pith. sign in

arxiv: 2403.11910 · v3 · submitted 2024-03-18 · 🧮 math.NA · cs.NA· math.OC· math.PR

Numerical method for nonlinear Kolmogorov PDEs via sensitivity analysis

Daniel Bartl , Ariel Neufeld , Kyunghyun Park This is my paper

Pith reviewed 2026-05-24 03:18 UTC · model grok-4.3

classification 🧮 math.NA cs.NAmath.OCmath.PR
keywords nonlinear Kolmogorov PDEsensitivity analysisMonte Carlo methodFeynman-Kac representationhigh-dimensional PDEHamiltonian nonlinearitynumerical approximation
0
0 comments X

The pith

Nonlinear Kolmogorov PDEs equal a baseline linear PDE plus ε times a second linear PDE as the neighborhood size shrinks to zero.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes a first-order sensitivity expansion for nonlinear Kolmogorov PDEs whose Hamiltonian is the pointwise supremum over an ε-neighborhood of given baseline drift and diffusion coefficients. It shows that the nonlinear equation reduces to the linear Kolmogorov PDE driven by those baseline coefficients plus an ε-scaled correction term whose value function satisfies a second linear Kolmogorov PDE. Because both linear equations admit Feynman-Kac representations, the expansion supplies a Monte Carlo scheme that remains tractable in high dimension. The authors supply error bounds and complexity estimates and illustrate the scheme numerically in dimensions up to 100.

Core claim

As ε tends to zero the solution of the nonlinear Kolmogorov PDE equals the solution of the linear Kolmogorov PDE with the baseline coefficients plus ε times the solution of an auxiliary linear Kolmogorov PDE whose coefficients are again built from the baseline process; the two linear equations are solved by Monte Carlo sampling of the associated diffusion.

What carries the argument

First-order sensitivity expansion of the nonlinear PDE value function with respect to the neighborhood radius ε, obtained by solving one auxiliary linear Kolmogorov PDE.

If this is right

  • The nonlinear PDE is approximated to order ε by solving two linear Kolmogorov equations.
  • Feynman-Kac Monte Carlo sampling yields a dimension-independent numerical scheme.
  • Error and complexity bounds follow from standard estimates on the two linear PDEs.
  • The method is demonstrated to remain stable in dimensions up to 100.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same expansion technique could be applied to other small perturbations of linear generators beyond the ε-neighborhood construction.
  • The two-step Monte Carlo procedure may be reused for related nonlinear problems in stochastic control whenever the nonlinearity admits a comparable first-order characterization.
  • Higher-order terms in the expansion could be derived by differentiating the auxiliary linear PDE again with respect to ε.

Load-bearing premise

The nonlinearity must arise exactly as the pointwise supremum of the generator over the ε-neighborhood of the baseline coefficients.

What would settle it

Compute the true nonlinear PDE solution for successively smaller ε and check whether the difference from the baseline linear solution divided by ε converges to the auxiliary linear solution.

Figures

Figures reproduced from arXiv: 2403.11910 by Ariel Neufeld, Daniel Bartl, Kyunghyun Park.

Figure 1
Figure 1. Figure 1: Comparative analysis between the approximated solution v 0 + ε · ∂εv 0 and the actual counterpart v ε over varying ε. 10 20 30 40 50 60 70 80 90 100 N 14.000 14.025 14.050 14.075 14.100 14.125 14.150 14.175 14.200 v 0 + ε · ∂ε v 0 v 0 + ε · ∂εv 0 (95% C.I.) v 0 + ε · ∂εv 0 (Avg.) 0.1 0.2 0.3 0.4 0.5 L 2-error L 2 -error (a) Stability w.r.t. N in [10, . . . , 100]. 0.5 1.0 1.5 2.0 M0 ×106 14.04 14.05 14.06 … view at source ↗
Figure 2
Figure 2. Figure 2: Stability analysis with respect to the model parameters N, M0, M1, M2 in Algorithm 1. We fix ε = 0.1 and γ = η = 1, and consider the basis parameters N = 100, M0 = 2.4 × 106 and M1 = M2 = 2.4 × 104 , as specified in Section 3.1. In each plot, all but one parameters are fixed (e.g. in the left hand image, M0, M1, M2 are fixed and N is varied). The plots features blue lines to represent the average of estima… view at source ↗
Figure 3
Figure 3. Figure 3: The running maximum for the error between the approximation of v 0 and ∂εv 0 across k runs with k ∈ {1, . . . , 1000}, where the benchmark for these errors is the average value for the scaled solution for d = 1 according to (3.2)-(3.4) across the 1,000 runs. (b o , σo ) are generated randomly for every run and every d. function of k ≤ K that shows that maximum made error of the estimator over all j ≤ k inp… view at source ↗
read the original abstract

We examine nonlinear Kolmogorov partial differential equations (PDEs). Here the nonlinear part of the PDE comes from its Hamiltonian where one maximizes over all possible drift and diffusion coefficients which fall within a $\varepsilon$-neighborhood of pre-specified baseline coefficients. Our goal is to quantify and compute how sensitive those PDEs are to such a small nonlinearity, and then use the results to develop an efficient numerical method for their approximation. We show that as $\varepsilon\downarrow 0$, the nonlinear Kolmogorov PDE equals the linear Kolmogorov PDE defined with respect to the corresponding baseline coefficients plus $\varepsilon$ times a correction term which can be also characterized by the solution of another linear Kolmogorov PDE involving the baseline coefficients. As these linear Kolmogorov PDEs can be efficiently solved in high-dimensions by exploiting their Feynman-Kac representation, our derived sensitivity analysis then provides a Monte Carlo based numerical method which can efficiently solve these nonlinear Kolmogorov equations. We establish an error and complexity analysis for our numerical method. Moreover, we provide numerical examples in up to 100 dimensions to empirically demonstrate the applicability of our numerical method.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper examines nonlinear Kolmogorov PDEs whose nonlinearity arises from a Hamiltonian that takes the pointwise supremum of the generator over an ε-neighborhood of given baseline drift and diffusion coefficients. It derives an asymptotic expansion showing that, as ε↓0, the nonlinear solution equals the solution of the corresponding linear Kolmogorov PDE plus ε times a correction term that itself solves a linear Kolmogorov PDE driven by the directional derivative of the Hamiltonian. This expansion is then used to construct a Monte Carlo numerical method based on Feynman-Kac representations of the two linear PDEs; error and complexity bounds are established for the method, and numerical experiments are reported in dimensions up to 100.

Significance. If the expansion and its numerical realization hold under the stated assumptions, the work supplies a practical route to high-dimensional nonlinear Kolmogorov equations by reducing them to two linear problems that admit efficient Monte Carlo solution. The combination of a parameter-free first-order expansion, explicit error/complexity analysis, and empirical tests in 100 dimensions constitutes a concrete contribution to the numerical analysis of high-dimensional nonlinear PDEs.

major comments (2)
  1. [§3] §3 (derivation of the first-order term): the envelope differentiation step that produces the inhomogeneous term for the correction PDE v requires explicit verification that the directional derivative of the Hamiltonian exists and is continuous at the baseline coefficients; the manuscript should state the precise regularity (e.g., Lipschitz or C^1) assumed on the coefficients to justify differentiation under the supremum.
  2. [§5] §5 (numerical error tables): the complexity analysis claims an overall cost of order ε^{-2} or better, yet the reported experiments do not include a table that juxtaposes observed L^2 errors against the predicted rates for successive values of ε and dimension; without such a table the empirical support for the complexity claim remains incomplete.
minor comments (2)
  1. [Abstract] The abstract states that the correction term 'can be also characterized by the solution of another linear Kolmogorov PDE' but does not name the inhomogeneous source term; adding a one-line description of this source would improve readability.
  2. [Introduction] Notation for the baseline coefficients (a0,b0) and the ε-ball is introduced only in the problem formulation; repeating the definition once in the statement of the main theorem would aid readers who skim the introduction.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive evaluation and the constructive comments on our manuscript. We address each major comment below, indicating the revisions we plan to incorporate.

read point-by-point responses

  1. Referee: [§3] §3 (derivation of the first-order term): the envelope differentiation step that produces the inhomogeneous term for the correction PDE v requires explicit verification that the directional derivative of the Hamiltonian exists and is continuous at the baseline coefficients; the manuscript should state the precise regularity (e.g., Lipschitz or C^1) assumed on the coefficients to justify differentiation under the supremum.

    Authors: We agree that an explicit statement of the required regularity is needed to rigorously justify the envelope differentiation. In the revised manuscript we will insert a new remark immediately after the statement of the Hamiltonian in Section 3. The remark will assume that the baseline drift and diffusion coefficients are Lipschitz continuous in the state variable (uniformly in the control) and continuous in the control parameters, and will verify that these conditions guarantee the existence and continuity of the directional derivative of the Hamiltonian at the baseline point, thereby justifying differentiation under the supremum. revision: yes

  2. Referee: [§5] §5 (numerical error tables): the complexity analysis claims an overall cost of order ε^{-2} or better, yet the reported experiments do not include a table that juxtaposes observed L^2 errors against the predicted rates for successive values of ε and dimension; without such a table the empirical support for the complexity claim remains incomplete.

    Authors: We concur that a dedicated table would provide clearer empirical support for the complexity bounds. In the revised version we will add a new table in Section 5 that reports observed L^2 errors for successive values of ε (0.1, 0.05, 0.01) across dimensions 10, 50 and 100, together with the theoretically predicted rates, allowing direct comparison with the complexity analysis. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

The paper defines the nonlinear Hamiltonian explicitly as the pointwise supremum of the generator over an ε-neighborhood of baseline coefficients. The claimed first-order expansion as ε↓0 is obtained by standard directional differentiation of this definition, yielding a linear Kolmogorov PDE for the correction whose inhomogeneous term is the directional derivative of the Hamiltonian. This step uses only the well-posedness already assumed for the linear equations and does not invoke fitted parameters, self-citations for uniqueness theorems, or ansatzes smuggled from prior work. The Monte Carlo method follows directly from the Feynman-Kac representations of the resulting linear PDEs. No load-bearing step reduces to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The expansion implicitly assumes existence of the maximizing coefficients inside the ε-ball and sufficient regularity for differentiation under the expectation.

pith-pipeline@v0.9.0 · 5724 in / 1273 out tokens · 30101 ms · 2026-05-24T03:18:50.966873+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Robust SGLD algorithm for solving non-convex distributionally robust optimisation problems

    math.OC 2024-03 unverdicted novelty 5.0

    Develops robust SGLD with non-asymptotic convergence bounds for non-convex DRO and applies it to neural network regression under adversarial corruption.

Reference graph

Works this paper leans on

66 extracted references · 66 canonical work pages · cited by 1 Pith paper

  1. [1]

    G. Barles. Convergence of numerical schemes for degenerate parabolic equations arising in finance theory. In Rogers L. C. G., Talay D. eds. Numerical Methods in Finance

    work page

  2. [2]

    Barles, R

    G. Barles, R. Buckdahn, and E. Pardoux. Backward stochastic differential equations and integral-partial differential equations.Stochastics, 60(1-2):57–83, 1997

    work page 1997

  3. [3]

    Bartl, S

    D. Bartl, S. Drapeau, J. Obłój, and J. Wiesel. Sensitivity analysis of Wasserstein distributionally robust optimization problems.Proc. R. Soc. A, 477(2256):20210176, 2021

    work page 2021

  4. [4]

    Bartl, A

    D. Bartl, A. Neufeld, and K. Park. Sensitivity of robust optimization problems under drift and volatility uncertainty.arXiv preprint arXiv:2311.11248, 2023

    work page arXiv 2023

  5. [5]

    Bartl and J

    D. Bartl and J. Wiesel. Sensitivity of multiperiod optimization problems with respect to the adapted Wasserstein distance.SIAM J. Financial Math., 14(2):704–720, 2023

    work page 2023

  6. [6]

    C. Beck, S. Becker, P. Cheridito, A. Jentzen, and A. Neufeld. Deep splitting method for parabolic PDEs. SIAM J. Sci. Comput, 43(5):A3135–A3154, 2021

    work page 2021

  7. [7]

    Black and M

    F. Black and M. Scholes. The pricing of options and corporate liabilities.J. Polit. Econ., 81(3):637–654, 1973

    work page 1973

  8. [8]

