Ebook An Econometric Model of Credit Spreads with Rebalancing, ARCH and Jump Effects
Accessing and managing credit risk of risky debt instruments has been a major area of interest and concern to academics, practitioners and regulators, especially in the aftermath of a series of recent credit crises such as the Russian default and the Enron and WorldCom collapses. In particular, there has been a fast growing literature on models of credit risk measurement. One measurement issue that is both interesting and challenging is the credit risk of a portfolio of risky bonds. This is especially relevant for banks, pension funds, insurance companies, and bond mutual funds. However, the literature on portfolio credit risk is still new, especially in the area of empirical studies.
In this paper, we examine the time series behavior of credit spreads on corporate bond portfolios and propose an econometric model for these spreads that incorporates portfolio rebalancing, unit root, conditional heteroscedasticity, jumps, and Treasury bond and/or equity market factors. More specifically, we apply this model to option-adjusted spreads (OAS) of Merrill Lynch (ML) corporate bond indices for nine rating/maturity categories over January, 1997 through August, 2002.
There are a few benefits from examining credit spreads of corporate bond indices from a major dealer. First, the ML credit spread indices are representative portfolios with a given rating and maturity range, are updated daily, and are a leading index in credit markets. Yields on several Merrill Lynch corporate bond indices are in fact quoted in the Wall Street Journal. A credit spread index is also useful for corporate bond index funds that track a corporate bond index. The accuracy of a tracking model depends crucially on its assumptions on the interest rate and credit spread dynamics of the targeted index (see Jobst and Zenios (2003)).
Second, credit spread indices could serve as the underlying instrument for credit derivatives. There has been a rapid growing market for structured credit products recently, with the total value exceeding $1 trillion in 2001 (Berd and Howard, 2001). Although there are no credit derivatives written directly on credit spread indices in the market at this moment, there are ongoing efforts in the industry to make this happen. For example, Standard & Poor’s has announced that “S&P Credit Indices are now available for licensing in association with derivative products, structured notes, reference obligations and other applications” (Standard & Poor’s, 1999). To design and price credit derivatives based on a credit spread index, it is important to investigate first the time series properties of the index.
Third, the ML high-yield indices each cover a substantial amount of high-yield issues (see Appendix B). Existing studies on the dynamics of individual bond credit spreads have mostly focused on monthly prices of investment grade bonds because of data constraints (e.g., Duffee, 1999). A study of high-yield indices may help understand the general dynamics of credit spreads in the high-yield bond market.
Download
PDF Ebook An Econometric Model of Credit Spreads with Rebalancing, ARCH and Jump Effects
Posted in :
