End-to-End Neural Calibration of Stochastic Volatility Models via Option Panels
DOI:
https://doi.org/10.62051/w5dgtn92Keywords:
Stochastic volatility calibration; option panels; deep learning; Heston model; no-arbitrage constraints; Fourier pricing.Abstract
We propose an end-to-end neural architecture that calibrates stochastic volatility models directly from high-dimensional option panels. A 3-D CNN–Transformer hybrid learns the inverse mapping from implied-volatility surfaces to latent Heston parameters in milliseconds while enforcing no-arbitrage constraints via a differentiable penalty layer. Closed-form gradients of the Fourier pricing formula enable stable training on synthetic and market data. Extensive experiments on SPX and EURO STOXX 50 options show sub-0.3 % RMSE and a 100× speed-up over L-BFGS-B, with calibration times under 2ms on GPU.
Downloads
References
[1] X.-J. He and W. Chen, “A closed-form pricing formula for European options under a new stochastic volatility model with a stochastic long-term mean,” Math Finan Econ, vol. 15, no. 2, pp. 381–396, Mar. 2021, doi: 10.1007/s11579-020-00281-y. DOI: https://doi.org/10.1007/s11579-020-00281-y
[2] A. W. Van Der Stoep, L. A. Grzelak, and C. W. Oosterlee, “COLLOCATING VOLATILITY: A COMPETITIVE ALTERNATIVE TO STOCHASTIC LOCAL VOLATILITY MODELS,” Int. J. Theor. Appl. Finan., vol. 23, no. 06, p. 2050038, Sep. 2020, doi: 10.1142/S0219024920500387. DOI: https://doi.org/10.1142/S0219024920500387
[3] R. Merino, J. Pospíšil, T. Sobotka, T. Sottinen, and J. Vives, “DECOMPOSITION FORMULA FOR ROUGH VOLTERRA STOCHASTIC VOLATILITY MODELS,” Int. J. Theor. Appl. Finan., vol. 24, no. 02, p. 2150008, Mar. 2021, doi: 10.1142/S0219024921500084. DOI: https://doi.org/10.1142/S0219024921500084
[4] B. Jourdain and A. Zhou, “Existence of a calibrated regime switching local volatility model,” Mathematical Finance, vol. 30, no. 2, pp. 501–546, Apr. 2020, doi: 10.1111/mafi.12231. DOI: https://doi.org/10.1111/mafi.12231
[5] A. Sepp and P. Rakhmonov, “LOG-NORMAL STOCHASTIC VOLATILITY MODEL WITH QUADRATIC DRIFT,” Int. J. Theor. Appl. Finan., vol. 26, no. 08, p. 2450003, Dec. 2023, doi: 10.1142/S0219024924500031. DOI: https://doi.org/10.1142/S0219024924500031
[6] S.-C. Necula and V.-D. Păvăloaia, “AI-Driven Recommendations: A Systematic review of the state of the art in E-Commerce,” Applied Sciences, vol. 13, no. 9, p. 5531, 2023. DOI: https://doi.org/10.3390/app13095531
[7] S. Acharya, “Study of the effectiveness of chatbots in customer service on e-commerce websites,” 2023, Accessed: Jan. 08, 2025. [Online]. Available: https://www.theseus.fi/bitstream/handle/10024/807998/Acharya_Shiva.pdf?sequence=4.
[8] A. Brace, K. Gellert, and E. Schlögl, “SOFR term structure dynamics—Discontinuous short rates and stochastic volatility forward rates,” Journal of Futures Markets, vol. 44, no. 6, pp. 936–985, Jun. 2024, doi: 10.1002/fut.22499. DOI: https://doi.org/10.1002/fut.22499
[9] A. S. George, “AI Supremacy at the Price of Privacy: Examining the Tech Giants’ Race for Data Dominance,” Partners Universal Innovative Research Publication, vol. 3, no. 1, pp. 26–43, 2025.
[10] J.-L. Dupret, J. Barbarin, and D. Hainaut, “Impact of rough stochastic volatility models on long-term life insurance pricing,” Eur. Actuar. J., vol. 13, no. 1, pp. 235–275, Jun. 2023, doi: 10.1007/s13385-022-00317-1. DOI: https://doi.org/10.1007/s13385-022-00317-1
[11] M. Madanchian, “The Impact of Artificial Intelligence Marketing on E-Commerce Sales,” Systems, vol. 12, no. 10, p. 429, 2024. DOI: https://doi.org/10.3390/systems12100429
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Transactions on Computer Science and Intelligent Systems Research
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
