A Thriving Market in CDS for Sovereign Debt

Have you ever really taken a moment to think about the implications of a thriving CDS market on sovereign debt?

First, a word about the Credit Default Swap market size.  Since CDS are private bilateral agreements, there is really no single source to determine the outstanding notional amounts. However, much has been done to centralize trades, and the DTCC has illustrated a view of market size with some figures that were published in the Financial Times (FT) this morning (11/23/09 - “Bets Rise on Rich Country Defaults”). As per the FT, the Gross Notional of CDS outstanding on developed and emerging sovereign debt, by my simple count, was well over $500 billion. Of course, there are many trillion in outstanding sovereign debt, so $500 bn is a small position, but still, it is a large enough number to take pause.

In my opinion, the CDS market can do some good things, but I think it can also lead to some strange behavior.   I will argue today in the case of Sovereign Debt CDS, it may lead to the mis-perception of risk.

It is useful to remember that Credit risk is not like Market risk. Market risks have been subject to risk analysis and measurement which has a random normal assumption. Even though this has been hotly debated, the general sense that markets go up and down with equal likelihood has an intuitive appeal despite the statistical inaccuracies. The purpose of my note is not to start a debate of market risk, nor to advocate for random walks; rather, my goal is to contrast market risk with credit risk. Think in lay-person terms – I buy a government bond and hold to maturity. Either I get a small gain (especially these days) from the interest, or there is a default and I may lose the principal.  Now, let’s leave aside currency devaluation, inflation and other more subtle forms of debt destruction to keep it simple. The facts are that taking on credit risk results in a relatively smaller dispersion of gains compared to losses, and that is the main point I want to highlight – Credit Risk has some low probability, but really bad outcomes in comparison to its upside – interest payments.  Many would agree that credit risk follows a Poisson distribution, not a random normal distribution. In simple terms, when we apply this distribution to the dispersion of returns on credit risk, it means that there is a large “tail” or a lot of “remote” chances for very large losses whereas the upside gains are limited.

Enter the Credit Default Swap. Thanks to this derivative contract, I can purchase insurance against the uglies.  I can buy “protection” from a bank to pay me in case the government defaults on its bonds.  So, I can cut off that nasty long tail that comes along with credit risk. Again, my long-winded note is to highlight that this kind of thinking can lead to a mis-perception of risk. While the insurance has a nominal coverage in case of the bond default, one must wonder whether a CDS would truly provide the protection in case there really were a sovereign default. One hypothesis is that, in the event that a sovereign default were to happen, then uncertainty would be too great and the markets would be in too much systemic turmoil for these contracts to be useful.

As we learned with AIG, when I try to chop off default risk, I get counterparty risk in return – ie., I get the risk that the company from whom I bought the swap protection cannot make good on their promise. Today, many swap counterparties have the implied backing of…the government. So, if the government defaults, then the total credit rating of the financial contract must also decline because the counterparty’s “backer” has declined in credit quality. I am sure my previous statement can incite many explanations of my erroneous assumptions, but in my opinion, moral hazard is moral hazard. We have entered the abyss and banks have government backing unless and until proven otherwise. So, even if it is at a minute level, aren’t banks betting against themselves when they sell a CDS on sovereign default? (I can hear the boos, screams and hisses).

Back to the credit default swap…In cases where a counterparty makes a financial commitment that we feel may have a risk of being too great, we require that the customer put up collateral for the guarantee. The CDS market has a lot of collateralization to support its swaps, especially since the credit crunch. Now, in case you have not already guessed, in many cases the collateral supporting swaps is Government bonds.  To recap:

• We want protection against sovereign default, so we buy a CDS
• We buy it from a bank, whose credit is tied to the Government (too big to fail)
• And we collateralize the default insurance with bonds which, if the default happens, would be completely useless in terms of providing us something liquid to sell.

Also in today’s FT, Gilian Tett opines that outright default is not the issue, but rather the more subtle losses from sell-offs (significant price declines), inflation, and devaluation - “Will Sovereign Debt be the new Subprime?” – FT, 11/23/09.   While I agree with Mr. Tett, my simple point is about mis-perception of risk.  Clearly the risks Mr. Tett outlines are bad enough, and, as he suggests, “default seems highly unlikely” in the case of sovereign bonds.  In fact, throughout all cases of Credit Risk, a default seems unlikely (especially at inception of the transaction).

From a fundamental view, I am not sure how one can opine on the probability of sovereign risk default in the US, UK, Japan or other large debt market.   Measures such as Debt as % of GDP are too crude to really assign a viable ranking to the risk, but somehow the default risk can be measured in basis points.   By looking at the market, it seems like it is a little more than 30 basis points this month (Nov. 2009).  So, for 30 basis I can “hedge away” the default risk of my US Treasury note and forget about the default risk (on my “risk-free” bond…). Now, prudential risk management practices would suggest that, even if the default is unlikely, one should take the moment to contemplate the actual implications of a sovereign default.  Here is my simple view:

• If there was REALLY a sovereign default in one of the indebted “Industrialized” nations; quite simply, all rules would break loose.  I remember how bad the market behaved when the small Russian GKO market melted down – a US or UK default would be almost unimaginable, but let’s try…

• With the news of a sovereign default, the counterparties providing the protection would be under distress. Besides having their backer going bust, they are probably chock-full of bonds in their portfolios too.

• The validity of swap collateral would come into question, especially if it was sovereign debt (heretofore considered the safest of safe).

• Other asset markets would have massive volatility, illiquidity, and general panic.

• During this calamity, I am not sure how much solace (or protection) would come from the CDS.

Again, my point is that there can be a mis-perception of risk of a sovereign default.  It may cause you to wonder – do these swaps create a systemic link that would cause even more disruption than the sovereign default?

Let’s not go there….today