The lead article in Volume 23, Issue 3 of the Review of Finance is Do Credit Default Swaps Mitigate the Impact of Credit Downgrades? by Sudheer Chava, Rohan Ganduri, and Chayawat Ornthanalai. When a firm’s debt is downgraded, its stock price falls, it subsequently raises less debt, and its cost of debt increases. This paper finds that, when credit default swaps (CDS) trade on its debt, all three effects are reduced – suggesting that CDS alleviates the financial frictions arising from downgrades.
The CDS Controversy
In a Credit Default Swap, the CDS seller pays the CDS buyer when a borrower (such as a company) defaults on its debt. In theory, CDSs can reduce the risk in financial markets by providing insurance. For example, if a bank has lent to a company, it can buy CDS to protect itself from that company going bankrupt.
But in practice, CDSs are highly controversial, for several reasons. First, CDS insurance pays out far less frequently than buyers hope, since sellers often get away with claiming that a borrower has not in fact defaulted. This point was made forcefully in a 2018 Financial Times Op-Ed, entitled “Time to wipe out the absurd credit default swap market” with the strapline “Someone must deliver a coup de grâce to this global joke”. Second, buying insurance removes the incentive for a borrower to monitor the lender, potentially increasing the likelihood of defaults. Bolton and Oehmke (2011) termed this the empty creditor problem. Third, CDSs can be used for excessive speculation – they’re a type of derivative, which Warren Buffett famously called “financial weapons of mass destruction”.
A Novel Role for CDS
Defenders of CDSs have typically focused on how they benefit investors. Buyers of CDS obtain insurance; sellers of CDS generate revenue. But Sudheer, Rohan, and Chayawat identify a new benefit – to the borrower. This benefit is particularly interesting since the borrower isn’t a party to a CDS transaction – it’s a side-bet between investors – yet gains from positive spillovers.
Here’s how such benefits might arise. Credit rating downgrades (e.g. a company’s rating falling from BBB to BB) are typically bad events. The downgrade signals that the company is less creditworthy than investors previously thought, making it difficult to borrow in the future. The downgrade may make existing debt more expensive, if its interest rate is linked to the credit rating. All of these effects hurt shareholders, causing the stock price to drop.
However, if a firm has CDS trading on its debt, these negative effects may be muted through two channels. The information channel argues that active CDS trading already provides investors with up-to-date information about a firm’s credit rating. If a firm’s credit position starts to deteriorate, then CDS spreads will widen (the cost of CDS insurance will increase) in advance of the actual downgrade. Thus, the actual downgrade is less of a surprise, and so its effects are less negative. The financial frictions channel argues that CDS reduces the impact of a downgrade on a company’s ability to borrow in the future. If lenders can buy CDS, they may still be willing to lend to a borrower even after it has been downgraded. For example, when a borrower is downgraded, a bank typically faces a higher regulatory capital requirement, but this is not the case if it has bought CDS.
While both channels suggest that CDS trading dampens the impact of credit rating downgrades, the empty creditor problem predicts the opposite. A downgrade is more problematic if creditors aren’t actively monitoring the borrower to address the problems (e.g. poor management decisions) that may have led to the downgrade. So the overall effect is unclear – providing the motivation for the paper.
The authors find that CDS trading significantly alleviates the damage caused by a credit rating downgrade. Their evidence is consistent across many outcomes:
- Stock price reaction. The negative stock price reaction to a downgrade is 44-52% lower if a firm has traded CDS.
- Future debt issuance. In the year after a downgrade, firms without traded CDS reduce their net debt issuance by 3.72 percentage points. Traded CDS reduces this impact by 40% to 2.28 percentage points.
- Cost of debt. In the year after a downgrade, the loan spreads of non-CDS firms rises by 22-29%. For CDS firms, this rise is almost 50% lower, at 12-16%
This alleviation has further knock-on benefits. If a borrower knows that the consequences of a downgrade are less severe, it may engage in less “manipulation” to prevent a downgrade. In particular, credit ratings depend on discrete thresholds for financial ratios. For example, for firms with an excellent business risk profile, a Debt/EBITDA ratio of 1.5-2.0 typically leads to a rating of AA; a ratio of 2.0-3.0 typically leads to a rating of A. Thus, a firm with a Debt/EBITDA ratio of close to 2.0 will ensure that it remains below 2.0 by reducing its net debt issuance, potentially at the expense of investment.
Indeed, the authors find that firms closer to Debt/EBITDA thresholds issue less debt than firms further from thresholds. However, this effect is lower for CDS firms. As a result, they issue 2.8 percentage points more net debt per year than non-CDS firms. Thus, CDS affects not only a firm’s reaction to a downgrade, but also its financing behaviour in general.
Correlation or Causation?
As with almost all studies, a concern is whether the results are correlations or causations. Whether a firm has CDS trading is not random – investors choose to trade CDS on particular firms. If CDS firms are fundamentally different from non-CDS firms, this may explain the former’s lower reactions to downgrades. For example, investors typically trade CDS on better-rated firms (as they’re less willing to sell insurance on lowly-rated firms), and better-rated firms are typically less affected by downgrades. Thus, the authors compare borrowers with the same initial credit rating that were downgraded by the same amount, and only differ in whether they have CDS trading.
Of course, this only controls for one potential difference between CDS and non-CDS firms – the initial credit rating. These firms could differ among other dimensions. Thus, the authors use instruments that affect the likelihood that a company has CDS trading, but should have no effect on the reaction to credit downgrades. One is a time-series instrument, that takes a given company and predicts when CDS trading will be introduced on its debt. This is the aggregate global trading volume of CDS – when market demand for CDS is high in general, CDS trading is more likely to be introduced at a specific firm. The second is a cross-sectional instrument, that predicts which firms are likely to have traded CDS at a particular point in time. This is the average foreign exchange traded for hedging purposes by a downgraded firm’s lending banks, which captures the demand for CDS. Using both instruments, the results continue to hold, suggesting that CDS trading causes the muted reactions.
Finally, the authors study whether the information channel or the financial frictions channel is more likely to be driving their results. This difference is important. If the information channel is the main driver, the benefits of CDS trading are lessened – while the actual downgrade causes less damage, the negative effects may have transpired beforehand. For example, if the downgrade was foreshadowed by a rise in CDS spreads, the fall in the stock price, reduction in borrowing, and rise in the cost of debt may have commenced once CDS spreads started to rise. But if the financial frictions channel is the main driver, the destabilising effects of downgrades are indeed weaker. And this is what the authors find.
First, they find little evidence for the information channel. If this channel is at work, the market reaction to credit downgrades should be smaller if CDS trading is more active (e.g. greater CDS spread changes, volatility, or liquidity) – as this suggests that the CDS market is more informative. This is not the case.
Second, the authors find affirmative evidence for the financial frictions channel. Specifically, their results are stronger where credit rating downgrades are more likely to create financial frictions – where the downgrade takes the borrower from “investment grade” to junk status, where the cost of existing debt is linked to the credit rating, and where the borrower has more existing loans.
Overall, the results highlight a novel benefit of CDS. Not only can they be a stabilising force to investors, providing them with insurance, but also to companies, easing the financial frictions normally caused by credit downgrades.