If you are unable to see the message below, click here to view.  
HOME Recent Events RMI in the News ALUMNI
  Issue 7 | Archive  

May 2011

The CDS-Bond Basis and the Cross Section of Corporate Bond Returns

A paper by Haitao Li (University of Michigan, Ann Arbor), Weina Zhang (National University of Singapore), and Gi Hyun Kim (University of Michigan, Ann Arbor)

In a recent working paper, RMI affiliated researcher Dr. Weina Zhang, with her co-authors, provided a comprehensive analysis on the CDS-Bond basis and its implications for the pricing of corporate bonds in the U.S.

The credit default swap (CDS) is the most liquid and popular product and accounts for more than two thirds of all outstanding credit derivatives. The appearance of credit derivatives since late 1990s have fundamentally changed market practices in the investment, trading, and management of credit risk. Traditionally, institutional investors, such as pension funds and insurance companies, typically adopt a buy-and-hold strategy in their investments in cash corporate bonds. Nowadays, speculators, such as hedge funds and proprietary trading desks of investment banks, can easily long and short the credit risk of individual companies or portfolios of companies using credit derivatives.

CDS has been widely used to "arbitrage" the mispricing of the credit risk of the same company in the cash and derivatives markets through the so-called CDS-Bond basis trade. The CDS-Bond basis is defined as the difference between the CDS spread of a reference firm and the spread of the firm's cash corporate bond with similar maturity. Prior studies have shown that CDS and bond spread should follow a co-integrated process since they measure the credit risk of the same company. Investors can easily arbitrage away non-zero basis if the two markets are expected to converge in the future. When the basis is negative (positive), one can long (short) the underlying corporate bond and buy (sell) CDS to bet on the narrowing of the basis. Since it is generally more difficult to short corporate bonds, the negative basis trade has been more popular in practice.

Unlike a standard textbook riskless arbitrage, the CDS-Bond basis trade faces a wide variety of new risks, such as deleveraging risk, counterparty risk, and funding liquidity risk. During the recent financial crisis, the basis of investment grade index in late 2008 is about -250 basis points (bps). Many banks and hedge funds, such as Deutsche Bank, Merrill Lynch, and Citadel, have lost billions of dollars due to the blow up of the basis trade. The widening of the basis has also created serious disruptions in the credit market even for investors who have not invested in CDS. Traditional passive investors in cash bonds suffer huge losses as well due to the unwinding of the basis trade. As a result, investment-grade corporate credit spreads, such as CDX.IG index rose from 50 bps in early 2007 to about 250 bps by the end of 2008. The spread of even the safest tranche, such as CDX.IG super senior tranche, widened to about 100 bps from 5 bps. Such a disruption caused by basis trade would not happen before the introduction of CDS because passive buy-and-hold cash bond investors are not exposed to these risks.

Instead of explicitly discussing each individual component of the risks in the basis trade, the authors use the basis as a reduced-form measure of all the risks involved in the basis trade. To some extent, the basis risk is reflected in the magnitude of the basis because arbitrageurs demand discounts to enter the trade to compensate for the risks they bear.

The empirical analysis shows that the basis risk factor, constructed as the return differential between LOW and HIGH quintile bond portfolios formed according to the level of CDS-Bond basis, is a new risk factor. The basis factor is priced in the cross section of investment grade bonds with an annual return of about 3% in normal periods and a negative return of 8% during the recent financial crisis. There is no evidence that the basis risk is priced in speculative grade bonds, which are not widely used in the basis trade.

Further decomposition analysis shows that the basis risk factor is modestly correlated with the empirical proxies of default risk and liquidity risk, funding liquidity risk, collateral risk, and counterparty risk. Nevertheless, the correlation coefficients vary significantly across different time periods and there still remains a significant portion that cannot be explained by the existing known systematic risk proxies in corporate bond market. These findings indicate that the new basis risk factor is an important new systematic risk factor to be included in addition to the risk factors such as Fama-French three factors (e.g., Market, SMB and HML factors), default and term factors, and liquidity risk factor used by the existing asset pricing models for corporate bond.

Back To Newsletter

Published quarterly by Risk Management Institute, NUS
Editor: Ivy Wang (rmiwy@nus.edu.sg)