Use this URL to cite or link to this record in EThOS:
Title: Essays in quantitative finance on risk management and credit portfolio optimisation
Author: Wang, Zhi
ISNI:       0000 0004 1924 3715
Awarding Body: University of Essex
Current Institution: University of Essex
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
This thesis discusses three topics in the area of quantitative finance in relation to risk and credit portfolio management. Chapter 2 investigates the issue of estimating and testing the goodness-of-fit of a model for a dependence break. The dependence is modelled by copulas and an unknown break of dependence structure is allowed for by including a dummy variable in the copula. The model is selected by minimizing the Akaike Information Criterion (AIC) of each candidate breaking point. The candidate models are estimated by a well-established two-step Maximum Likelihood (ML) approach, namely "Inference Function for Margin" (IFM). Moreover, we examine 5 single-factor copulas and compare them to each other by AIC criteria. A parametric bootstrap goodness-of-fit test is also proposed. Empirically, the dependence structures of stock indices between the US-UK and US-Japan markets during the Subprime crisis are examined. We found breaks in both dependence structures. In Chapter 3, a new general approach is developed for optimizing a credit portfolio by minimizing the default risk of a whole credit portfolio subject to a certain target premium. The approach is rooted in concepts from Modem Portfolio Theory. The default risk is measured by a quadratic form of weights and a matrix containing information about default correlations between any two single-names and default intensities of each single-name. The default correlation and the default intensities are modelled by a new binomial intensity model. A Genetic Algorithm (GA) approach is also introduced to optimize a credit portfolio with the purpose of overcoming limitations of the analytical method and the traditional numerical method based on the first order condition. Empirically, the approach is applied to optimize Credit Default Swap (CDS) portfolios consisting of members of iTraxx and CDX indices. In Chapter 4, we focus on modelling counterparty risks of two important financial instruments: the Interest Rate Swap (IRS) and the CDS. Analytical solutions are derived for the theoretical fair prices of the IRS and the CDS under various assumptions of defaults of counterparties. Also a Monte Carlo approach is proposed as a numerical solution for the fair prices. Numerical experiments are designed to study the effects of various factors on the fair price. Empirically, we examine the counterparty risk of a CDS portfolio, composed of randomly selected single-names from iTraxx series 10.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available