Numerical Finance with Backward Stochastic Differential Equations: An Exploration of Three Schemes

Ki Wai Chau

Research output: ThesisDissertation (TU Delft)

64 Downloads (Pure)

Abstract

The main aims of this research are to study various numerical schemes in the approximation of the occurring expectations and their applications in numerically solving BSDEs. We focus on numerical expectation/finite measure integration since the majority of the BSDE solvers consists of two parts, conditional expectations computations, and deterministic functions to map these expectations to target approximations. By simply changing the approximation for conditional expectations, we can effectively generate various schemes for BSDEs that can suit different requirements. Furthermore, our results carry implications in numerical integration too. In this thesis, we focus on the mathematical properties of these approximations. We will discuss the fundamental assumptions for them, give complete descriptions, derive
error bounds and conduct numerical experiments. The main goal is to analyze these approximations. We will also touch upon the financial applications of BSDEs.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Delft University of Technology
Supervisors/Advisors
  • Oosterlee, C.W., Supervisor
Award date16 Jan 2020
Print ISBNs978-94-028-1886-4
DOIs
Publication statusPublished - 2020

Fingerprint

Dive into the research topics of 'Numerical Finance with Backward Stochastic Differential Equations: An Exploration of Three Schemes'. Together they form a unique fingerprint.

Cite this