Dissertation/Thesis Abstract

Synthetic collateral debt obligation pricing
by Zhanyong, Liu, M.Sc., Singapore Management University (Singapore), 2007, 74; 1483222
Abstract (Summary)

Portfolio credit products, such as CDO and Single Tranche CDO (STCDO) have gained their popularity in financial industry. The key problem facing by the financial engineers is how to price these portfolio credit derivatives, especially how to model the dependent default structure. Copula model proposed by Li (2000) is widely used in practice. Comparing with simulation, factor copula model and conditional independent framework provide good balance between accuracy and computational efficiency, but it is hard to achieve good performance if sticking to normal distribution. There are a few ways to improve it: introducing Levy distributions, using generic copula functions, and the semi parametric estimation. In this paper the Levy distribution and conditional independent factor copula model are examined. The flexibility and accuracy improvement comes from calibrating the skewness and heavy tail of Levy distribution for the underlying marginal distributions. The simulation result and short period prediction result are discussed too.

One of the other benefits of this model is that once calibrating to the standard market tranches spreads, the model can handle the customized CDO, e.g. Single Tranche CDO.

Key words: factor copula model, portfolio credit derivatives, Levy process

Indexing (document details)
Advisor: Lim, Kian Guan
Commitee:
School: Singapore Management University (Singapore)
Department: Lee Kong Chian School of Business
School Location: Republic of Singapore
Source: MAI 49/03M, Masters Abstracts International
Source Type: DISSERTATION
Subjects: Finance, Banking
Keywords: Factor copula model, Finance and financial management, Levy process, Portfolio and security analysis, Portfolio credit derivatives
Publication Number: 1483222
ISBN: 9781124352176
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest