Big Risks with Small Price Impact – Counterparty Risk Management in CDS

Why the discussion about central clearing of credit default swap (CDS) contracts? For one thing, central clearing increases transparency. Once transactions have to go through a central clearing counterparty (CCP), the regulator can simply monitor the CCP to get an idea of who is creating exposures. Remember the 2008 AIG default?

Market participants and regulators had no idea about the number of triggered CDS contracts nor about the repercussions throughout the financial system via CDS exposures.

Central clearing counterparties eliminate counterparty risk

But the AIG example also shows another advantage of central clearing. AIG was an important protection seller in CDS markets – and after the default, protection purchased from AIG was virtually worthless. This counterparty risk is a general feature of bilateral contracts: if your contract partner goes bankrupt, they cannot fulfill their part of their contract any more. Replacing a potentially risky counterparty with a well-capitalized, hedged CCP can eliminate this counterparty risk.

Costly counterparty risk management

Managing counterparty risk in non-cleared contract can be bothersome: maintaining separate margin accounts for all underlyings and counterparties, monitoring the risk evolution, and making and receiving margin accounts is capital-intensive and time-consuming. Hence, a CDS contract with counterparty risk is, ceteris paribus, less attractive than one without this risk. Therefore, the price you are willing to purchase protection (the CDS spread) can be affected by counterparty risk.

Practical implications: CDS spreads as default risk indicators

This is not only of academic interest. Investors often use observed CDS spreads to quantify the default risk of the underlying. But are CDS spreads low because the default risk of the underlying is low, or because the protection seller is really, really risky? In their 2012 article published in the Journal of Financial Economics “Counterparty credit risk and the credit default swap market”, Navneet Arora, Priyank Gandhi and Francis Longstaff explore this question.

Identifying the price impact of counterparty risk

In order to be able to identify the impact of counterparty risk on CDS spreads, Arora et al. (2012) run a regression of the CDS transaction spread on underlying u on the (lagged) spread of the dealer d who sells protection. The higher the spread of d, the higher is the dealer’s default risk. Hence, the negative and signficant coefficient estimate in the regression means that counterparty risk is priced. Interestingly, the economic effect is small. An increase in the CDS spread of the dealer of 645 bps (corresponding to a downgrade from AAA to CCC, and an increase in the default probability from 0% to 33%) decreases the transaction spread by only 1 bp!

Why is the price impact so small?

How can such a huge increase in counterparty risk result in such a tiny decrease in the CDS spread? Arora et al. (2012) explain this by collateralization: Both counterparties regularly post collateral in a CDS transaction. This collateral mitigates losses when a counterparty default occurs. The authors suggest that the vast majority of CDS transactions are fully collateralized. Therefore, the price effect of counterparty risk is very small. Apart from the price effect, Arora et al. (2012) find evidence for a “flight to quality”: the safest dealers are in high demand as protection sellers, and can charge a premium.

Summary: Counterparty choice and collateralization limit the price impact

In summary, counterparty choice and collateralization mitigate counterparty risk in the CDS market. This suggests that a central counterparty (link to article) may not reduce counterparty risk.