Electronic Theses and Dissertations


Yi LuFollow



Document Type


Degree Name

Doctor of Philosophy



Committee Chair

David Kemme

Committee Member

Joonhyung Lee

Committee Member

Jose Lopez

Committee Member

Jacob Kleinow


This dissertation provides a comprehensive study of the credit-default swaps (CDS). In the short-run, we review and synthesize literature across disciplines and offer a theoretical framework of the empirical relationship between CDS premiums, bond yields, and stock prices. Theoretically, the credit risk captured by CDS and bond spreads should be equal, since both CDS and bond spreads are the risk premiums associated with corporations risky instruments. However, data reveal a significant difference between the two spreads. Stock prices, as well, should impact CDS and bond spreads. In a panel time-series framework with daily data of 96 corporations from North America, Europe, and Asia during the period from 2005 to 2013, we examine the impact of credit events from one market on the other markets in the short-run fluctuation using a Vector Autoregression (VAR) model. We also analyze the volatility of the correlation between CDS and bond markets with a multivariate GARCH model.In the long-run, we find that changes in CDS spreads generally precede bond spreads and stock prices in the price-discovery process, but there were significant reversals in information flows across the global financial crisis. With high-quality time-series data of CDS premiums, bond yields, and stock prices that are available, we find the time-series to be cointegrated and examine the long-run adjustment process between CDS, bond and stock markets with a Vector Error Correction model (VECM). We use rolling window samples to estimate the time-varying long-run adjustment process and the speed of adjustment in response to information shocks with Gonzales and Granger measures of price discovery. We find overall the CDS market leads the bond or stock markets in the price-discovery process. But, importantly the results are firm-specific, they vary by geographic locations of the market, sectors of the industry and occurrences of the credit event. There are significant reversals of the information flow prior to and after the 20072008 financial crisis.


Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest