Abstract
In this article we design a novel quasi-regression Monte Carlo algorithm in order to approximate the solution of discrete time backward stochastic differential equations, and we analyze the convergence of the proposed method. The algorithm also approximates the solution to the related semi-linear parabolic partial differential equation obtained through the well known Feynman–Kac representation. For the sake of enriching the algorithm with high order convergence a weighted approximation of the solution is computed and appropriate conditions on the parameters of the method are inferred. With the challenge of tackling problems in high dimensions we propose suitable projections of the solution and efficient parallelizations of the algorithm taking advantage of powerful many core processors such as graphics processing units.
Similar content being viewed by others
Notes
Fully coupled FBSDEs have the general form
$$\begin{aligned} {\mathrm{d}}X_t&= b(t,X_t,Y_t,Z_t) {\mathrm{d}}t + \sigma (t,X_t,Y_t,Z_t) \mathrm{d}W_t, \quad X_0 = x_0, \\ -{\mathrm{d}}Y_t&= f(t,X_t,Y_t,Z_t) {\mathrm{d}}t - Z_t {\mathrm{d}}W_t, \quad Y_T = g(X_T). \end{aligned}$$Single precision data type is assumed.
References
Bally V (1997) Approximation scheme for solutions of BSDE. In: Backward stochastic differential equations (Paris, 1995–1996), Pitman research notes in mmathematics series, vol 364, pp 177–191. Longman, Harlow
Bally V, Pagès G (2003) A quantization algorithm for solving multi-dimensional optimal stopping problems. Bernoulli 9(6):1003–1049
Barles G, Buckdahn R, Pardoux E (1997) Backward stochastic differential equations and integral-partial differential equations. Stoch Stoch Rep 60(1–2):57–83
Bellmann R (1961) Adaptive control processes: a guided tour. Princeton University Press, Princeton
Bender C, Denk R (2007) A forward scheme for backward SDEs. Stoch Process Appl 117(12):1793–1823
Bender C, Moseler T (2010) Importance sampling for backward SDEs. Stoch Anal Appl 28(2):226–253
Bergman Y (1995) Option pricing with differential interest rates. Rev Financ Stud 8(2):475–500
Bismut JM (1973) Conjugate convex functions in optimal stochastic control. J Math Anal Appl 44(2):384–404
Bouchard B, Tan X, Warin X, Zou Y (2017) Numerical approximation of BSDEs using local polynomial drivers and branching processes. Monte Carlo Methods Appl 23(4):241–263
Bouchard B, Touzi N (2004) Discrete time approximation and Monte-Carlo simulation of backward stochastic differential equations. Stoch Process Appl 111:175–206
Canuto C, Hussaini MY, Quarteroni A, Zang TA (2006) Spectral methods fundamentals in single domains. Springer, Berlin
Chan-Wai-Nam Q, Mikael J, Warin X (2019) Machine learning for semi linear PDEs. J Sci Comput. https://doi.org/10.1007/s10915-019-00908-3
Cheridito P, Soner HM, Touzi N, Victoir N (2007) Second-order backward stochastic differential equations and fully nonlinear parabolic PDEs. Commun Pure Appl Math 60(7):1081–1110
Chevance D (1997) Numerical methods for backward stochastic differential equations. In: Moffatt HK (ed) Numerical methods in finance. Cambridge University Press, Cambridge, pp 232–244
Constantine GM, Savits TH (1996) A multivariate Faa di Bruno formula with applications. Trans Am Math Soc 348(2):503–520
Crepey S (2013) Financial modeling: a backward stochastic differential equations perspective. Springer, Berlin
Cvitanic J, Ma J (1996) Hedging options for a large investor and Forward-Backward SDEs. Ann Appl Probab 6(2):370–398
Delarue F, Menozzi S (2006) A forward backward algorithm for quasi-linear PDEs. Ann Appl Probab 16(1):140–184
Delarue F, Menozzi S (2008) An interpolated stochastic algorithm for quasi-linear PDEs. Math Comput 77(261):125–158
Delong L (2013) Backward stochastic differential equations with jumpsand their actuarial and financial applications: BSDEs with jumps. Springer, Berlin
Di Pietro D, Ern A (2011) Mathematical aspects of discontinuous Galerkin methods, vol 69. Springer, Berlin
Duffie D, Epstein L (1992) Stochastic differential utility. Econometrica 60(2):353–394
E W, Han J, Jentzen A (2017) Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations. Commun Math Stat 5(4):349–380
El Karoui N, Hamadène S, Matoussi A (2008) Backward stochastic differential equations and applications. In: Carmona R (ed) Indifference pricing: theory and applications, vol 8. Springer, Berlin, pp 267–320
El Karoui N, Kapoudjian C, Pardoux E, Peng S, Quenez M (1997) Reflected solutions of BSDEs and related obstacle problems for PDE’s. Ann Probab 25(2):702–737
El Karoui N, Peng S, Quenez M (1997) Backward stochastic differential equations in finance. Math Financ 7(1):1–71
El Karoui N, Quenez M (1995) Dynamic programming and pricing of contingent claims in an incomplete market. SIAM J Control Optim 33(1):29–66
Funaro D (1992) Polynomial approximation of differential equations. Lecture notes in physics. Monographs 8. Springer, Berlin
Gavalas G (1968) Nonlinear differential equations of chemically reacting systems, vol 17. Springer tracts in natural philosophy. Springer, New York
Geiss C, Geiss S, Gobet E (2012) Generalized fractional smoothness and \(L_p\)-variation of BSDEs with non-Lipschitz terminal condition. Stoch Process Appl 122(5):2078–2116
Gobet E (2016) Monte-Carlo methods and stochastic processes: from linear to non-linear. CRC Press, Boca Raton
Gobet E, Lemor JP, Warin X (2005) A regression-based Monte-Carlo method to solve backward stochastic differential equations. Ann Appl Probab 15(3):2172–2202
Gobet E, López-Salas JG, Turkedjiev P, Vázquez C (2016) Stratified regression Monte-Carlo scheme for semilinear PDEs and BSDEs with large scale parallelization on GPUs. SIAM J Sci Comput 38(6):C652–C677
Gobet E, Turkedjiev P (2016) Approximation of backward stochastic differential equations using Malliavin weights and least-squares regression. Bernoulli 22(1):530–562
Gobet E, Turkedjiev P (2016) Linear regression MDP scheme for discrete backward stochastic differential equations under general conditions. Math Comput 85(299):1359–1391
Gobet E, Turkedjiev P (2017) Adaptive importance sampling in least-squares Monte-Carlo algorithms for backward stochastic differential equations. Stoch Process Appl 127(4):1171–1203
Henry D (1981) Geometric theory of semilinear parabolic equations, Lecture notes in mathematics, vol 840. Springer, Berlin
Henry-Labordere P, Tan X, Touzi N (2014) A numerical algorithm for a class of BSDEs via the branching process. Stoch Process Appl 124(2):1112–1140
Jackman S (2009) Bayesian analysis for the social sciences. Wiley Series in probability and statistics. Wiley, Sussex
Johnson W (2002) Combinatorics of higher derivatives of inverses. Am Math Mon 109(3):273–277
Johnson W (2002) The curious history of Faà di Bruno’s formula. Am Math Mon 109(3):217–234
Kloeden PE, Platen E (1992) Numerical solution of stochastic differential equations. Springer, Berlin
Krommer AR, Ueberhuber CW (1998) Comput Integr. SIAM, Philadelphia
L’Ecuyer P (1999) Good parameters and implementations for combined multiple recursive random number generators. Oper Res 47(1):159–164
Lee VW, Kim C, Chhugani J, Deisher M, Nguyen AD, Satish N, Smelyanskiy M, Chennupaty S, Hammarlund P, Singhal R, Dubey P (2010) Debunking the 100X GPU vs. CPU Myth: an evaluation of throughput computing on CPU and GPU. Newsl ACM SIGARCH Comput Archit News 10(38):451–460
Lemor JP, Gobet E, Warin X (2006) Rate of convergence of an empirical regression method for solving generalized backward stochastic differential equations. Bernoulli 12(5):889–916
Ma J, Protter P, Yong J (1994) Solving forward-backward stochastic differential equations explicitly—a four step scheme. Probab Theory Relat Fields 98(3):339–359
Menozzi S, Lemaire V (2010) On some non asymptotic bounds for the Euler scheme. Electron J Probab 15(53):1645–1681
Nevai P (1986) Géza Freud, orthogonal polynomials and Christoffel functions. A case study. J Approx Theory 48(1):3–167
Nvidia Coorporation (2014) NVIDIA's Next Generation CUDA Compute Architecture: Kepler GK110/210. http://www.nvidia.com/content/PDF/kepler/NVIDIA-Kepler-GK110-GK210-Architecture-Whitepaper.pdf
OpenMP Architecture Review Board (2018) OpenMP Application Program Interface Version 5.0. https://www.openmp.org/wp-content/uploads/OpenMP-API-Specification-5.0.pdf
Pardoux E, Peng S (1990) Adapted solution of a backward stochastic differential equation. Syst Control Lett 14(1):55–61
Pardoux E, Peng S (1992) Backward stochastic differential equations and quasilinear parabolic partial differential equations. In: Stochastic partial differential equations and their applications. Lecture notes in control and information sciences, vol 176, pp 200–217. Springer
Pardoux E, Rascanu A (2014) Stochastic differential equations, backward SDEs, partial differential equations, vol 69. Stochastic modelling and applied probability. Springer, Berlin
Peng S (2003) Nonlinear expectations, nonlinear evaluations and risk measures. In: Lectures presented in CIME EMS Bressanone, Italy, July 6 13, 2003
Pham H (2009) Continuous-time stochastic control and optimization with financial applications, vol 61. Series stochastic modeling and applied probability. Springer, Berlin
Shigesada N, Kawasaki K (1997) Biological invasions : theory and practice. Oxford series in ecology and evolution. Oxford University Press, Oxford
Smoller J (1994) Shock waves and reaction-diffusion equations, vol 258, 2nd edn. Fundamental principles of mathematical sciences. Springer, New York
Soner HM, Touzi N, Zhang J (2012) Well-posedness of second order backward SDEs. Probab Theory Relat Fields 153(1–2):149–190
NVIDIA Thrust. http://developer.nvidia.com/thrust
NVIDIA cuRAND. http://developer.nvidia.com/curand
Visintin A (1996) Models of phase transitions, vol 28. Progress in nonlinear differential equations and their applications. Birkhäuser Boston Inc., Boston
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of Interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The first author research is part of the Finance for Energy Markets (FiME) lab, of the Chair Financial Risks of the Risk Foundation and of the ANR project CAESARS (ANR-15-CE05-0024). The second author has been financially supported by the Chair Financial Risks of the Risk Foundation, the Spanish Grant MTM2016-76497-R and the Xunta de Galicia 2018 postdoctoral grant. The third author was partially supported by Spanish Grant MTM2016-76497-R.
Appendix A: Technical Estimates on Student’s t-Distribution
Appendix A: Technical Estimates on Student’s t-Distribution
In this appendix, estimates on the CDF of the Student’s t-distribution, its inverse and the derivatives of the inverse of the CDF are computed. As we have been doing in the article, we use the same notation C for any generic constants, we will keep the same C from line to line, in order the alleviate the computations, although its value changes. In this appendix \(\nu _l\) stands for Student’s t-distribution density (2.11) with parameter \(\mu \), \(F_{\nu _{l}}\) its marginal CDF and \(F^{-1}_{\nu }\) its inverse.
Lemma 11
(Estimates on \(F_{\nu _{l}}\) and \(F^{-1}_{\nu _{l}}\)) Let\(x\in {\mathbb {R}}\), \(u\in (0,1)\). The following estimates hold
with\({\tilde{c}}_\mu = \left( \dfrac{c_\mu }{\mu }\right) ^{1/\mu }.\)
Proof
In view of (2.11) we know that \(\nu _{l}(x_l)\sim _{x_l\rightarrow \pm \infty }\dfrac{c_\mu }{|x_l|^{\mu +1}}\). Besides,
The distribution \(\nu _l\) being symmetric, we have \(F_{\nu _{l}}(x)+F_{\nu _{l}}(-x)=1\) and \(F^{-1}_{\nu _{l}}(u)=-F^{-1}_{\nu _{l}}(1-u)\). Therefore, the estimate \(1- F_{\nu _{l}}(x)\) for \(x\rightarrow +\infty \) follows from the case \(x\rightarrow -\infty \).
In order to compute \(F^{-1}_{\nu _{l}}\) we use that \(F_{\nu _{l}}(F^{-1}_{\nu _{l}}(u))=u\):
If \(u\rightarrow 0^+\), \(F^{-1}_{\nu _{l}}(u)\rightarrow -\infty \), \(u = F_{\nu _{l}}(F^{-1}_{\nu _{l}}(u)) \sim _{u\rightarrow 0^+} \dfrac{c_\mu }{\mu |F^{-1}_{\nu _{l}}(u)|^\mu }\). Therefore,
which is the advertised result. The case \(u\rightarrow 1^-\) follows from the case for \(u\rightarrow 0^+\) using again the symmetry of the distribution \(\nu _l\). \(\square \)
Lemma 12
(Estimates on derivatives of \(\frac{1}{(1+|\mathbf {x}|^2)^{q/2}}\)) Let\(\mathbf {x}\in {\mathbb {R}}^d\), \(q\in {\mathbb {R}}^+\)and\(n\in {\mathbb {N}}\). The following upper bound holds
Proof
We only have to consider the case \(n\ge 1\). In order to compute \(\partial ^n_{x_l} \dfrac{1}{(1+|\mathbf {x}|^2)^{q/2}}\) we use the following Faà di Bruno’s formula (as long as the partial derivative is only with respect to one variable we can extend in the following way the one-dimensional Faà di Bruno’s formula (2.34)):
Here, consider \(f(u)=u^{-q/2}\) and \(g(\mathbf {x})=1+|\mathbf {x}|^2\): \(\mathrm{d}^m_u f(u) = C_{m,q} u^{-q/2-m}\) and \(\partial _{x_l}^{j} g(\mathbf {x})=2 x_l{\mathbf {1}}_{j=1}+2{\mathbf {1}}_{j=2}\). Observe that in the Faà di Bruno’s formula (A.3), the sum over \(\mathbf {j}\in J_{m,n}\) will be made only for \(j_i=1\) with \(m_1:=\#\{j_i:\mathbf {j}\in J_{m,n}, j_i=1\}\) terms, and for \(j_i=2\) with \(m_2:=\#\{j_i:\mathbf {j}\in J_{m,n}, j_i=2\}\) terms. Owing to this property and by definition of \(J_{m,n}\), we have \(m_1+2m_2=n\) and \(m_1+m_2=m\), which results in \(m_1=2m-n\) and \(m_2=n-m\). Invoking the form of the derivatives \(\partial _{x_l}^{j_i} g(\mathbf {x})\), it readily follows that
All in all, we deduce
\(\square \)
Lemma 13
(Estimates on derivatives of \(\frac{1}{(1+|\mathbf {x}|^2)^{q/2}}\)) Let\(\mathbf {x}\in {\mathbb {R}}^d\), \(q\in {\mathbb {R}}^+\), \(\mathbf {n} = (n_1,\ldots ,n_d) \in {\mathbb {N}}_+^d\). The following upper bound holds
Proof
We only have to consider the case \(\overline{\mathbf {n}}\ge 1\). For such \(\mathbf {n}\), we use the following multivariate Faà di Bruno’s formula derived from [15, Corollary 2.10]
where the positive constants \(C_{m,J}\) depend on m and the matrix J under consideration.
Note that non coordinate-wise derivatives of \(g(\mathbf {x})=1+|\mathbf {x}|^2\) are all zero. Besides, coordinate-wise derivatives of g are \(\partial _{x_l}^{j} g(\mathbf {x})=2 x_l{\mathbf {1}}_{j=1}+2{\mathbf {1}}_{j=2}\). For a given J, let \(a_{l} \in {\mathbb {N}}\, \forall l=1,\ldots ,d\) denote the number of vectors in J with 1 in the l-th coordinate and zero anywhere else. It holds that
for \(2m\ge \overline{\mathbf {n}}\). All in all, we deduce
\(\square \)
Lemma 14
(Estimates on derivatives of Student’s t-distribution) Let\(n\in {\mathbb {N}}\). The following upper bound holds
Proof
This result follows readily using the previous estimate (A.2). \(\square \)
Lemma 15
(Estimates on derivatives of \(F^{-1}_{\nu _{l}}\)) Let\(n\in {\mathbb {N}}\). The following upper bounds hold
Cis a constant depending on\(\mu \)andn.
Proof
Using the combinatorial formula for higher derivatives of inverses [40] and the fact that for \(b\in {\mathbb {N}}\), \(b\ge 1\), \(\mathrm{d}^b_{x}F_{\nu _{l}}(x) = \mathrm{d}_x^{b-1}\nu _l(x)\),
Therefore, using (A.6),
Finally, combining the previous upper bound with (A.1) yields
where the constant C may change value along computations. \(\square \)
Rights and permissions
About this article
Cite this article
Gobet, E., López-Salas, J.G. & Vázquez, C. Quasi-Regression Monte-Carlo Scheme for Semi-Linear PDEs and BSDEs with Large Scale Parallelization on GPUs. Arch Computat Methods Eng 27, 889–921 (2020). https://doi.org/10.1007/s11831-019-09335-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11831-019-09335-x