Bailout

The US default-risk meme

By Felix Salmon

There’s a constant drumbeat of stories about how the price of credit protection on the USA says little if anything about America’s creditworthiness: Dan Gross had one just last week. I’ve written much the same story myself, but at least I tried to explain what was actually going on in that market; what I’m still waiting for is a journalist who can find someone who’s actually buying or selling these things, so that we can find out from the horse’s mouth why they’re doing so.

What’s new is stories looking at yields on Treasury notes and extrapolating default probabilities from those. Bloomberg had a headline on Monday saying “Obama Pays More Than Buffett as U.S. Risks AAA Rating”, which led with this:

The bond market is saying that it?s safer to lend to Warren Buffett than Barack Obama.

I’m pretty sure it’s safe to say that sentence simply isn’t true. For one thing, it’s predicated on the idea that Berkshire Hathaway bonds were yielding 3.5bp less than equivalent-maturity Treasuries, “according to data compiled by Bloomberg”, but when you start looking at gaps that small, you have to be very careful with the reliability of your data, and I suspect that the people at Bloomberg weren’t. In the comments over at Marginal Revolution, Ricardo pulls up all the actual trades of the Berkshire bond in question on March 18, the day that Bloomberg says it yielded less than Treasuries; there are seven of them altogether, starting at 1.206% and finishing at 1.023%. All but one are above the 0.93% that Bloomberg says Treasuries were yielding, and it’s not clear how much Treasuries were yielding at exactly that point in the day.

But that doesn’t stop Bloomberg from taking one suspect datapoint and constructing a large edifice of rhetoric atop it, talking about “concerns whether the U.S. deserves its AAA credit rating” and so forth.

Today, it’s the 10-year swap spread, which has followed the 30-year swap spread into negative territory, and suddenly everybody’s talking about “the fiscal situation in the US” and how “increased sovereign risk” is the culprit.

Well, it might be (or it might not be: Bond Girl has good non-default-risk reasons why this might have happened) but I’m going to need a hell of a lot more evidence before I start thinking of the Treasury market as including a measurable or even extant premium for default risk. In theory, it might well be there, and indeed the existence of that CDS market on US debt is prima facie evidence that it exists. But if it’s there, there have got to be better ways of finding it than by pouncing on curious anomalies in the bond market.

Full post as published by Macroprudential on March 24, 2010 (boomark / email).