    Blanchet and K

    J. Blanchet and K. Murthy. Quantifying distributional model risk via optimal transport.Math. Oper. Res., 44(2):565–600, 2019

    work page 2019

  9. [9]

    V. I. Bogachev, N. V. Krylov, M. Röckner, and S. V. Shaposhnikov.Fokker–Planck–Kolmogorov Equations, volume 207 of Mathematical Surveys and Monographs. American Mathematical Society, 2022

    work page 2022

  10. [10]

    Cerrai.Second Order PDE’s in Finite and Infinite Dimension: A Probabilistic Approach

    S. Cerrai.Second Order PDE’s in Finite and Infinite Dimension: A Probabilistic Approach. Lecture Notes in Math.Springer, Berlin, 2001

    work page 2001

  11. [11]

    Cheridito, H

    P. Cheridito, H. M. Soner, N. Touzi, and N. Victoir. Second-order backward stochastic differential equations and fully nonlinear parabolic PDEs.Comm. Pure Appl. Math., 60(7):1081–1110, 2007

    work page 2007

  12. [12]

    Coquet, Y

    F. Coquet, Y. Hu, J. Mémin, and S. Peng. Filtration-consistent nonlinear expectations and related g- expectations. Probab. Theory Relat. Fields, 123(1):1–27, 2002

    work page 2002

  13. [13]

    M. G. Crandall, H. Ishii, and P.-L. Lions. User’s guide to viscosity solutions of second order partial differ- ential equations.Bull. Amer. Math. Soc., 27(1):1–67, 1992

    work page 1992

  14. [14]

    Dellacherie and P

    C. Dellacherie and P. Meyer. Probabilities and Potential, B: Theory of Martingales

    work page

  15. [15]

    Denis, M

    L. Denis, M. Hu, and S. Peng. Function spaces and capacity related to a sublinear expectation: application to G-Brownian motion paths.Potential Anal., 34:139–161, 2011

    work page 2011

  16. [16]

    Dolinsky, M

    Y. Dolinsky, M. Nutz, and H. M. Soner. Weak approximation ofG-expectations. Stoch. Process. Appl., 122(2):664–675, 2012

    work page 2012

  17. [17]

    El Karoui and X

    N. El Karoui and X. Tan. Capacities, measurable selection and dynamic programming part ii: application in stochastic control problems.arXiv preprint arXiv:1310.3364, 2013

    work page arXiv 2013

  18. [18]

    L. C. Evans.Partial Differential Equations, volume 19. American Mathematical Society, 2010

    work page 2010

  19. [19]

    Fadina, A

    T. Fadina, A. Neufeld, and T. Schmidt. Affine processes under parameter uncertainty.Probab. Uncertain. Quant. Risk, 4:1–35, 2019

    work page 2019

  20. [20]

    W. H. Fleming and H. M. Soner.Controlled Markov processes and viscosity solutions, volume 25. Springer Science & Business Media, 2006

    work page 2006

  21. [21]

    Friedman

    A. Friedman. Stochastic differential equations and applications. InStochastic differential equations, pages 75–148. Springer, 1975

    work page 1975

  22. [22]

    Fuhrmann, M

    S. Fuhrmann, M. Kupper, and M. Nendel. Wasserstein perturbations of Markovian transition semigroups. Ann. Inst. Henri Poincaré Probab. Stat., 59(2):904–932, 2023

    work page 2023

  23. [23]

    R. Gao, X. Chen, and A. J. Kleywegt. Wasserstein distributionally robust optimization and variation regularization. Oper. Res., 2022

    work page 2022

  24. [24]

    Gao and A

    R. Gao and A. Kleywegt. Distributionally robust stochastic optimization with wasserstein distance.Math. Oper. Res., 48(2):603–655, 2023

    work page 2023

  25. [25]

    M. B. Giles. Multilevel Monte Carlo path simulation.Oper. Res., 56(3):607–617, 2008

    work page 2008

  26. [26]

    E. Gobet. Monte-Carlo methods and stochastic processes: from linear to non-linear. Chapman and Hal- l/CRC, 2016

    work page 2016

  27. [27]

    Graham and D

    C. Graham and D. Talay.Stochastic simulation and Monte Carlo methods: mathematical foundations of stochastic simulation, volume 68. Springer Science & Business Media, 2013. 30 DANIEL BARTL, ARIEL NEUFELD, AND KYUNGHYUN PARK

    work page 2013

  28. [28]

    J. Han, A. Jentzen, and W. E. Solving high-dimensional partial differential equations using deep learning. Proc. Natl. Acad. Sci., 115(34):8505–8510, 2018

    work page 2018

  29. [29]

    Hasselmann

    K. Hasselmann. Stochastic climate models part I. Theory.Tellus, 28(6):473–485, 1976

    work page 1976

  30. [30]

    Herrmann and J

    S. Herrmann and J. Muhle-Karbe. Model uncertainty, recalibration, and the emergence of delta–vega hedg- ing. Finance Stoch., 21:873–930, 2017

    work page 2017

  31. [31]

    Herrmann, J

    S. Herrmann, J. Muhle-Karbe, and F. T. Seifried. Hedging with small uncertainty aversion.Finance Stoch., 21:1–64, 2017

    work page 2017

  32. [32]

    C. Huré, H. Pham, and X. Warin. Deep backward schemes for high-dimensional nonlinear PDEs.Math. Comp., 89(324):1547–1579, 2020

    work page 2020

  33. [33]

    Hutzenthaler, A

    M. Hutzenthaler, A. Jentzen, and T. Kruse. On multilevel Picard numerical approximations for high- dimensional nonlinear parabolic partial differential equations and high-dimensional nonlinear backward stochastic differential equations.J. Sci. Comput., 79(3):1534–1571, 2019

    work page 2019

  34. [34]

    Y. Jiang. Duality of causal distributionally robust optimization: the discrete-time case.arXiv preprint arXiv:2401.16556, 2024

    work page arXiv 2024

  35. [35]

    N. G. V. Kampen.Stochastic processes in physics and chemistry, volume 1. North-Holland, Oxford, 1981

    work page 1981

  36. [36]

    Karatzas and S

    I. Karatzas and S. Shreve.Brownian motion and stochastic calculus, volume 113. Springer, 1991

    work page 1991

  37. [37]

    M. Kimura. Some problems of stochastic processes in genetics.Ann. Math. Stat., 28:882–901, 1957

    work page 1957

  38. [38]

    N. V. Krylov. On Kolmogorov’s equations for finite-dimensional diffusions. In Stochastic PDE’s and Kol- mogorov Equations in Infinite Dimensions (Cetraro, 1998). Lecture Notes in Math.1715, 1–63. Springer, Berlin, 1999

    work page 1998

  39. [39]

    J. D. Logan and W. Wolesensky.Mathematical methods in biology, volume 96. John Wiley & Sons, 2009

    work page 2009

  40. [40]

    Matoussi, D

    A. Matoussi, D. Possamaï, and C. Zhou. Robust utility maximization in nondominated models with 2BSDE: the uncertain volatility model.Math. Finance, 25(2):258–287, 2015

    work page 2015

  41. [41]

    Mohajerin Esfahani and D

    P. Mohajerin Esfahani and D. Kuhn. Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations.Math. Programming, 171(1-2):115–166, 2018

    work page 2018

  42. [42]

    Nendel and A

    M. Nendel and A. Sgarabottolo. A parametric approach to the estimation of convex risk functionals based on Wasserstein distance.arXiv preprint arXiv:2210.14340, 2022

    work page arXiv 2022

  43. [43]

    Neufeld and M

    A. Neufeld and M. Nutz. Measurability of semimartingale characteristics with respect to the probability law. Stoch. Process. Appl., 124(11):3819–3845, 2014

    work page 2014

  44. [44]

    Neufeld and M

    A. Neufeld and M. Nutz. Nonlinear Lévy processes and their characteristics. Trans. Am. Math. Soc., 369(1):69–95, 2017

    work page 2017

  45. [45]

    Neufeld and M

    A. Neufeld and M. Nutz. Robust utility maximization with Lévy processes.Math. Finance, 28(1):82–105, 2018

    work page 2018

  46. [46]

    M. Nutz. RandomG-expectations. Ann. Appl. Probab., 23(5):1755–1777, 2013

    work page 2013

  47. [47]

    Nutz and R

    M. Nutz and R. van Handel. Constructing sublinear expectations on path space.Stoch. Process. Appl., 123(8):3100–3121, 2013

    work page 2013

  48. [48]

    Obłój and J

    J. Obłój and J. Wiesel. Distributionally robust portfolio maximization and marginal utility pricing in one period financial markets.Math. Finance, 31(4):1454–1493, 2021

    work page 2021

  49. [49]

    Pascucci

    A. Pascucci. Kolmogorov equations in physics and in finance. Elliptic and parabolic problems.Prog. Non- linear Differ. Equ. Appl., 63:353–364, 2005

    work page 2005

  50. [50]

    S. Peng. BSDE and relatedg-expectations. InBackward Stochastic Differential Equations, ed. N. El Karoui and L. Mazliak. pages 141–159. Longman, Harlow, 1997

    work page 1997

  51. [51]

    S. Peng. G-expectation, G-Brownian motion and related stochastic calculus of Itô type. In Stochastic Analysis and Applications, Abel Symp., pages 541–567. Springer, Berlin, Heidelberg, 2007

    work page 2007

  52. [52]

    Pflug and D

    G. Pflug and D. Wozabal. Ambiguity in portfolio selection.Quant. Finance, 7(4):435–442, 2007

    work page 2007

  53. [53]

    H. Pham. Continuous-time stochastic control and optimization with financial applications, volume 61. Springer Science & Business Media, 2009

    work page 2009

  54. [54]

    Possamaï, X

    D. Possamaï, X. Tan, and C. Zhou. Stochastic control for a class of nonlinear kernels and applications. Ann. Probab., 46(1):551–603, 2018

    work page 2018

  55. [55]

    P. E. Protter.Stochastic Integration and Differential Equations, volume 21. Springer Science & Business Media, 2005

    work page 2005

  56. [56]

    Sirignano and K

    J. Sirignano and K. Spiliopoulos. DGM: A deep learning algorithm for solving partial differential equations. J. Comput. Phys., 375:1339–1364, 2018. 31

    work page 2018

  57. [57]

    S. A. Smolyak. Quadrature and interpolation formulas for tensor products of certain classes of functions. In Doklady Akademii Nauk, volume 148, pages 1042–1045. Russian Academy of Sciences, 1963

    work page 1963

  58. [58]

    H. M. Soner, N. Touzi, and J. Zhang. Quasi-sure stochastic analysis through aggregation.Electron. J. Probab., 16:1844–1879, 2011

    work page 2011

  59. [59]

    H. M. Soner, N. Touzi, and J. Zhang. Wellposedness of second order backward SDEs.Probab. Theory Related Fields, 153(1-2):149–190, 2012

    work page 2012

  60. [60]

    H. M. Soner, N. Touzi, and J. Zhang. Dual formulation of second order target problems.Ann. Appl. Probab., 23(1):308–347, 2013

    work page 2013

  61. [61]

    E. Tadmor. A review of numerical methods for nonlinear partial differential equations.Bull. Amer. Math. Soc., 49(4):507–554, 2012

    work page 2012

  62. [62]

    Thomée.Galerkin finite element methods for parabolic problems, volume 25

    V. Thomée.Galerkin finite element methods for parabolic problems, volume 25. Springer Science & Business Media, 2007

    work page 2007

  63. [63]

    J. Thuburn. Climate sensitivities via a Fokker–Planck adjoint approach. Q. J. R. Meteorol. Soc., 131(605):73–92, 2005

    work page 2005

  64. [64]

    D. V. Widder.The heat equation. Academic Press, New York, 1976

    work page 1976

  65. [65]

    Wilmott, J

    P. Wilmott, J. Dewynne, and S. Howison.Option Pricing: Mathematical Models and Computation. Oxford Financial Press, 1993

    work page 1993

  66. [66]

    Yong and X

    J. Yong and X. Y. Zhou.Stochastic controls: Hamiltonian systems and HJB equations, volume 43. Springer Science & Business Media, 2012. F aculty of Mathematics, University of Vienna Email address: daniel.bartl@univie.ac.at Division of Mathematical Sciences, Nanyang Technological University Email address: ariel.neufeld@ntu.edu.sg Division of Mathematical Sc...

    work page 2012