Abstract
High-dimensional partial differential equations (PDEs) appear in a number of models from the financial industry, such as in derivative pricing models, credit valuation adjustment models, or portfolio optimization models. The PDEs in such applications are high-dimensional as the dimension corresponds to the number of financial assets in a portfolio. Moreover, such PDEs are often fully nonlinear due to the need to incorporate certain nonlinear phenomena in the model such as default risks, transaction costs, volatility uncertainty (Knightian uncertainty), or trading constraints in the model. Such high-dimensional fully nonlinear PDEs are exceedingly difficult to solve as the computational effort for standard approximation methods grows exponentially with the dimension. In this work, we propose a new method for solving high-dimensional fully nonlinear second-order PDEs. Our method can in particular be used to sample from high-dimensional nonlinear expectations. The method is based on (1) a connection between fully nonlinear second-order PDEs and second-order backward stochastic differential equations (2BSDEs), (2) a merged formulation of the PDE and the 2BSDE problem, (3) a temporal forward discretization of the 2BSDE and a spatial approximation via deep neural nets, and (4) a stochastic gradient descent-type optimization procedure. Numerical results obtained using TensorFlow in Python illustrate the efficiency and the accuracy of the method in the cases of a 100-dimensional Black–Scholes–Barenblatt equation, a 100-dimensional Hamilton–Jacobi–Bellman equation, and a nonlinear expectation of a 100-dimensional G-Brownian motion.
Similar content being viewed by others
References
Amadori, A.L.: Nonlinear integro-differential evolution problems arising in option pricing: a viscosity solutions approach. Differ. Integral Equ. 16(7), 787–811 (2003)
Avellaneda, M., Arnon, L., Parás, A.: Pricing and hedging derivative securities in markets with uncertain volatilities. Appl. Math. Finance 2, 73–88 (1995)
Bally, V., Pagès, G.: A quantization algorithm for solving multi-dimensional discrete-time optimal stopping problems. Bernoulli 9(6), 1003–1049 (2003)
Barron, A.R.: Universal approximation bounds for superpositions of a sigmoidal function. IEEE Trans. Inf. Theory 39(3), 930–945 (1993)
Bayraktar, E., Young, V.: Pricing options in incomplete equity markets via the instantaneous Sharpe ratio. Ann. Finance 4(4), 399–429 (2008)
Bayraktar, E., Milevsky, M.A., Promislow, S.D., Young, V.R.: Valuation of mortality risk via the instantaneous Sharpe ratio: applications to life annuities. J. Econ. Dyn. Control 33(3), 676–691 (2009)
Bender, C., Denk, R.: A forward scheme for backward SDEs. Stoch. Process. Appl. 117(12), 1793–1812 (2007)
Bender, C., Schweizer, N., Zhuo, J.: A primal-dual algorithm for BSDEs. Math. Finance 27(3), 866–901 (2017)
Bengio, Y.: Learning deep architectures for AI. Found. Trends Mach. Learn. 2(1), 1–127 (2009)
Bergman, Y.Z.: Option pricing with differential interest rates. Rev. Financ. Stud. 8(2), 475–500 (1995)
Bismut, J.-M.: Conjugate convex functions in optimal stochastic control. J. Math. Anal. Appl. 44, 384–404 (1973)
Bouchard, B., Elie, R., Touzi, N.: Discrete-time approximation of BSDEs and probabilistic schemes for fully nonlinear PDEs. In: Advanced financial modelling, vol. 8 of Radon Series on Computational and Applied Mathematics, Walter de Gruyter, Berlin, pp. 91–124 (2009)
Bouchard, B.: Lecture notes on BSDEs: main existence and stability results. Ph.D. thesis, CEREMADE-Centre de Recherches en MAthématiques de la DEcision (2015)
Bouchard, B., Touzi, N.: Discrete-time approximation and Monte-Carlo simulation of backward stochastic differential equations. Stoch. Process. Appl. 111(2), 175–206 (2004)
Briand, P., Labart, C.: Simulation of BSDEs by Wiener chaos expansion. Ann. Appl. Probab. 24(3), 1129–1171 (2014)
Cai, Z.: Approximating quantum many-body wave-functions using artificial neural networks (2017). arXiv:1704.05148
Carleo, G., Troyer, M.: Solving the quantum many-body problem with artificial neural networks. Science 355(6325), 602–606 (2017)
Chang, D., Liu, H., Xiong, J.: A branching particle system approximation for a class of FBSDEs. Probab. Uncertain. Quant. Risk 1, 9 (2016). 34
Chassagneux, J.-F.: Linear multistep schemes for BSDEs. SIAM J. Numer. Anal. 52(6), 2815–2836 (2014)
Chassagneux, J.-F., Crisan, D.: Runge-Kutta schemes for backward stochastic differential equations. Ann. Appl. Probab. 24(2), 679–720 (2014)
Chassagneux, J.-F., Richou, A.: Numerical stability analysis of the Euler scheme for BSDEs. SIAM J. Numer. Anal. 53(2), 1172–1193 (2015)
Chassagneux, J.-F., Richou, A.: Numerical simulation of quadratic BSDEs. Ann. Appl. Probab. 26(1), 262–304 (2016)
Cheridito, P., Soner, H.M., Touzi, N., Victoir, N.: Second-order backward stochastic differential equations and fully nonlinear parabolic PDEs. Comm. Pure Appl. Math. 60(7), 1081–1110 (2007)
Chiaramonte, M., Kiener, M.: Solving differential equations using neural networks. Machine Learning Project (2013)
Crépey, S., Gerboud, R., Grbac, Z., Ngor, N.: Counterparty risk and funding: the four wings of the TVA. Int. J. Theor. Appl. Finance 16(2), 1350006 (2013)
Crisan, D., Manolarakis, K.: Probabilistic methods for semilinear partial differential equations. Applications to finance. M2AN Math. Model. Numer. Anal. 44 44(5), 1107–1133 (2010)
Crisan, D., Manolarakis, K.: Solving backward stochastic differential equations using the cubature method: application to nonlinear pricing. SIAM J. Financ. Math. 3(1), 534–571 (2012)
Crisan, D., Manolarakis, K.: Second order discretization of backward SDEs and simulation with the cubature method. Ann. Appl. Probab. 24(2), 652–678 (2014)
Crisan, D., Manolarakis, K., Touzi, N.: On the Monte Carlo simulation of BSDEs: an improvement on the Malliavin weights. Stoch. Process. Appl. 120(7), 1133–1158 (2010)
Cybenko, G.: Approximation by superpositions of a sigmoidal function. Math. Control Signals Syst. 2(4), 303–314 (1989)
Darbon, J., Osher, S.: Algorithms for overcoming the curse of dimensionality for certain Hamilton–Jacobi equations arising in control theory and elsewhere. Res. Math. Sci. 3, 19 (2016)
Dehghan, M., Nourian, M., Menhaj, M.B.: Numerical solution of Helmholtz equation by the modified Hopfield finite difference techniques. Numer. Methods Partial Differ. Equ. 25(3), 637–656 (2009)
Delarue, F., Menozzi, S.: A forward–backward stochastic algorithm for quasi-linear PDEs. Ann. Appl. Probab. 16(1), 140–184 (2006)
Douglas Jr., J., Ma, J., Protter, P.: Numerical methods for forward–backward stochastic differential equations. Ann. Appl. Probab. 6(3), 940–968 (1996)
E, W., Hutzenthaler, M., Jentzen, A., Kruse, T.: Linear scaling algorithms for solving high-dimensional nonlinear parabolic differential equations. (2017a). arXiv:1607.03295
E, W., Hutzenthaler, M., Jentzen, A., Kruse, T.: On multilevel Picard numerical approximations for high-dimensional nonlinear parabolic partial differential equations and high-dimensional nonlinear backward stochastic differential equations. (2017b). arXiv:1708.03223
E, W., Han, J., Jentzen, A.: Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations. Commun. Math. Stat. 5(4), 349–380 (2017c)
Ekren, I., Muhle-Karbe, J.: Portfolio choice with small temporary and transient price impact (2017). arXiv:1705.00672
El Karoui, N., Peng, S., Quenez, M.C.: Backward stochastic differential equations in finance. Math. Finance 7(1), 1–71 (1997)
Fahim, A., Touzi, N., Warin, X.: A probabilistic numerical method for fully nonlinear parabolic PDEs. Ann. Appl. Probab. 21(4), 1322–1364 (2011)
Forsyth, P.A., Vetzal, K.R.: Implicit solution of uncertain volatility/transaction cost option pricing models with discretely observed barriers. Appl. Numer. Math. 36(4), 427–445 (2001)
Fu, Y., Zhao, W., Zhou, T.: Efficient spectral sparse grid approximations for solving multi-dimensional forward backward SDEs. Discrete Contin. Dyn. Syst. Ser. B 22(9), 3439–3458 (2017)
Geiss, S., Ylinen, J.: Decoupling on the Wiener space, related Besov spaces, and applications to BSDEs. (2014). arXiv:1409.5322
Geiss, C., Labart, C.: Simulation of BSDEs with jumps by Wiener chaos expansion. Stoch. Process. Appl. 126(7), 2123–2162 (2016)
Glorot, X., Bordes, A., Bengio, Y.: Deep sparse rectifier neural networks. In: Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics (Fort Lauderdale, FL, USA, 11–13 Apr 2011), G. Gordon, D. Dunson, and M. Dudk, Eds., vol. 15 of Proceedings of Machine Learning Research, PMLR, pp. 315–323
Gobet, E., Lemor, J.-P.: Numerical simulation of BSDEs using empirical regression methods: theory and practice. (2008). arXiv:0806.4447
Gobet, E., Labart, C.: Solving BSDE with adaptive control variate. SIAM J. Numer. Anal. 48(1), 257–277 (2010)
Gobet, E., Turkedjiev, P.: Approximation of backward stochastic differential equations using Malliavin weights and least-squares regression. Bernoulli 22(1), 530–562 (2016a)
Gobet, E., Turkedjiev, P.: Linear regression MDP scheme for discrete backward stochastic differential equations under general conditions. Math. Comp. 85(299), 1359–1391 (2016b)
Gobet, E., Lemor, J.-P., Warin, X.: A regression-based Monte Carlo method to solve backward stochastic differential equations. Ann. Appl. Probab. 15(3), 2172–2202 (2005)
Gobet, E., López-Salas, J.G., Turkedjiev, P., Vázquez, C.: Stratified regression Monte-Carlo scheme for semilinear PDEs and BSDEs with large scale parallelization on GPUs. SIAM J. Sci. Comput. 38(6), C652–C677 (2016)
Grohs, P., Hornung, F., Jentzen, A., von Wurstemberger, P.: A proof that artificial neural networks overcome the curse of dimensionality in the numerical approximation of Black–Scholes partial differential equations (2018)
Guo, W., Zhang, J., Zhuo, J.: A monotone scheme for high-dimensional fully nonlinear PDEs. Ann. Appl. Probab. 25(3), 1540–1580 (2015)
Guyon, J., Henry-Labordère, P.: The uncertain volatility model: a Monte Carlo approach. J. Comput. Finance 14(3), 37–61 (2011)
Han, J., Jentzen, A., E, W.: Overcoming the curse of dimensionality: solving high-dimensional partial differential equations using deep learning (2017). arXiv:1707.02568
Han, J., E, W.: Deep learning approximation for stochastic control problems (2016). arXiv:1611.07422
Henry-Labordère, P., Oudjane, N., Tan, X., Touzi, N., Warin, X.: Branching diffusion representation of semilinear PDEs and Monte Carlo approximation (2016). arXiv:1603.01727
Henry-Labordère, P.: Counterparty risk valuation: a marked branching diffusion approach (2012). arXiv:1203.2369
Henry-Labordère, P., Tan, X., Touzi, N.: A numerical algorithm for a class of BSDEs via the branching process. Stoch. Process. Appl. 124(2), 1112–1140 (2014)
Hinton, G., Deng, L., Yu, D., Dahl, G.E., Mohamed, A-r, Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T.N., Kingsbury, B.: Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups. IEEE Signal Process. Mag. 29(6), 82–97 (2012)
