The cost of CDS liquidity

Yields on the new Greek 2042 bonds have jumped 3% this week to 16.93%. Prices, which move inversely to yields, on these new bonds are going down is because market participants are unable to hedge the risk inherent in these sovereign debt instruments. The reason for this lack of hedging ability stems from the ISDA‘s 2009 Supplement to the 2003 ISDA Credit Derivatives Definitions document. Let me explain.

After the 2003 Definitions were published, a credit event that occurred on or after the trade date of the CDS could trigger payout. While this provided a healthy amount of protection to buyers, who were hedged immediately upon buying a CDS, this same immediacy could make some CDS contracts not fungible. For example, if a CDS transaction occurs on January 1st and a credit event occurs on January 12th and a second CDS transaction occurs on January 20th, the first CDS and the second CDS are not fungible because the first CDS can be triggered by the January 12th credit event while the second CDS cannot be triggered by the January 12th credit event.

At first blush, this seems like the way the market should work. However, the problem occurs when you examine how derivatives function in the real world. A simple derivative like a call option grants the buyer the right, but not the obligation, to buy the underlying security for a predetermined price by a predetermined date. In reality, the underlying security is rarely bought and sold in this situation. What actually happens is the call option seller offsets their obligation by buying an identical call option when the call option buyer executes their option. Eventually, somebody somewhere must buy the underlying security but not in the vast majority of party/counter-party pairs. In order for our hypothetical seller to be able to offset their obligation, the two call options must be identical; they must be fungible. This was not the case after the 2003 Definitions came into being because CDS contracts with different trade dates were treated differently if a credit event occurred between the two trade dates.

This begs a more fundamental question, of course: why should a CDS purchased after a credit event be interchangeable with a CDS purchased prior to a credit event? Answer: to allow sell-side market participants to cover their obligations to their counter-parties after a credit event has occurred. Because CDS settlement is a long process involving complex procedures like an auction, market makers need time to cover their obligations to their counter-parties.

The 2009 Supplement changed this by allowing market participants more time and thus giving CDS contracts more liquidity. According to law firm Allen & Overy, “the Supplement provides that any credit event that occurred more than 60 calendar days prior to the date on which a request [to declare a credit event] is submitted to [the] ISDA … will not be able to trigger settlement of a credit derivative.” This means that sell-side market participants have 60 days to cover their obligations to their counter-parties by buying CDS contracts.

The feared corollary to this is that a credit event that occurs less than 60 days prior to the ISDA request date can be used to trigger CDS payout. As the FT reported, “Bankers want the ISDA to make a swift ruling that [backstop] clauses in CDS contracts cannot be used to activate fresh payouts on any new protection.” Without this assurance from the ISDA, buyers of new CDS contracts could trigger payout immediately under the 2009 Supplement 60 day backstop provision.

As a result, the last time Markit, who requires at least three banks to provide CDS prices in order to give CDS quotes, quoted CDS prices was on March 9th. Simply put, no sell-side market participant wants to take the risk of having a freshly written CDS executed on them and thus none of them are giving CDS quotes. This state will remain until either the ISDA rules that 60 day backstop provisions cannot be used to trigger payout on new CDS contracts or 61 days elapses from the March 9th credit event date.

Until one of these two conditions is met, it seems likely to me that that Greek yields will continue climb as bond buyers are unable to hedge the risk of another Greek credit event.

My interest in all this was piqued after reading the previously linked FT article. Of course, my explanation above is merely my understanding of the forces moving Greek yields and may be incorrect or incomplete and should not be taken as legal or financial advice of any kind.

No related posts.

25. March 2012 by Chris
Categories: Finance, Thought Experiments | Leave a comment

Leave a Reply

Required fields are marked *