Investment optimization in the Heston model

Ya. I. Belopolskaya   ; A. A. Chubatov

Zapiski Nauchnykh Seminarov POMI, Probability and statistics. Part 35, Tome 526 (2023), pp. 29-51

Résumé

The investment portfolio optimization problem in the Heston model is solved via several reductions. Namely, we reduce the original problem to the Cauchy problem for a new fully nonlinear parabolic equation and construct its probabilistic representation via solution of a forward–backward stochastic differential equation (FBSDE). Next we reduce solution of the FBSDE to a new optimization problem and construct its numerical solution applying the neural network technique.

Bibliographie

[1] T. Bork, Teoriya arbitrazha v nepreryvnom vremeni, Izd. MTsNMO, M., 2010

[2] S. L. Heston, “A closed-form solution for options with stochastic volatility with applications to bond and currency options”, Review Financ. Stud., 6 (1993), 327–343 | DOI | MR | Zbl

[3] N.Touzi, Stochastic Control and Application to Finance, Ecole Polytechnique, Paris, 2018 | MR

[4] J. Han, A. Jentzen, E. Weinan, “Solving high-dimensional partial differential equations using deep learning”, Proc. Nat. Acad. Sci. USA, 115:34 (2018), 8505–8510 | DOI | MR | Zbl

[5] H. Pham, X. Warin, M. Germain, “Neural networks-based backward scheme for fully nonlinear PDEs”, SN Partial Diff. Eq. Appl., 2 (2021), 1–21 | DOI | MR

[6] C. Beck, M Hutzenthaler, A. Jentzen, B. Kuckuck, “An overview on deep learning-based approximation methods for partial differential equations”, Discrete Contin. Dyn. Syst., Ser. B, 28:6 (2023), 3697–3746 | DOI | MR

[7] M. Hutzenthaler, A. Jentzen, T. Kruse, “Overcoming the curse of dimensionality in the numerical approximation of parabolic partial differential equations with gradient-dependent nonlinearities”, Found. Comput. Math., 22 (2022), 905–966 | DOI | MR | Zbl

[8] Ya. Belopolskaya, Yu. Dalecky, “Investigation of the Cauchy problem with quasilinear systems with finite and infinite number of arguments by means of Markov random processes”, Izv. VUZ Math., 38:12 (1978), 6–17

[9] Ya. Belopolskaya, Yu. Dalecky, Stochastic Equations and Differential Geometry, Kluwer, 1990 | MR | Zbl

[10] Ya. Belopolskaya, W. A. Woyczynski, “SDEs, FBSDEs and fully nonlinear parabolic systems”, Rend. Sem. Mat. Univ. Politec. Torino, 71:2 (2013), 209–217 | MR | Zbl

[11] Ya. Belopolskaya, “Probabilistic interpretations of quasilinear parabolic equations”, AMS. Contemporary Mathematics, 734, 2019, 39–56 | DOI | MR

[12] E. Pardoux, S. Peng, “Backward stochastic differential equations and quasilinear parabolic partial differential”, Lect. Notes CIS, 176, 1992, 200–217 | MR | Zbl

[13] E. Pardoux, “Backward stochastic differential equations and viscosity solutions of systems of semilinear parabolic and elliptic pdes of second order”, Stoch. Anal. Relat. Topics: The Geilo Workshop, Birkháuser, 1996, 79–127 | MR

[14] P. Cheridito, M. Soner, N. Touzi, N. Victoir, “Second order backward stochastic differential equations and fully nonlinear parabolic PDEs”, Commun. Pure Appl. Math., 60:7 (2007), 1081–1110 | DOI | MR | Zbl

[15] M. Soner, N. Touzi, J. Zhang, “Wellposedness of second order backward SDEs”, Probab. Theory Relat. Fields, 153 (2012), 149–190 | DOI | MR | Zbl

[16] S. Ji, S. Peng, Y. Peng, X. Zhang, A deep learning method for solving stochastic optimal control problems driven by fully-coupled FBSDEs, 2022, arXiv: 2204.05796

[17] M. Raissi, Forward-backward stochastic neural networks: deep learning of high dimensional partial differential equations, 2018, arXiv: 1804.07010

[18] A. G. Baydin, B. A. Pearlmutter, A. A. Radul, J. M. Siskind, Automatic differentiation in machine learning: a survey, 2015, arXiv: 1502.05767 | MR

[19] G. Cybenko, “Approximation by superpositions of a sigmoidal function”, Math. Control Signals Syst., 2:4 (1989), 303–314 | DOI | MR | Zbl

[20] I. Goodfellow, Y. Bengio, A. Courville, Deep Learning, MIT Press, 2016 | MR | Zbl