Skip to main content Accessibility help

A Comparison of Higher-Order Weak Numerical Schemes for Stopped Stochastic Differential Equations

  • Francisco Bernal (a1) (a2) and Juan A. Acebrón (a1) (a3)


We review, implement, and compare numerical integration schemes for spatially bounded diffusions stopped at the boundary which possess a convergence rate of the discretization error with respect to the timestep h higher than . We address specific implementation issues of the most general-purpose of such schemes. They have been coded into a single Matlab program and compared, according to their accuracy and computational cost, on a wide range of problems in up to ℝ48. The paper is self-contained and the code will be made freely downloadable.


Corresponding author

*Corresponding author. Email addresses: (F. Bernal), (J. A. Acebrón)


Hide All
[1] Acebrón, J.A., Busico, M.P., Lanucara, P., and Spigler, R., Domain decomposition solution of elliptic boundary-value problems via Monte Carlo and quasi-Monte Carlo methods. SIAM J. Sci. Comp. 27(2), 440457 (2005).
[2] Baldi, P., Exact asymptotics for the probability of exit from a domain and applications to simulation. Ann. Probab. 23(4), 16441670 (1995).
[3] Bayer, C., Szepessy, A., and Tempone, R., Adaptive weak approximation of reflected and stopped diffusions. Monte Carlo Methods and Applications 16, 167 (2010).
[4] Bernal, F., Acebrón, J.A., and Anjam, I., A stochastic algorithm based on fast marching for automatic capacitance extraction in non-Manhattan geometries. SIAM J. Imag. Sci. 7(4), 26572674 (2014).
[5] Buchmann, F.M. and Petersen, W.P., An exit probability approach to solving high dimensional Dirichlet problems. SIAM J. Sci. Comput. 28, 11531166 (2006).
[6] Buchmann, F.M., Simulation of stopped diffusions. J. Comput. Phys. 202(2), 446462 (2005).
[7] Constantini, C., Pacchiarotti, B., and Sartoretto, F., Numerical approximation for functionals of reflecting diffusion processes. SIAM J. Appl. Math. 58, 73102 (1998).
[8] Deaconu, M. and Lejay, A., A Random Walk on Rectangles algorithm. Methodol. Comput. Appl. Probab. 8, 135151 (2006).
[9] Freidlin, M., Functional Integration and Partial Differential Equations. Princeton University Press (1985).
[10] Giles, M.B., Multi-levelMonte Carlo path simulation. Operations Research, 56(3), 607617 (2008).
[11] Giles, M.B. and Bernal, F., Multilevel simulations of expected exit times and other functionals of stopped diffusions. In preparation. (Unpublished)
[12] Giraudo, M.T. and Sacerdote, L., An improved technique for the simulation of first passage times for diffusion processes. Comm. Statist. Simulation Comput. 28(4), 11351163 (1999).
[13] Glasserman, P., Monte Carlo Methods in Financial Engineering. Springer (2003).
[14] Gobet, E. and Menozzi, S., Stopped diffusion processes: Overshoots and boundary correction. Stoch. Proc. Appl., 120, 130162, (2010).
[15] Gobet, E., Euler schemes for the weak approximation of killed diffusion. Stoch. Process. Appl. 87, 167197 (2000).
[16] Gobet, E., Euler schemes and half-space approximation for the simulation of diffusions in a domain. ESAIM: Probab. Statist. 5, 261297 (2001).
[17] Higham, D.J., Mao, X., Roj, M., Song, Q., and Yin, G.. Mean exit times and the multilevel Monte Carlo method. SIAM J. Uncertainty Quantification, 1(1), 218 (2013).
[18] Iyer, G., Novikov, A., Ryzhik, L., and Zlatos, A., Exit times for diffusions with incompressible drift. SIAM J. Math. Anal. 42, 24842498 (2010).
[19] Jansons, K.M. and Lythe, G.D., Efficient numerical solution of stochastic differential equations using exponential timestepping. J. Statist. Phys. 100(5-6), 10971109 (2000).
[20] Kloeden, P.E. and Platten, E., Numerical Solution of Stochastic Differential Equations. Springer (1999).
[21] Lemaire, P. and Pagés, G., Multilevel Richardson-Romberg extrapolation. To appear in Bernoulli, (2015).
[22] Mairé, S. and Tanré, E., Some new simulation schemes for the evaluation of Feynman-Kac representations. Monte Carlo Methods and Applications, 14(1), 2951 (2008).
[23] Mairé, S. and Simon, M., A partially reflecting random walk on spheres algorithm for electrical impedance tomography. Preprint in arXiv:1502.04318 (2015). (Unpublished)
[24] Mancini, S., Bernal, F., and Acebrón, J.A., An efficient algorithm for accelerating Monte Carlo approximations of the solution to boundary value problems. J. Sci. Comput. 66(2), 577597 (2016).
[25] Mannella, R., Absorbing boundaries and optimal stopping in a stochastic differential equation. Phys. Lett. A 254(5), 257262 (1999).
[26] Mascagni, M. and Simonov, N.A., Monte Carlo methods for calculating some physical properties of large molecules, SIAM J. Sci. Comput. 26, 339357 (2004).
[27] Michael, J.R., Schucany, W.R., and Haas, R.W., Generating random variates using transformations with multiple roots. Am. Statist. 30, 8890 (1976).
[28] Milstein, G.N., Weak approximation of a diffusion process in a bounded domain. Stoch. Stoch. Rep. 62, 147200 (1997).
[29] Milstein, G.N. and Tretyakov, M.V., Stochastic Numerics for Mathematical Physics. Springer, Berlin (2004).
[30] Milstein, G.N. and Tretyakov, M.V., The simplest random walks for the Dirichlet problem. Teor. Veroyatnost. I Primenen. 47(1), 3958 (2002).
[31] Miranda, C., Partial Differential Equations of Elliptic Type. Springer (1970).
[32] Muller, M.E., Some continuous Monte Carlo methods for the Dirichlet problem. Ann. Math. Statist. 27, 569589 (1956).
[33] Redner, S., A Guide to First-Passage Processes. Cambridge University Press (2001).
[34] Schwabedal, J.T.C. and Pikovsky, A., Phase description of stochastic oscillations, Phys. Rev. Lett. 110(20), 204102 (2013).
[35] Tamborrino, M., Sacerdote, L., and Jacobsen, M., Weak convergence of marked point processes generated by crossings of multivariate jump processes. Application to neural network modeling. Physica D, 288, 4552 (2014).
[36] Talay, D. and Tubaro, L., Expansion of the global error for numerical schemes solving stochastic differential equations. Stoch. Anal. App., 8(4), 94120 (1990).


MSC classification


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed