Abstract
In this paper, we present a novel approach towards variance reduction for discretised diffusion processes. The proposed approach involves specially constructed control variates and allows for a significant reduction in the variance for the terminal functionals. In this way, the complexity order of the standard Monte Carlo algorithm (\(\varepsilon ^{-3}\)) can be reduced down to \(\varepsilon ^{-2}\sqrt{\left| \log (\varepsilon )\right| }\) in case of the Euler scheme with \(\varepsilon \) being the precision to be achieved. These theoretical results are illustrated by several numerical examples.
The study has been funded by the Russian Academic Excellence Project ‘5–100’.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Performing the full complexity analysis via Lagrange multipliers one can see that these parameter values are not optimal if \(2(p+1)\le d\) or \(\nu \le \frac{2d(p+1)}{2(p+1)-d}\) (a Lagrange multiplier corresponding to a ‘\(\le 0\)’ constraint is negative, cf. proof of Theorem 8). Therefore, the recommendation is to choose the power p for our basis functions according to \(p>\frac{d-2}{2}\). The opposite choice is allowed as well (the method converges), but theoretical complexity of the method would be then worse than that of the SMC, namely, \(\varepsilon ^{-3}\).
- 2.
For the Euler scheme, there is an additional logarithmic factor in the complexity of the MLMC algorithm (see [3]).
- 3.
Compare with footnote 1 on p. 16.
- 4.
References
Akahori, J., Amaba, T., Okuma, K.: A discrete-time Clark-Ocone formula and its application to an error analysis (2013). arXiv:1307.0673v2
Belomestny, D., Häfner, S., Nagapetyan, T., Urusov, M.: Variance reduction for discretised diffusions via regression (2016). arXiv:1510.03141v3
Giles, M.B.: Multilevel Monte Carlo path simulation. Op. Res. 56(3), 607–617 (2008)
Györfi, L., Kohler, M., Krzyżak, A., Walk, H.: A distribution-free theory of nonparametric regression. Springer Series in Statistics. Springer, New York (2002)
Karatzas, I., Shreve, S.: Brownian Motion and Stochastic Calculus, vol. 113. Springer Science & Business Media, Berlin (2012)
Kloeden, P., Platen, E.: Numerical Solution of Stochastic Differential Equations, vol. 23. Springer, Berlin (1992)
Milstein, G.N., Tretyakov, M.V.: Stochastic numerics for mathematical physics. Scientific Computation. Springer, Berlin (2004)
Milstein, G.N., Tretyakov, M.V.: Practical variance reduction via regression for simulating diffusions. SIAM J. Numer. Anal. 47(2), 887–910 (2009)
Müller-Gronbach, T., Ritter, K., Yaroslavtseva, L.: On the complexity of computing quadrature formulas for marginal distributions of SDEs. J. Complex. 31(1), 110–145 (2015)
Newton, N.J.: Variance reduction for simulated diffusions. SIAM J. Appl. Math. 54(6), 1780–1805 (1994)
Acknowledgements
Stefan Häfner thanks the Faculty of Mathematics of the University of Duisburg-Essen, where this work was carried out.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Belomestny, D., Häfner, S., Urusov, M. (2017). Regression-Based Variance Reduction Approach for Strong Approximation Schemes. In: Panov, V. (eds) Modern Problems of Stochastic Analysis and Statistics. MPSAS 2016. Springer Proceedings in Mathematics & Statistics, vol 208. Springer, Cham. https://doi.org/10.1007/978-3-319-65313-6_7
Download citation
DOI: https://doi.org/10.1007/978-3-319-65313-6_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-65312-9
Online ISBN: 978-3-319-65313-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)