Hedging credit: Equity liquidity matters

https://doi.org/10.1016/j.jfi.2008.08.005Get rights and content

Abstract

We theorize and confirm a new channel by means of which liquidity costs are embedded in CDS spreads. We show that credit default swap (CDS) spreads are directly related to equity market liquidity in the Merton [Merton, R.C., 1974. On the pricing of corporate debt: The risk structure of interest rates. J. Finance 29, 449–470] model via hedging. We confirm this relationship empirically using a sample of 1452 quarterly CDS spreads over 2001–2005. In the model, this relationship is monotone increasing when credit quality worsens. These results are robust to alternative measures of equity liquidity and other possible determinants of CDS spreads.

Introduction

A growing literature documents that illiquidity is a component of bond spreads. For instance, the “spread puzzle” where the spread between corporate bonds and Treasuries is too high to be explained by credit related factors (Collin-Dufresne et al., 2001, Huang and Huang, 2003) has been attributed mostly to illiquidity in the bond markets. Furthermore, credit risk is now being traded using credit default swaps (CDS) and the CDS-bond basis (the difference between the CDS spread and the bond yield) has been shown to be related to liquidity proxies (Longstaff et al., 2005, Mahanti et al., 2007). Hence the trading of credit risk through corporate bonds results in bearing liquidity risk.

In contrast to bond spreads, the natural assumption in the literature has been that CDS spreads contain minimal or no components of liquidity, and to a lesser extent, other non-default priced systematic risks. We investigate this assumption theoretically and empirically. The paper makes explicit the theoretical link between CDS spreads and illiquidity in the equity of the reference entity in the context of Merton's (1974) structural model. We take the theoretical predictions of the model to the data using a final sample of 1452 quarterly CDS spreads from 2001 to 2005 and find that the equity liquidity of the reference entity is negatively related to CDS spreads. These results are robust to different measures of liquidity and to other known determinants of CDS spreads. We believe this is the first paper to establish a link between CDS spreads and liquidity in the equity markets.1 Thus, this paper extends the literature which examines the role of liquidity in credit markets and the literature on explaining the cross-section of CDS spreads (see the papers by Berndt et al., 2003, Ericsson et al., in press, Das et al., 2006, Duffie et al., 2007).

There is an inherent dissimilarity between liquidity in corporate bonds and CDS liquidity based on differences in the market's use of these instruments. Whereas the average corporate bond does not trade frequently,2 and is held for portfolio reasons, default swaps are widely used in credit arbitrage, construction of CDOs, and risk management. Therefore, even though there is a literature on liquidity effects in bond spreads (see Chen et al., 2007, Goldstein et al., 2006), it is necessary to investigate the same phenomenon separately in the CDS markets. The sellers of CDS contracts actively hedge their exposures through the equity markets and through the use of options and debt-related instruments. When liquidity in the equity markets dries up, it becomes more expensive for sellers of CDS contracts to delta hedge their short credit positions by taking short positions in equity or long positions in put options. These hedging costs are recovered through higher CDS spreads, even when illiquidity is not systematic. Indeed, our empirical results confirm that equity market illiquidity remains a strong explanatory variable for CDS spreads even after controlling for other default related factors.

That liquidity is priced as a factor has been established for equity markets in prior work, such as that of Acharya and Pedersen (2005). Equity market illiquidity is priced into bond spreads, as shown in de Jong and Driessen (2005). These papers examine overall market illiquidity in equity and bond markets. Other work looks at liquidity in individual securities. For example, Chen et al. (2007) examine bond-specific illiquidity using bond market measures, but do not consider equity market linkages. In a similar manner, our work here focuses particularly on explaining the cross-section of CDS spreads with liquidity measures on individual names. This paper is not focused on whether illiquidity is a priced factor in default swap markets. The hedging mechanism implies that transaction costs are transmitted into spreads even when this risk is not systematic, in the same manner in which default likelihoods are components of spreads. There may be additional liquidity premia arising from correlated risks in CDS spreads, suggested in the work of Acharya et al. (2007), but this is not the focus of our investigation.

