The Real Effects Of Credit Default Swaps: Evidence From Firm Valuations
October 25, 2015
Credit default swaps (CDSs) may have real effects on the underlying firms when markets are incomplete. For instance, they may increase firm value by reducing supply-side credit constraints or decrease firm value by introducing inefficiencies in resolving financial distress. I find that the initiation of CDS trading is associated with lower firm valuations. The effect is stronger in active CDS markets, for riskier firms, and during periods of low counterparty risk. Consistent with a risk-based explanation, systematic risk of equity and stock returns increase following CDS initiation. I address endogeneity concerns using standard econometric methods and a quasi-natural experiment.
The Real Effects Of Credit Default Swaps: Evidence From Firm Valuations – Introduction
Although only 10% of public U.S. firms had credit default swap (CDS) contracts on their debt in 2013, these firms accounted for nearly 70% of public U.S. firms based on their market capitalizations. Given the importance of these large firms on the economy, it is crucial to quantify the effects of CDS-induced externalities, if any, on firm value. In this paper, I identify these externalities and investigate their net effect on firm valuations.
The extant literature suggests several reasons why CDSs may affect the value of the underlying assets. For example, CDSs may increase bankruptcy risk (e.g., Subrahmanyam, Tang, and Wang, 2014), reduce the liquidity of related securities (e.g., Boehmer, Chava, and Tookes, 2015), undermine lenders’ monitoring incentives and increase firm risk-taking (e.g., Parlour and Winton, 2013), reduce supply-side credit constraints (e.g., Saretto and Tookes, 2013), help complete markets (e.g., Ross, 1976), and stimulate information production (e.g., Acharya and Johnson, 2007). Taken altogether, these mechanisms provide mixed predictions about the net effect of CDSs on firm value.
I study a sample of U.S. industrial firms over the period 2001–2013, and use Tobin’s q as the primary valuation measure. I identify the CDS initiation date for each firm as the earliest date of its CDS pricing in the Bloomberg and Credit Market Analysis (CMA) CDS databases. As the effects of CDSs on firm value would be more evident in active CDS markets, I also compute continuous CDS activity proxies: Gross Notional and CDS Liquidity.
I find that CDS initiation is associated with a reduction in firm valuations, and the effect is more pronounced for firms with active CDS contracts, suggesting that CDSs introduce costs for the underlying firms that outweigh their potential benefits. A major concern is that credit default swaps initiation can be endogenous to firm value because unobserved firm characteristics related to credit risk may drive both the selection of firms for CDS trading and the changes in firm value.
I address endogeneity concerns by estimating regressions with firm fixed effects and implementing a two-stage instrumental variable approach. Following Saretto and Tookes (2013), I use the average foreign exchange hedging positions of financial institutions affiliated with a firm as an instrument for CDS initiation.1 Moreover, I study the effect of CDS activity on firm valuations within firms with CDSs and implement a propensity score-matching method.
To further mitigate endogeneity concerns, I exploit the implementation of the Big Bang Protocol (BBP) on April 8, 2009, as a quasi-natural experiment. BBP standardized the restructuring clause in the North American credit default swaps market to “No Restructuring” so that debt renegotiations (e.g., out-of-court debt workouts) no longer trigger CDS payouts. Prior to BBP, debt renegotiations triggered CDS payouts for the majority of investment-grade entities, but not for high-yield entities.2 BBP may affect firm value through the empty creditor channel.
When debt renegotiations do not trigger credit default swaps payouts, creditors hedged with CDSs (empty creditors) would have incentives to resist debt renegotiations in order to force the firm into bankruptcy (e.g., Bolton and Oehmke, 2011). This is because while renegotiations require empty creditors to accept concessions, a bankruptcy would trigger credit default swaps payouts and grant them potentially higher recoveries. Empty creditor resistance to debt renegotiations may reduce firm value because renegotiating the distressed debt (out of court) is a more efficient and less costly method of resolving distress than filing for bankruptcy (e.g., Bris, Welch, and Zhu, 2006). Since BBP changes the restructuring clause to “No Restructuring” for the majority of CDSs on investment-grade entities, I expect CDSs to decrease firm value more for investment-grade firms, relative to high yield firms, following the implementation of BBP. I find that BBP affects firm valuations in the expected direction.
See full PDF below.