We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.NA

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Numerical Analysis

Title: Pathwise Iteration for Backward SDEs

Abstract: We introduce a novel numerical approach for a class of stochastic dynamic programs which arise as discretizations of backward stochastic differential equations or semi-linear partial differential equations. Solving such dynamic programs numerically requires the approximation of nested conditional expectations, i.e., iterated integrals of previous approximations. Our approach allows us to compute and iteratively improve upper and lower bounds on the true solution starting from an arbitrary and possibly crude input approximation. We demonstrate the benefits of our approach in a high dimensional financial application.
Subjects: Numerical Analysis (math.NA); Probability (math.PR); Computational Finance (q-fin.CP)
Cite as: arXiv:1605.07500 [math.NA]
  (or arXiv:1605.07500v2 [math.NA] for this version)

Submission history

From: Christian Gaertner [view email]
[v1] Tue, 24 May 2016 15:28:07 GMT (26kb)
[v2] Thu, 23 Jun 2016 15:06:21 GMT (26kb)

Link back to: arXiv, form interface, contact.