Unlike the past literature, we focus on the mechanism for the transmission of illiquidity from equity markets to CDS spreads. We posit that, since CDS contracts are actively hedged, unlike bonds, and because hedging costs are incurred whether or not liquidity risk is systematic, we should anticipate that illiquidity costs from the equity markets are transmitted into CDS spreads. Using standard measures of illiquidity and transactions costs in the equity markets, such as the ILLIQ measure of price impact of Amihud (2002), the LOT measure of Lesmond et al. (1999), and bid–ask spreads, regressions show that individual variations in illiquidity across firms explain the cross-section of CDS spreads even after controlling for default and other explanatory variables.3 By controlling for common time-series effects across firms, we isolate the firm-specific component of the impact of equity market illiquidity on CDS spreads.

There is also growing evidence that default risk and liquidity risk are correlated. Acharya et al. (2007) present a model where declining credit quality results in the drying up of liquidity in the corporate debt markets. Similar relationships are observed in Downing et al. (2007). Credit spreads and illiquidity are positively correlated in the empirical record. In this paper, we present a model in which such an effect is theoretically supported at the level of individual credit names. Upon testing, this theoretical proposition is also found to be supported by the data, i.e. liquidity components increase as the credit risk of an individual issuer increases. Thus, our theoretical and empirical results complement those of the literature.

The rest of the paper proceeds as follows. The model in Section 2 establishes the link between equity market illiquidity and CDS spreads via a hedging mechanism. It also posits that the impact of illiquidity in the equity markets on CDS spreads will increase when credit quality worsens. This relationship is monotone and convex. These relationships are tested in Section 3, where we also present the data and the variables used in the study. The results confirm the theoretical predictions. Section 4 provides concluding comments.

Section snippets

Hedging CDS in a structural model

Ericsson and Renault (2006) develop a structural model to connect bond market liquidity with default risk. In their model, bond spreads are related to costs of having to trade when it is not optimal to do so. Random liquidity shocks force suboptimal bond trades resulting in potential reductions in value of the bonds. This cost is embedded in bond spreads. Hence bond illiquidity is related to the lack of immediacy in liquidating a bond position (see Chacko, 2005, Chacko et al., 2008 for more

Data

Our sample of credit default swap spreads is obtained from Bloomberg. It consists of 2860 quarterly credit default swap spreads over the period 2001–2005. This sample was further restricted to include only CDS securities where the notional value is dollar denominated and where the reference entity is a publicly traded firm. We further restricted the sample to CDS contracts of five-year maturity as these are the most actively traded maturity. Financial information on the reference entity is then

Conclusion

Whereas it is widely accepted that liquidity is a major component of the spreads of corporate bonds, there is almost no literature on liquidity in CDS spreads. This paper studies whether individual firm liquidity can further explain the cross-section of CDS spreads, after controlling for default risk, using market-based and firm-specific variables. We find strong evidence that CDS spreads contain liquidity components. We used three different proxies for equity market liquidity that are commonly

Acknowledgments

We are grateful to the editor, S. Viswanathan, an anonymous referee, as well as Edith Hotchkiss and participants at the Moodys Credit Risk Conference 2008 for excellent comments that helped in improving the paper.

References (31)

  • M. Blume et al.

    The declining credit quality of U.S. corporate debt: Myth or reality?

    J. Finance

    (1998)
  • Carr, P., Wu, L., 2007. A simple robust link between American puts and credit insurance. Working paper, Baruch...
  • Chacko, G., 2005. Liquidity risk in the corporate bond markets. Working paper, Harvard Business...
  • G. Chacko et al.

    The price of immediacy

    J. Finance

    (2008)
  • L. Chen et al.

    Corporate yield spreads and bond liquidity

    J. Finance

    (2007)
  • Cited by (52)

    • ESG, liquidity, and stock returns

      2022, Journal of International Financial Markets, Institutions and Money
    • Liquidity tail risk and credit default swap spreads

      2018, European Journal of Operational Research
      Citation Excerpt :

      CDS contracts are often traded together with equity securities, for example, in the context of capital structure arbitrage.28 Therefore, equity liquidity complements CDS liquidity and affects CDS spreads, because, for instance, an illiquid equity market leads to more expensive hedging costs for sellers of CDS contracts, which then are recovered through higher CDS spreads (see, e.g., Das & Hanouna, 2012; Lesplingart et al., 2012; Tang & Yan, 2006). Some of the empirical studies on the pricing of CDS have already focused on potential liquidity spillover effects between the equity and the CDS market (see, e.g., Junge & Trolle, 2015; Lesplingart et al., 2012; Meine et al., 2015; Tang & Yan, 2006).

    View all citing articles on Scopus
    View full text