{"id":"479c7f8b-3640-48de-afe0-50d67e97ca4f","arxiv_id":"2606.12450","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"A Legendre polynomial dimension reduction followed by Tikhonov regularization reconstructs terminal option prices from noisy current data in forward-time Black-Scholes with state-dependent volatility, with stability proofs for fixed truncation.","lead":"The paper develops a numerical method to recover future option prices at expiration from today's observed price profile by solving the Black-Scholes equation forward in time. It stabilizes this ill-posed problem using a finite Legendre polynomial expansion in the asset price combined with Tikhonov regularization.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No significant objection identified","rationale":"The reader's weakest_assumption correctly isolates the approximation quality of the Legendre reduction as the key unverified link between the reduced well-posedness results and recovery of the original PDE solution. Because the abstract states that proofs are supplied for the reduced problem and the numerical section reports successful recovery on several payoffs, the argument is internally consistent at the level of detail available; no stronger load-bearing concern emerges.","tokens_in":1778,"tokens_out":330,"duration_ms":24951,"concrete_test":"Reproduce the reduced ODE system for N=8 (or the smallest N used in the experiments) with the European-put payoff, integrate the Tikhonov-regularized forward problem from noisy initial data, and compare the recovered terminal profile against the known exact payoff; if the L2 recovery error scales consistently with the noise level as predicted by the stability estimate, the reduced-system analysis is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that existence, uniqueness, data stability, and convergence hold for the dimension-reduced Legendre--Tikhonov method at each fixed truncation level, with the finite shifted Legendre projection serving as a spectral cutoff that relaxes the S^2 degeneracy and yields a stable reduced ODE system. The abstract describes a technically coherent construction: projection onto the basis produces an ODE system in the coefficients, Tikhonov regularization is applied to this finite-dimensional inverse problem, and standard well-posedness arguments then apply for fixed N. No internal gap, hidden assumption, or inconsistency in this logic is visible from the given description.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper studies the ill-posed forward-time Black-Scholes equation with state-dependent volatility, where the current option-price profile is given and the terminal profile at T must be recovered. It projects the PDE onto a finite shifted Legendre basis in the asset-price variable to obtain a reduced ODE system, applies Tikhonov regularization to this finite-dimensional inverse problem, and claims to prove existence, uniqueness, data stability, and convergence for each fixed truncation level N. Numerical experiments recover terminal profiles for smooth, butterfly-spread, and European put payoffs from noisy data, with comparisons to a reduced PINN solver and the physical-space quasi-reversibility method.","tokens_in":1907,"tokens_out":375,"duration_ms":14007,"significance":"If the well-posedness and convergence results hold, the Legendre reduction provides a spectral cutoff that relaxes the S^2 degeneracy at S=0 and yields a stable reconstruction method for an important class of inverse option-pricing problems. The combination of dimension reduction, regularization, and secondary PINN benchmark is technically coherent and offers a concrete alternative to existing stabilization techniques.","major_comments":[],"minor_comments":[{"comment":"The abstract and introduction state that proofs of existence, uniqueness, stability, and convergence are given for fixed N, but the manuscript should explicitly reference the relevant theorem numbers and indicate where the spectral-cutoff argument for the degeneracy relaxation is formalized.","section":null},{"comment":"Numerical section: the reported error metrics and noise levels should be tabulated with explicit values of N and the Tikhonov parameter for each payoff to allow direct reproducibility.","section":null},{"comment":"The comparison with quasi-reversibility would benefit from a brief statement of the precise discretization and regularization parameters used in that baseline.","section":null}],"recommendation":"minor_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful reading and positive assessment of the manuscript. The recommendation for minor revision is noted. Since no specific major comments were raised, we interpret the request as pertaining to minor editorial or presentational adjustments that can be addressed in the revised version.","responses":[],"tokens_in":1309,"tokens_out":72,"duration_ms":7961,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The core contribution is a shifted Legendre projection that reduces the forward-time Black-Scholes PDE with state-dependent volatility to a finite ODE system, then applies Tikhonov regularization to recover the terminal price profile from noisy current data. For each fixed truncation they prove existence, uniqueness, stability, and convergence of the regularized solution.\n\nThe reduction does two useful things at once: it supplies a spectral cutoff that controls the instability and it removes the S^2 degeneracy at the zero-price boundary, so the reduced system is well-behaved. The numerical tests on smooth payoffs, butterfly spreads, and European puts show that the method recovers the terminal profile under noise, and the comparisons to quasi-reversibility and a reduced PINN give a practical sense of where it helps.\n\nThe soft spots are modest. The well-posedness results are standard once the problem is finite-dimensional, so the real work is in the application rather than new theory. The truncation error relative to the original PDE is not given explicit bounds, which means practical accuracy still depends on choosing N large enough without reintroducing instability. Details on how state-dependent volatility enters the projected coefficients are light in the description. No circularity or hidden fitting appears.