Abstract
The paper develops a new deep learning based scheme for solving high-dimensional nonlinear forward-backward stochastic differential equations (FBSDE) and associated partial differential equations. Firstly, the original BSDE is split into the linear dominant BSDE part and the nonlinear residual BSDE part. Then the linear BSDE part is approximated with high accuracy using a weak approximation technique. To approximate the nonlinear BSDE part, Deep BSDE solver is applied with asymptotic expansions which work as control variates. A sharp error estimate provides how the new scheme improves the original Deep BSDE method. Numerical experiments for high-dimensional nonlinear models show the validity and the effectiveness of the new scheme in financial application.
Similar content being viewed by others
References
E, W., Han, J., & Jentzen, A. (2017). Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations. Communications in Mathematics and Statistics, 5(4), 349–380.
El Karoui, N., Peng, S., & Quenez, M. C. (1997). Backward stochastic differential equations in finance. Mathematical Finance, 7(1), 1–71.
Fujii, M., Takahashi, A., & Takahashi, M. (2019). Asymptotic expansion as prior knowledge in deep learning method for high dimensional BSDEs. Asia-Pacific Financial Markets, 26(3), 391–408.
Han, J., & Long, J. (2020). Convergence of the Deep BSDE method for coupled FBSDEs. Probability, Uncertainty and Quantitative Risk, 5(5), 1–33.
Han, J., Jentzen, A., & E, W. (2018). Solving high-dimensional partial differential equations using deep learning. Proceedings of the National Academy of Sciences, 115(34), 8505–8510.
Han, J., Zhang, L., & E, W. (2019). Solving many-electron Schrödinger equation using deep neural networks. Journal of Computational Physics, 399, 108929.
Han, J., Lu, J., & Zhou, M. (2020). Solving high-dimensional eigenvalue problems using deep neural networks: A diffusion Monte Carlo like approach. Journal of Computational Physics, 423, 109792.
Hu, Y., & Watanabe, S. (1996). Donsker's delta functions and approximation of heat kernels by the time discretization methods. Journal of Mathematics of Kyoto University, 36(3), 499–518.
Iguchi, Y., & Yamada, T. (2021). Operator splitting around Euler-Maruyama scheme and high order discretization of heat kernels. ESAIM: Mathematical Modelling and Numerical Analysis, 55, 323–367.
Iguchi, Y., Naito, R., Okano, Y., Takahashi, A., & Yamada, T. (2021). Deep asymptotic expansion: Application to financial mathematics. IEEE CSDE 2021 (to appear).
Kloeden, P. E., & Platen, E. (1992). Numerical Solution of Stochastic Differential Equations. Springer.
Kunitomo, N., & Takahashi, A. (2003). On validity of the asymptotic expansion approach in contingent claim analysis. Annals of Applied Probability, 13(3), 914–952.
Li, Y., Lu, J., & Mao, A. (2020). Variational training of neural network approximations of solution maps for physical models. Journal of Computational Physics, 409(15), 109338.
Naito, R., & Yamada, T. (2019). A third-order weak approximation of multidimensional Itô stochastic differential equations. Monte Carlo Methods and Applications, 25(2), 97–120.
Naito, R., & Yamada, T. (2020). An acceleration scheme for deep learning-based BSDE solver using weak expansions. International Journal of Financial Engineering, 7, 2050012.
Naito, R., & Yamada, T. (2021). A higher order weak approximation of McKean-Vlasov type SDEs. BIT Numerical Mathematics (published online first).
Nualart, D. (2006). The Malliavin Calculus and Related Topics. Springer.
Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019). Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378, 686–707.
Raynal, P. E. C., & Trillos, C. A. G. (2015). A cubature based algorithm to solve decoupled McKean-Vlasov forward-backward stochastic differential equations. Stochastic Processes and their Applications, 125(6), 2206–2255.
Takahashi, A., & Yamada, T. (2012). An asymptotic expansion with push-down of Malliavin weights. SIAM Journal on Financial Mathematics, 3, 95–136.
Takahashi, A., & Yamada, T. (2015). An asymptotic expansion of forward-backward SDEs with a perturbed driver. International Journal of Financial Engineering, 2(2), 1550020.
Takahashi, A., & Yamada, T. (2016). A weak approximation with asymptotic expansion and multidimensional Malliavin weights. Annals of Applied Probability, 26(2), 818–856.
Takahashi, A., Tsuchida, Y., & Yamada, T. (2022). A new efficient approximation scheme for solving high-dimensional semilinear PDEs: Control variate method for Deep BSDE solver. Journal of Computational Physics (published online first).
Yamada, T. (2019). An arbitrary high order weak approximation of SDE and Malliavin Monte Carlo: Application to probability distribution functions. SIAM Journal on Numerical Analysis, 57(2), 563–591.
Yamada, T. (2021). High order weak approximation for irregular functionals of time-inhomogeneous SDEs. Monte Carlo Methods and Applications, 27(2), 117–136.
Yamada, T. (2022). A Gaussian Kusuoka approximation without solving random ODEs. SIAM Journal on Financial Mathematics (published online first).
Yamada, T., & Yamamoto, K. (2019). Second order discretization of Bismut-Elworthy-Li formula: Application to sensitivity analysis. SIAM/ASA Journal on Uncertainty Quantification, 7(1), 143–173.
Yamada, T., & Yamamoto, K. (2020). A second order discretization with Malliavin weight and Quasi Monte Carlo method for option pricing. Quantitative Finance, 20(11), 1825–1837.
Zhang, J. (2017). Backward Stochastic Differential Equations. Springer.
Acknowledgements
I would like to express my sincere gratitude to Prof. Akihiko Takahashi (University of Tokyo) and Prof. Toshihiro Yamada (Hitotsubashi University) for their comments and advice on the method of this paper. Also, I would like to thank Mr. Riu Naito (Japan Post Insurance Co., Ltd. & Hitotsubashi University) for his support on numerical experiments. Lastly, I thank two anonymous reviewers for their grateful comments and suggestions.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no conflict of interest.
Ethical standards
The authors declare that the experiments comply with the current laws of the country in which they were performed.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Tsuchida, Y. Control Variate Method for Deep BSDE Solver Using Weak Approximation. Asia-Pac Financ Markets 30, 273–296 (2023). https://doi.org/10.1007/s10690-022-09374-8
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10690-022-09374-8