Extensions of the Cross-Entropy Method with Applications to Diffusion Processes and Portfolio Losses
Degree
Doctor of Philosophy
Program
Applied Mathematics
Supervisor
Adam Metzler
Abstract
Rare event simulation is a crucial part of simulations. In financial mathematics, the study of rare events appear naturally when we consider risk measures such as the conditional value at risk. This thesis is composed of three related papers treating the rare event simulations subject: the first paper addresses rare event simulations using for diffusion processes, the second paper addresses rare event simulations for the normal and the Student t-copula model while the last paper addresses rare event simulations for a portfolio model where there is a correlation structure between the loss-given-default and the probability of default.
Recommended Citation
Scott, Alexandre, "Extensions of the Cross-Entropy Method with Applications to Diffusion Processes and Portfolio Losses" (2015). Electronic Thesis and Dissertation Repository. 2858.
https://ir.lib.uwo.ca/etd/2858