Hornik, K., Stinchcombe, M., White, H.: Multilayer feedforward networks are universal approximators. Neural Netw. 2(5), 359–366 (1989)
Huijskens, T.P., Ruijter, M.J., Oosterlee, C.W.: Efficient numerical Fourier methods for coupled forward–backward SDEs. J. Comput. Appl. Math. 296, 593–612 (2016)
Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. In: Proceedings of The 32nd International Conference on Machine Learning (ICML) (2015)
Jentzen, A., Kuckuck, B., Neufeld, A., von Wurstemberger, P.: Strong error analysis for stochastic gradient descent optimization algorithms (2018). arXiv:1801.09324
Karatzas, I., Shreve, S.E.: Brownian motion and stochastic calculus, 2nd edn, vol. 113 of Graduate Texts in Mathematics. Springer, New York, (1991)
Khoo, Y., Lu, J., Ying, L.: Solving parametric PDE problems with artificial neural networks (2017). arXiv:1707.03351
Kingma, D., Ba, J.: Adam: a method for stochastic optimization. In: Proceedings of the International Conference on Learning Representations (ICLR) (2015)
Kloeden, P.E., Platen, E.: Numerical Solution of Stochastic Differential Equations, vol. 23 of Applications of Mathematics (New York). Springer, Berlin (1992)
Kong, T., Zhao, W., Zhou, T.: Probabilistic high order numerical schemes for fully nonlinear parabolic PDEs. Commun. Comput. Phys. 18(5), 1482–1503 (2015)
Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks. Adv. Neural Inf. Process. Syst. 25, 1097–1105 (2012)
Labart, C., Lelong, J.: A parallel algorithm for solving BSDEs. Monte Carlo Methods Appl. 19(1), 11–39 (2013)
Lagaris, I.E., Likas, A., Fotiadis, D.I.: Artificial neural networks for solving ordinary and partial differential equations. IEEE Trans. Neural Netw. 9(5), 987–1000 (1998)
Laurent, J.-P., Amzelek, P., Bonnaud, J.: An overview of the valuation of collateralized derivative contracts. Rev. Deriv. Res. 17(3), 261–286 (2014)
LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. Proc. IEEE 86(11), 2278–2324 (1998)
LeCun, Y., Bottou, L., Orr, G.B., Müller, K.R.: Efficient BackProp, pp. 9–50. Springer, Berlin, Heidelberg (1998)
LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature 521, 436–444 (2015)
Lee, H., Kang, I.S.: Neural algorithm for solving differential equations. J. Comput. Phys. 91(1), 110–131 (1990)
Leland, H.E.: Option pricing and replication with transactions costs. J. Finance 40(5), 1283–1301 (1985)
Lemor, J.-P., Gobet, E., Warin, X.: Rate of convergence of an empirical regression method for solving generalized backward stochastic differential equations. Bernoulli 12(5), 889–916 (2006)
Lionnet, A., dos Reis, G., Szpruch, L.: Time discretization of FBSDE with polynomial growth drivers and reaction-diffusion PDEs. Ann. Appl. Probab. 25(5), 2563–2625 (2015)
Ma, J., Yong, J.: Forward–backward stochastic differential equations and their applications, vol. 1702 of Lecture Notes in Mathematics. Springer, Berlin (1999)
Ma, J., Protter, P., Yong, J.M.: Solving forward-backward stochastic differential equations explicitly—a four step scheme. Probab. Theory Relat. Fields 98(3), 339–359 (1994)
Ma, J., Protter, P., San Martín, J., Torres, S.: Numerical method for backward stochastic differential equations. Ann. Appl. Probab. 12(1), 302–316 (2002)
Maruyama, G.: Continuous Markov processes and stochastic equations. Rend. Circ. Mat. Palermo (2) 4, 48–90 (1955)
McKean, H.P.: Application of Brownian motion to the equation of Kolmogorov–Petrovskii–Piskunov. Commun. Pure Appl. Math. 28(3), 323–331 (1975)
Meade Jr., A.J., Fernández, A.A.: The numerical solution of linear ordinary differential equations by feedforward neural networks. Math. Comput. Model. 19(12), 1–25 (1994)
