Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.556250
Title: Credit risk pricing with quadratic term structure model
Author: Li, Yao Dong
Awarding Body: University of York
Current Institution: University of York
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
Abstract:
This thesis is an empirical credit risk study, developing a multi-factor quadratic term structure model in discrete time and investigating credit risk. There are three main contributions. First, I examine the corporate credit default swap premia. I estimate parameters using an extended Kalman filter. The results show that the discrete quadratic term structure model is able to avoid some of the drawbacks associated with continuous models and affine term structure models. In addition, it was found that two state variables are enough to explain most of the variation in corporate credit spread. Second, I use an empirical study to analyze how market priced default component and nondefault component were incorporated into corporate bonds during 2005 - 2009. I jointly model the corporate bonds and credit default swaps using a three-factor quadratic term structure model and an extended Kalman filter. The evidence shows that the default component is not the major part of the corporate spread for most firms in my dataset, rather it is the time-varying nondefault component. Lastly, I study a large range of sovereign and bank credit default swap premia. I use a quadratic term structure model to define credit risk and decompose bank credit default swap premia into three factors: sovereign, common and bank-specific factors. The results indicate that banks in the dataset in all countries outside the United States (US) are very sensitive to common factors, and relatively sensitive to sovereign factors. The US banks, however, show more sensitivity to sovereign factors. Although sovereign and common factors explain a large portion of bank credit default swap premia, bank- specific factors played a major role, especially after the Lehman's failure. I also examine how the market priced the default component in both the dollar and sterling LIBOR (London Interbank Offered Rate). The evidence shows that the default component of LIBOR no longer reflects the panel banks' credit risks. Before September 2008, the default component of LIBOR was mainly driven by the coeffects of the sovereign and common factors. After that, both dollar and sterling LIBOR were reduced to an historic low. The sterling LIBOR are almost driven by sovereign factors while the dollar LIBOR reflects an even smaller credit risk than the sovereign risk. This evidence confirms market participants' thoughts that the dollar LIBOR should be higher than the observed rate.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.556250  DOI: Not available
Share: