The US default-risk meme

By Felix Salmon
March 25, 2010
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.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

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.

Update: Mark Gongloff has a good roundup of the various different explanations for negative swap spreads; sovereign default risk doesn’t even make the cut.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Is the swap-spread meme assuming that the rate differences between equal maturity swaps treasuries reflects ONLY default risk? Shouldn’t we also calculate any embedded options?

After all, what about prepayment risk? Treasuries are callable aren’t they?

Posted by Sprizouse | Report as abusive

Let’s ask some of the characters in Michael Lewis’ book, and Michael Lewis himself. They are the ones that can see the shapes in the mist. And there’s always that Bill Gross character. See if he’ll take your call.

Posted by Uncle_Billy | Report as abusive

Bloomberg articles tend to be written by 25 year old Ivy Leaguers who couldn’t find work in an investment bank, and tend to remain somewhat bitter about that fact. As regards their journalism, the gap between level of expertise and breadth of readership is surpassed only by The Economist.

Posted by dewey2999 | Report as abusive

History has shown that not too many revolutionary governments honor the previous government’s debts.

(This is the only way I could come up with a plausible default risk for the US.)

Posted by high_al | Report as abusive

1. US treasury rates are usually below the risk-free rate. The liquidity premium you pay is higher than whatever default premium you receive. In the typical IR environment of the last 25 years, swap rates are not as far above risk-free (arithmetically) than treasuries are below.

2. In any case, have you bothered to compare swap spreads with with sub-debt bond spreads from the same issuer? A true anomaly would be if the swap spread were not lower than the bond spread. The reason is that (absent disguised loan tricks) the credit risk of the bond is much greater because of the exposure to principal. With the swap, half the time the other side has exposure to you.

3. It follows that you really want to compare bank vs government credit (still a pointless exercise in my opinion), you have to compare like for like: swap vs swap or bond vs bond. (CDS vs CDS won’t work because the terms aren’t comparable e.g. EUR payoff from EUR bank in the case of the sovereign.)

Bond Girl has some speculation on the technical mechanics of putting on an arb but no matter how you do it, you will founder on this problem of credit risk.

Posted by Greycap | Report as abusive

Greycap, the swap settles off of 3m libor, so the payoff incorporates 3m bank credit risk, independently of any counterparty risk in the swap. Besides, interest rate swaps between dealers are margined and have very little counterparty risk.
And if you add the funding angle (you can borrow against treasuries at below libor rates) the effective spread between the treasury rate and the swap rate is even greater than the quoted one.

Posted by niveditas | Report as abusive

Thank you for sharing excellent informations. Your web site is so cool. I am impressed by the details that you’ve on this site. It reveals how nicely you perceive this subject. Bookmarked this website page, will come back for more articles. You, my friend, ROCK! I found simply the information I already searched all over the place and simply could not come across. What a great web-site.