Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.745832
Title: Essays on dependence modelling with vine copulas and its applications
Author: Xia, Xinghua
ISNI:       0000 0004 7228 1163
Awarding Body: University of Leicester
Current Institution: University of Leicester
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis contains three essays on dependence modelling with high dimension vine copulas and its applications in credit portfolio risk, asset allocation and international financial contagion. In the first essay, we demonstrate the superiority of vine copulas over multivariate Gaussian copula when modelling the dependence structure of a credit portfolio risk factors. We introduce the vine copulas to modelling the dependence structure of multi risk factors log returns in the combined framework of both threshold model and mixture model credit risk modelling. The second essay studies asset allocation decisions in the presence of regime switching on asset allocation with alternative investments. We find evidence that two regimes, characterized as bear and bull states, are required to capture the joint distribution of stock, bond and alternative investments returns. Optimal asset allocation varies considerably across these states and changes over time. Therefore, in order to capture observed asymmetric dependence and tail dependence in financial asset returns, we introduce high dimensional vine copula and construct a multivariate vine copula regime-switching model, which account for asymmetric dependence and tail dependence in high dimensional data. The third essay explores the cross-market dependence between six popular equity indices (S&P 500, NASDAQ 100, FTSE 100, DAX 30, Euro Stoxx 50 and Nikkei 225), and their corresponding volatility indices (VIX, VXN, VFTSE, VDAX, VSTOXX and VXJ). In particular, we propose a novel dynamic method that combine the Generalised Autoregressive Score (GAS) Method with high dimension R-vine copula approach which is able to capture the time-varying tail dependence coefficient (TDC) of index returns.
Supervisor: Dias, Alexandra ; Haven, Emmanuel Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.745832  DOI: Not available
Share: