 8, without additional fits or simulations. Table 4 compares them with Monte Carlo estimates on a test set, and Figure 4 shows them along the first price.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":20,"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":3,"extracted_at":"2026-11-17T15:29:17Z"},"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/"}