Mehrkanoon, S., Suykens, J.A.: Learning solutions to partial differential equations using LS-SVM. Neurocomputing 159, 105–116 (2015)
Milstein, G.N.: Approximate integration of stochastic differential equations. Teor. Verojatnost. i Primenen. 19, 583–588 (1974)
Milstein, G.N., Tretyakov, M.V.: Numerical algorithms for forward–backward stochastic differential equations. SIAM J. Sci. Comput. 28(2), 561–582 (2006)
Milstein, G.N., Tretyakov, M.V.: Discretization of forward–backward stochastic differential equations and related quasi-linear parabolic equations. IMA J. Numer. Anal. 27(1), 24–44 (2007)
Moreau, L., Muhle-Karbe, J., Soner, H.M.: Trading with small price impact. Math. Finance 27(2), 350–400 (2017)
Øksendal, B.: Stochastic differential equations. An introduction with applications. Universitext. Springer, Berlin (1985)
Pardoux, E., Peng, S.: Backward, stochastic differential equations and quasilinear parabolic partial differential equations. In: Stochastic Partial Differential Equations and their Applications (Charlotte, NC 1991), vol. 176 of Lecture Notes in Control and Information Sciences, pp. 200–217. Springer, Berlin (1992)
Pardoux, É., Peng, S.: Adapted solution of a backward stochastic differential equation. Syst. Control Lett. 14(1), 55–61 (1990)
Pardoux, E., Tang, S.: Forward-backward stochastic differential equations and quasilinear parabolic PDEs. Probab. Theory Relat. Fields 114(2), 123–150 (1999)
Peng, S.: \(G\)-expectation, \(G\)-Brownian motion and related stochastic calculus of Itô type. In: Stochastic Analysis and Applications, vol. 2 of Abel Symposium, pp. 541–567. Springer, Berlin (2007)
Peng, S.: Nonlinear expectations and stochastic calculus under uncertainty (2010). arXiv:1002.4546
Peng, S.: Nonlinear expectations, nonlinear evaluations and risk measures. In: Stochastic Methodsin Finance, vol. 1856 of Lecture Notes in Mathematics, pp. 165–253. Springer, Berlin (2004)
Peng, S.: Probabilistic interpretation for systems of quasilinear parabolic partial differential equations. Stoch. Stoch. Rep. 37(1–2), 61–74 (1991)
Peng, S.: Nonlinear expectations and nonlinear Markov chains. Chin. Ann. Math. Ser. B 26(2), 159–184 (2005)
Petersen, P., Voigtlaender, F.: Optimal approximation of piecewise smooth functions using deep relu neural networks (2017). arXiv:1709.05289
Pham, H.: Feynman–Kac representation of fully nonlinear PDEs and applications. Acta Math. Vietnam. 40(2), 255–269 (2015)
Possamaï, D., Mete Soner, H., Touzi, N.: Homogenization and asymptotics for small transaction costs: the multidimensional case. Commun. Partial Differ. Equ. 40(11), 2005–2046 (2015)
Ramuhalli, P., Udpa, L., Udpa, S.S.: Finite-element neural networks for solving differential equations. IEEE Trans. Neural Netw. 16(6), 1381–1392 (2005)
Rasulov, A., Raimova, G., Mascagni, M.: Monte Carlo solution of Cauchy problem for a nonlinear parabolic equation. Math. Comput. Simulation 80(6), 1118–1123 (2010)
Ruder, S.: An overview of gradient descent optimization algorithms (2016). arXiv:1609.04747
Ruijter, M.J., Oosterlee, C.W.: A Fourier cosine method for an efficient computation of solutions to BSDEs. SIAM J. Sci. Comput. 37(2), A859–A889 (2015)
Ruijter, M.J., Oosterlee, C.W.: Numerical Fourier method and second-order Taylor scheme for backward SDEs in finance. Appl. Numer. Math. 103, 1–26 (2016)
Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Learning internal representations by error propagation. Technical report, California University of San Diego La Jolla, Institute for Cognitive Science (1985)
