Three essays on credit default swaps

• Main Author: Pu, Xiaoling
• Language: ENG
• Published: ScholarWorks@UMass Amherst 2008
• Subjects: Finance|Economic theory
• Online Access: https://scholarworks.umass.edu/dissertations/AAI3336926

The credit default swap (CDS) market has blossomed to become a major asset class in the capital markets. Once largely confined to banks, the market participants have expanded to include insurance companies, hedge funds, mutual funds, pension funds, and other investors looking for yield enhancement or credit risk transference. The applications have evolved from the financial institutions’ needs to manage their illiquid credit concentrations to hedge their credit exposure. The literature review presents the recent work on credit default swap valuation. Structural models, reduced form models, and incomplete information models are reviewed, together with the modeling frameworks and related empirical performances. The first essay examines the reason of low correlation observed between equity returns and credit spread changes. In a recent article, Collin-Dufresne, Goldstein and Martin (2001) point out that the correlation between equity and credit markets are low, considering the fact that both stocks and bonds are claims on the same underlying firm value. We investigate whether this is so because arbitrageurs face impediments to arbitrage. Using arbitrage impediments measures, we find support for the hypothesis that greater impediments lead to a lower correlation between the two markets of a given firm. The impediments include transaction costs and holding costs, which are measured as the liquidity in the credit and equity markets and the idiosyncratic risk exposure, respectively. In addition, we include the information measure, which is related to the unobservability of the underlying firm value in the arbitrages across stocks and bonds. The second essay examines various liquidity measures across the credit derivatives and corporate bond markets. The results, from the factor decompositions for individual liquidity measures and across various measures, show that there is a strong commonality in the liquidity measures of the fixed income markets. In addition, the CDS innovation, unexplained part in the credit spreads by structural models, is estimated from the linear and nonlinear regressions, respectively. The paper finds that the liquidity common factors have significant impact on the part of the credit spread changes unexplained by default risk factors. The third essay explores the components of the credit spread changes related to expected default, market risk premium and liquidity. The firm level factors, proxy for expected default, include equity returns, changes in volatility, changes in firm leverage, changes in book to market equity ratio, and changes in profitability. The common factors include changes in treasury yields, changes in the slope of the yield curve, changes in VIX, and three Fama-French factors. The liquidity factors include the aggregate stock market liquidity measure and the credit market depth. Furthermore, the paper examines whether the news from firm level or common factors that mainly drives the credit spread changes. The paper explores the relative impact of these factors to the credit spread changes in a vector autoregressive model. The results suggest that the innovation from the firm level factors have larger impact than that from the common factors for high yield firms while the investment grade firms are more affected by the news from the common factors.