The Global Mispricing of Risk

In previous posts I showed that debt and debt derivatives have expanded at explosive rates over the past few years, caused by an unprecedented under-pricing of credit risk. A BBB+ rated corporation can now theoretically borrow at just 32 basis points (0.32%) over what the US Treasury pays – according to where CDS’s are trading right now. Actual costs may be higher, however, because it is one thing to actually borrow a billion dollars in cash and quite another to sell a $10 million credit swap for $32,000. Yet, this last tiny transaction is deemed as indicative of actual risk as actually borrowing a billion - that’s the effect of “marking to market”.

Despite the explosive growth in notional amounts outstanding in the CDS market (according to ISDA, $26 trillion on June 30, 2006 and likely $35 trillion at year end), the market value of all such swaps was around $375 billion or 1.45% of the notional on June 30, 2006. This percentage has now dropped to probably around 1%, as risk premiums have since come down at least 35-45 bp for the whole market (i.e. average of investment plus non-investment grade). This means that 100% of debt, the whole of the bond market, is being marked to market and priced by swaps that cost a tiny 1%. It seems derivatives are now pricing bonds instead of the other way around, i.e. the tail is wagging the dog.

Apart from price, there are also serious quality concerns. Who is selling this credit insurance is even more important than how much they are selling it for. What good is it to pay just $100 for fire insurance if the insurer cannot pay in the event your house burns down? Credit swaps may be priced against US Treasuries or German Bunds (in Europe), but it is not the AAA+ US or German government that is the insurer. It is hedge and private equity funds and large banks that sell default insurance – what are their credit ratings? And how much have they reserved for potential losses against the trillions in insurance they have sold?

Unlike property and casualty insurance, this is essentially a totally unregulated business. A hedge fund can sell as much insurance as it can convince someone to buy from it – there are no statutory limits or reserve requirements. Flip the premium numbers upside down: by accepting just $320,000 per year a hedge fund has the obligation to come up with $100.000.000 in cash to fulfill its insurance obligation in the event of a default by what is today a BBB or A- borrower: this translates to 312-to-1 leverage. For insuring B rated “junk” bonds the leverage is “just” 44-to-1. How likely is it that all, or even most, hedge funds and other CDS sellers will come up with such enormous sums of money if real defaults occur? Remember that hedge funds and private equity firms have zero reserve requirements. They are the weakest link in what is already a very stretched chain.

What is even more concerning is that it is not even necessary for defaults to occur in order for the chain to snap. There are now tradable CDS indexes that can be bought or sold on the aggregate of 100 corporate names. Such index trades are margined and settled for cash: if the market goes against the speculator, he has to come up with money regardless of any default. If anything happens to suddenly raise the overall level of risk perception, margin calls alone could wipe out many players and even result in crushing losses, if margin calls are not met, for whoever took the other side of the trade (banks and investment banks?). According to BIS, about one third of all credit swaps are now against such multi-name instruments.

In the past it used to be that we measured risk perceptions indirectly by the exuberant or morose behavior of primary markets and their relative performance: stocks, bonds and commodities. We have now constructed a market for pricing and trading risk itself and it is telling us there is very little risk out there. Are we to believe it? I don’t know, but there is an obvious oxymoron here: the CDS market that is right now proclaiming that risk is very low is by itself creating more risk.