\n\nThis is aimed at people who work on numerical methods for inverse problems in option pricing. A reader who needs a concrete, stabilized spectral technique for this forward formulation will get usable ideas and benchmarks. It is a coherent, targeted piece that deserves a serious referee to check the proofs and the full experimental controls.","headline":"The paper stabilizes forward Black-Scholes reconstruction with Legendre projection and Tikhonov, proving well-posedness for fixed N and showing decent numerics.","tokens_in":2413,"tokens_out":383,"would_cite":false,"duration_ms":20695,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Projecting the forward Black-Scholes PDE onto a finite shifted Legendre basis yields a stable ODE system that Tikhonov regularization solves for the terminal price profile.","keywords":["Black-Scholes equation","forward problem","Legendre polynomials","Tikhonov regularization","ill-posed problem","option pricing","spectral methods","dimension reduction"],"falsifier":"A test in which, for successively smaller noise levels in the initial price profile, the recovered terminal profile does not converge to the known terminal payoff or in which the reconstruction error increases rather than decreases with higher truncation levels.","tokens_in":2673,"feed_emoji":"📈","tokens_out":732,"duration_ms":28173,"temperature":0.7,"pith_summary":"The paper addresses the ill-posed forward-time Black-Scholes problem, in which the current option-price profile is given and the profile at maturity must be recovered, by first reducing the spatial dimension via projection onto shifted Legendre polynomials. This produces a finite system of ODEs in time whose solution approximates the PDE while avoiding the S-squared degeneracy at the asset-price boundary. The authors then apply Tikhonov regularization to this reduced system and prove existence, uniqueness, stability with respect to noisy data, and convergence for each fixed truncation level. Numerical tests on smooth, butterfly-spread, and European-put payoffs show that the resulting Legendre-Tikhonov reconstructions recover the terminal profile accurately even when the input data contain noise, while also outperforming conventional quasi-reversibility in stability.","feed_headline":"Legendre reduction yields stable forward Black-Scholes recovery","feed_subtitle":"Finite shifted expansion produces an ODE system whose Tikhonov solution reconstructs terminal prices from noisy current data","key_machinery":"The finite shifted Legendre expansion in the asset-price variable, which projects the PDE onto a stable ODE system and relaxes the S-squared boundary degeneracy before Tikhonov regularization is applied.","core_discovery":"The dimension-reduced Legendre-Tikhonov method, formed by projecting the forward Black-Scholes equation with state-dependent volatility onto a finite shifted Legendre basis in the asset-price variable and then regularizing the resulting ODE system, admits a unique solution for each truncation level that is stable to perturbations in the initial data and converges to the true terminal profile as the noise level tends to zero.","pith_inferences":["The same Legendre projection could be inserted into other forward parabolic pricing equations that suffer from similar boundary degeneracy.","Choosing the truncation level adaptively from the noise amplitude might further improve practical accuracy without altering the proved stability properties.","Because the reduced system is low-dimensional, the method may allow repeated reconstructions inside calibration loops that would be prohibitive with full PDE solvers."],"forward_implications":["For any fixed truncation level the reduced problem possesses a unique solution that depends continuously on the initial data.","The reconstruction converges to the true terminal profile as the noise amplitude in the current price profile tends to zero.","The Legendre reduction demonstrably stabilizes the problem relative to the conventional physical-space quasi-reversibility method.","The same reduced system admits a PINN solver that serves as a consistent numerical benchmark for the Tikhonov reconstructions."],"fun_headline_variants":["Legendre reduction stabilizes forward Black-Scholes","Regularized basis recovers terminal prices from noise","Legendre-Tikhonov method for forward-time reconstruction","Spectral Legendre cutoff tames pricing ill-posedness"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The finite shifted Legendre expansion in the asset-price variable acts as an effective spectral cutoff that relaxes the S-squared degeneracy at the zero-price boundary and yields a stable reduced ODE system whose solution approximates the original PDE solution sufficiently well for reconstruction purposes.","fun_headline_variants_meta":{"raw":{"variants":["Legendre reduction stabilizes forward Black-Scholes","Regularized basis recovers terminal prices from noise","Legendre-Tikhonov method for forward-time reconstruction","Spectral Legendre cutoff tames pricing ill-posedness"]},"model":"grok-4.3","cost_usd":0.003927,"raw_usage":{"total_tokens":2033,"prompt_tokens":709,"num_sources_used":0,"completion_tokens":58,"cost_in_usd_ticks":39274500,"prompt_tokens_details":{"text_tokens":709,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1266,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":709,"tokens_out":58,"duration_ms":9670,"temperature":1.0,"reasoning_tokens":1266,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-28T17:26:24.623921+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A test in which, for successively smaller noise levels in the initial price profile, the recovered terminal profile does not converge to the known terminal payoff or in which the reconstruction error increases rather than decreases with higher truncation levels.","supporting_citations":[],"review_version":1}