mathbabe

My friend Nathan recently sent me this Credit Writedowns article on the markets in German risk. There’s basically a single, interesting observation in the article, namely that 5-year bond yields are going down while credit default (CDS) spreads are going up.

When I was working in risk, we’d use both the bond market prices and the CDS market prices to infer the risk of default of a debt issuer – mostly we thought about companies, but we also generalized to countries (although mostly not in Europe!).

For example with bonds, we would split up the yield (how much the bond pays) into two pieces. Namely, you’d get some money back simply because you have to wait for the money, so you’re kind of being compensated for inflation, and then the other part of the money is your compensation for taking on the risk that it might not be paid back at all. This second part is the default risk, and we’d measure default risk of a company like GM in the U.S., for example, by comparing U.S. bond yields to GM yields, the assumption being that there’s no default risk for the U.S. at all. Note this same calculation was typical in lots of countries, but especially the U.S. and Germany, which were considered the two least risky issuers in the world.

With the CDS market, it was a bit more complicated in terms of math but the same idea was underneath – CDS is kind of like an insurance on bonds, although you don’t need to buy the underlying bonds to buy the insurance (something like buying fire insurance on a house you don’t own). The amount you’d have to pay would go up if the perceived risk of default of the issuer went up, all other things being equal.

And that’s what I want to talk about now- in the case of Germany, are all other things equal? I’ve got a short list of things that might be coming into play here besides the risk of a German default.

• Counterparty risk – whereas you only have to worry about Germany defaulting on German bonds, you actually have to worry about whomever wrote the CDS when you buy a CDS. Remember AIG? They went down because they wrote lots of CDS they couldn’t possibly pay out on, and the U.S. taxpayer paid all their bills. But that may not happen again. The counterparty risk is real, especially considering the state of banks in Europe right now.
• People might be losing faith in the CDS market. There’s a group of people who call themselves ISDA and who decide when the issuer of the debt has “defaulted”, triggering the payment of the CDS. But when Greece took a haircut on their debt, it took ISDA a long time to decide it constituted a default. If I’m a would-be CDS buyer, I think hard about whether CDS is a proper hedge for my German bonds (or whatever).
• As the writer of the article mentioned, even though it looks like there’s an “arbitrage opportunity,” people aren’t piling into the trade. Part of this may be because it’s a five year trade and nobody thinks that far ahead when they’re afraid of the next 12 months, which is I think what the author was saying.
• There are rules for some funds about what they are allowed to invest in, and bonds are deemed more elemental and therefore safe than CDSs, for good reason. Another possibility for the German bond/ CDS discrepancy is that certain funds need exposure to highly rated bonds, so German bonds or U.S. bonds, and they can’t substitute writing CDSs for that long exposure.
• Finally, in the formula for how much big a CDS spread is compared to the price, there’s an assumption about how much of a “haircut” the debt owner would have to take on their bond – but this isn’t clear from the outset, it’s determined (as it was in Greece) through a long, drawn-out, political process. If the market thinks this number is changing the spreads on CDS could be moving without the perceived default risk moving.