28 May 2007

So I Can Refer To It Over At The Daily Duck

The yield curve on US Treasuries as of 5/25/07.

This chart shows the 6 month, 1, 2 and 3 year rates daily from January 3, 2005 through May 30, 2007. The inversion first occurs at the end of December 2005, widens throughout 2006 and now seems to be closing.


Hey Skipper said...

The x-axis is time?

If so, the dates look a little funny to me.

David said...

How so?

Susan's Husband said...

He's probably wondering what "01/01/37" mean as a date — 1 Jan 2037? If so, presumably the curve is annualized yield vs. maturity date for the security?

joe shropshire said...

Yeah, that's the curve out to 30 years. I don't follow this that closely, but IIRC a normal inverted yield curve is when short term (<5 year?) yields are higher than the 20 year. As David says over on the Duck, that's a funny-looking curve.

David said...

SH: You and Joe have it. The source, which is linked, is the Treasury's daily yield chart.

Oroborous said...

Cross-posted at the Duck:

I think that the yields are inverted. A couple of economic/business analysts/economists whose opinions I respect think that the best comparison is the spread between the 1yr vs. the 10yr, because the 20 and 30yrs attract a lot of money from pension funds and insurance companies who are looking for safety above all else, which, as David notes, skews the yields a bit.

By the 1yr/10yr standard, there is an inverted yield curve.

Further, look at the spreads for all maturities. The 2yr and 10yr are basically at par, and the biggest spread between the 3, 5, 7yrs vs. the 10yr is a paltry 6 basis points.

Also, the 6mo. pays more than the 1yr, which pays more than the 2yr, which pays more than the 3yr, which pays more than the 5yr...

To show further how screwy the yields are, here are some historical data points, from non-recessionary years selected at random (notice the orderly yield ladder):

............1mo 3mo..6mo 1yr..2yr...3yr..5yr
05/27/97 N/A 5.18 5.50 5.89 6.30 6.47 6.64
05/25/05 2.67 2.93 3.19 3.34 3.64 3.75 3.94
05/25/07 4.97 4.88 4.97 4.93 4.85 4.81 4.80

.............7yr..10yr 20yr..30yr
05/27/97 6.73 6.79 7.12 7.03
05/25/05 4.08 4.26 4.65 N/A
05/25/07 4.82 4.86 5.09 5.01

Hey Skipper said...

If so, the dates look a little funny to me.

Whereupon I expose the stygian depths of my treasuries ignorance.

Now further exposure, presuming such is possible.

If the yield is inverted, why, pray tell, would anyone bother purchasing the longer term security?

Oroborous said...

...why, pray tell, would anyone bother purchasing the longer term security?

One reason would be if you're investing with a long time horizon, and you think that interest rates are going to fall sometime in the next few years.

So, if you're going to be holding money for ten years, while you might be able to get an extra 7 - 11 basis points by buying for one year or six months, you run the risk of having to roll the money over into something paying less than today's 10 year.

Plus, if rates do go down, then the long bonds with the higher rates will have a higher face value, should you desire to sell them.

Oroborous said...

higher face value

Rather, higher market value.

Susan's Husband said...


Only semi-ignorance. Your question is precisely why an inverted yield curve is considered anamolous and interesting.

Oroborous said...

To state it more clearly:

In normal times, bond issuers pay a premium to investors to be allowed to hold the money for a longer time.

In times of economic slowing, investors pay a premium to bond issuers, to be permitted to lock in a certain rate for a longer time.