Ruszczynski, A., Yao, J.: A dual method for backward stochastic differential equations with application to risk valuation (2017). arXiv:1701.06234
Silver, D., Huang, A., Maddison, C.J., Guez, A., Sifre, L., Van Den Driessche, G., Schrittwieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., et al.: Mastering the game of go with deep neural networks and tree search. Nature 529(7587), 484–489 (2016)
Sirignano, J., Spiliopoulos, K.: DGM: a deep learning algorithm for solving partial differential equations (2017). arXiv:1708.07469
Skorohod, A.V.: Branching diffusion processes. Teor. Verojatnost. i Primenen. 9, 492–497 (1964)
Tadmor, E.: A review of numerical methods for nonlinear partial differential equations. Bull. Am. Math. Soc. (N.S.) 49(4), 507–554 (2012)
Thomée, V.: Galerkin finite element methods for parabolic problems, vol. 25 of Springer Series in Computational Mathematics. Springer, Berlin (1997)
Turkedjiev, P.: Two algorithms for the discrete time approximation of Markovian backward stochastic differential equations under local conditions. Electron. J. Probab. 20(50), 49 (2015)
von Petersdorff, T., Schwab, C.: Numerical solution of parabolic equations in high dimensions. M2AN Math. Model. Numer. Anal. 38(1), 93–127 (2004)
Warin, X.: Variations on branching methods for non linear PDEs (2017). arXiv:1701.07660
Watanabe, S.: On the branching process for Brownian particles with an absorbing boundary. J. Math. Kyoto Univ. 4, 385–398 (1965)
Windcliff, H., Wang, J., Forsyth, P.A., Vetzal, K.R.: Hedging with a correlated asset: solution of a nonlinear pricing PDE. J. Comput. Appl. Math. 200(1), 86–115 (2007)
Zhang, J.: A numerical scheme for BSDEs. Ann. Appl. Probab. 14(1), 459–488 (2004)
Zhang, G., Gunzburger, M., Zhao, W.: A sparse-grid method for multi-dimensional backward stochastic differential equations. J. Comput. Math. 31(3), 221–248 (2013)
Zhao, W., Zhou, T., Kong, T.: High order numerical schemes for second-order FBSDEs with applications to stochastic optimal control. Commun. Comput. Phys. 21(3), 808–834 (2017)
Acknowledgements
Sebastian Becker and Jiequn Han are gratefully acknowledged for their helpful and inspiring comments regarding the implementation of the deep 2BSDE method.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Paul Newton.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix: Source Codes
Appendix: Source Codes
1.1 A.1: A Python Code for the Deep 2BSDE Method used in Subsection 4.1
The following Python code, Python code 1, is a simplified version of Python code 3 in Appendix A.3.
1.2 A.2: A Matlab Code for the Branching Diffusion Method used in Subsection 4.1
The following Matlab code is a slightly modified version of the Matlab code in (E et al. (2017c), Subsection 6.2).
1.3 A.3: A Python Code for the Deep 2BSDE Method used in Subsection 4.3
The following Python code is based on the Python code in (E et al. (2017c), Subsection 6.1).
1.4 A.4: A Matlab Code for the Classical Monte Carlo Method used in Subsection 4.4
The following Matlab code is a slightly modified version of the Matlab code in (E et al. (2017c), Subsection 6.3).
1.5 A.5: A Matlab Code for the Finite Differences Method used in Subsection 4.6
The following Matlab code is inspired by the Matlab code in (E et al. (2017b), MATLAB code 7 in Section 3).
Rights and permissions
About this article
Cite this article
Beck, C., E, W. & Jentzen, A. Machine Learning Approximation Algorithms for High-Dimensional Fully Nonlinear Partial Differential Equations and Second-order Backward Stochastic Differential Equations. J Nonlinear Sci 29, 1563–1619 (2019). https://doi.org/10.1007/s00332-018-9525-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00332-018-9525-3