Slog News & Arts

Line Out

Music & Nightlife

« Ted Tripp, the Paradox, and Ho... | Savage Love Letter of the Day »

Monday, September 22, 2008

A Visit to the Financial Alamo: The US Government Treasury Bond

posted by on September 22 at 13:35 PM

(Borrowed from today’s New York Times.)

You need to understand this chart. Given our government will be borrowing a trillion or two dollars (about a tenth or a fifth of the entire economic output of the United States in a year), you should understand how expensive this is, right now.

Start with the dark blue line. This is the annual yield the US Government (er, us the taxpayers) must pay to borrow money for three months, six months, two years, five years, ten years and thirty years.

As the length of borrowing goes up, the yield the government must pay goes up. This makes sense. If you’re going to tie up the investor’s money for longer, you should have to pay more. Why?

1. If the economy does better in the future compared to right now, interest rates are likely to go up. Therefore, convincing investors to give up their money now, for a long period of time, means you have to do a bit of bribery and pay a higher yield. (This is the optimistic interpretation.)

2. Investors might be concerned about the long-term financial health of the US government. Such investors will be more willing to lend for short periods of time, and increasingly anxious about longer commitments. If investors think the financial health of the US is at risk, they’ll expect to be paid to accept this risk.

Well, my friends, the yield curve today is indeed steep.

The yield on a 3-month bond is incredibly paltry—perhaps even less than the current rate of inflation. A yield like this makes the 3-month (and even the 6-month) bonds the equivalent of the big global mattress, where everyone is stuffing their cash—not to make it grow, just to keep it from being lost.

The fact that the short-term yields are so low, despite the massive increase in government borrowing, tells us just how spooked global investors are. Whatever money there is to be invested right now, much of it is going to the big mattress.

A year ago, the curve was really flat. (Look at the gray line at the top of the chart.) Flat—or inverted curves where yields are higher for shorter term commitments—often proceed times of economic upheaval.

So, people are still lending money to the US government for now. Maybe because they think things will get better soon. If the cost of long-term debt starts creeping up, it might be because they thing it’ll be much much worse in a few years. Should be fun to see which proves to be true.

Or, to quote our fourth branch of government, Dick Cheney, “Reagan proved deficits don’t matter.” We’ll all see, Dick. We’ll all see.

RSS icon Comments


People are still lending long term too. Look at your own damn graph. The yields on longer term T-bills are not up, there down, albeit less than the shorter term ones.

People buy long term debt for the long term and short term debt for the short term. Given the problems with everything, short term treasury debt is one of the few safe places around. This ups demand and lowers price.

We have a sharp yield curve not because of changes in long term outlooks vis-a-vis the US government or even economy, but because of a hug spike up in demand for safe short term debt.

Posted by Giffy | September 22, 2008 1:42 PM


I agree. That's why I included the graph, and went on to use the mattress metaphor. Was I not clear that it's #1 for now?

Posted by Jonathan Golob | September 22, 2008 1:47 PM

And you haven't even begun to question how the municipal bond and corporate bond markets are going to react. Cities, states and companies have to borrow to fund their plans and projects, and now they have to compete for funds with the US Treasury. Throw those notions in the blender and mix.

Posted by Cranky Old Man | September 22, 2008 1:51 PM

So, people are still lending money to the US government—at least for short time commitments—for now.

That's what I am objecting too and your entire second claim. There has been increased demand for short term debt and no decrease (in fact a small increase) in demand for long term debt. Both are historically low.

So there is not evidence from yield curves to support your second claim. None at all.

Posted by Giffy | September 22, 2008 1:59 PM

Ok Giffy.

I modified the post to reflect this and let people know what to look for if things go south.

Posted by Jonathan Golob | September 22, 2008 2:02 PM

Go get that guy Malarkey to settle it.

Posted by PC | September 22, 2008 2:03 PM

@5 Cool. You might want to take a look at this graph as well^TNX&t=my&l=on&z=m&q=l&c=

When we hit 10 or 15 then its getting time to freak.

Posted by Giffy | September 22, 2008 2:04 PM

Dick won't see. He'll be dead by then and couldn't give two shits.

Posted by Smade | September 22, 2008 2:05 PM

you could have just used two words to describe why long term interest pays more than short term; time preference. people should be able to figure it out from there.

Posted by Bellevue Ave | September 22, 2008 2:28 PM

Your graph shows that the yield curve was, and still is, almost flat from 10 years to 30 years. There seems to be no support for your second claim, that investors are worried about the US being able to pay its obligations, unless for some reason many think the risk of default GIVEN the government doesn't default through 2018 is almost zero, but the risk of default before then is greater.

Posted by Andrew | September 22, 2008 2:31 PM

I'm too lazy to do this, but you could determine the hazard function that these data impose under the 2nd hypothesis (greater risk of default over longer time-horizons). I can tell right now that it won't be constant (which would be the result of the most parsimonious hypothesis you can make: a poisson process) since that gives you a exponential distribution for mean time between failures, which would be a straight line on the log-scale you've got. Would be an interesting bit of geekery.

But anyway, posting the the graph and implying that it supports your second hypothesis is pretty unscientific.

Posted by Andrew | September 22, 2008 2:40 PM


Yes. For now US Government debt is selling well. Actually, it's selling better than usual. That means investors are pulling money into treasury bonds--bad for the overall economy.

The curve is asymptotic. If the asymptote increases, then investors are losing confidence in the debt. It hasn't happened yet. It might not.

The slipping of the dollar today--against almost all other currencies--is a worrisome sign.

Posted by Jonathan Golob | September 22, 2008 2:44 PM


It's "Dear SCIENCE"

not "Dear ECONOMICS"


Posted by Non | September 22, 2008 2:46 PM

Actually, the yield on 3-mo T-bills last Thursday was 0.02 percent. Only marginally better than the cost of hiring some thug with an AK-47 to guard a pile of cash in a locked room.

Posted by Toe Tag | September 22, 2008 3:34 PM

Non @13--thou weed of dill--economics is a branch of social science, well within the Golob purview. I'm awfully glad it is.

Posted by tomasyalba | September 22, 2008 3:39 PM

"The slipping of the dollar today--against almost all other currencies--is a worrisome sign."

Perhaps. Did you see what happended to the price of oil today? The price of oil and
US dollar have recently been closely tied.

Posted by Cranky Old Man | September 22, 2008 3:45 PM

Wait a minute. The Treasury Bond hasn't been issued since 2001. T-Notes and T-Bills are still being issued, so this must reflect the secondary market. Bond bought at a discount have a higher yield. The increase in yields is bad news, because it reflects investor confidence, but only with T-Bills and T-Notes does it have a direct effect on our borrowing. T-Bills are sold by the government at a discount to face and redeemed at face, and in that respect, we're making out. People are so all about the security that a $1,000 T-Bill that ten years ago would have gone for $930 is now going for $990. T-Notes on the other hand, the interest we have to pay is going up, and that's fucking with us.

Posted by Gitai | September 22, 2008 4:39 PM


You, as usual, are totally right.

I didn't want to get into explaining the difference between T-Bills, T-Notes. I'm being a bit slippery by using the secondary market to indicate the confidence of investors in the US government's solvency. T-Bills are a better indication, but a bit trickier for me to explain.

I happen to really like the TIPS and the Series I savings bonds, but that's a discussion better for another time.

Posted by Jonathan Golob | September 22, 2008 5:29 PM

@17 I thought they began reissuing them in 06.

Posted by Giffy | September 22, 2008 9:10 PM

Comments Closed

Comments are closed on this post.