Zerohedge (Tyler Durden’s nom de finance) was the first to notice this. At least he was the first I saw. The 30 year Treasury bond auction June 10 was as close as we’ve come to a failed auction in my memory. Let Mr. Hedge explain in his own words:
After yesterday’s stellar 10Y auction, which saw the 5th highest Indirect take down on record, today’s reopening of $22BN in 30Y paper (via Cusip UU0) was a mirror image: ugly, poor foreign demand, and tailing.
Starting at the top, the auction priced at a high yield of 5.02%, down fractionally from 5.046% last month (which was the first 5% coupon auction in history). And just like last month, today’s auction also tailed the When Issued 5.008% by 1.2bps; this was the third tailing 30y auction and was also the biggest tail since August 2025.
Our pals at Charles Schwab provide a nice translation:
What’s the score?
The end of a Treasury auction, like the end of a game, comes with its own version of a box score. There are a few terms that show up in the results that investors should know.
Bid-to-cover ratio: Measures demand at the auction by comparing the number of bids received relative to the amount of Treasuries being auctioned. A higher ratio generally suggests stronger demand.
The tail: Measures the difference between the yield awarded at auction and the yield implied by pre-auction trading. A positive tail can indicate weak auction demand that required the government to offer a higher yield to attract buyers. This can cause Treasury yields to rise when the results of the auction are released.
Primary dealers: Measures the percentage of debt bought by primary dealers, who buy the portion of the offering not purchased by other investors. A high percentage here can also indicate weak auction demand.
The Treasury Department regularly publishes auction results on its site and lets investors look back at recent history to see what each one yielded
Back to Zerohedge
This graph looks at auctions since 2019. The green bars mean robust demand with no tails. The red bars are tails; the most recent was 0.01250. (The graph’s y axis is inverted. Values above the x-axis are negative, indicating robust demand. Below the x-axis are tails as positive numbers meaning demand is less than forecast.)
Fig01 (click for larger image)
The next graph is confusing (to put it mildly).
Fig02 (click for larger image)
This graph violates Dr. Lima’s law of time-series graphs: never use a single graph to measure two, wildly different topics. The left axis is percentages. The right is a ratio (the BTC bids received divided by auction volume) and the high yield (%). This compounds the felony by putting two different units of measure on a single axis: a ratio and a percentage. This only works if you write 5.68% as 5.68 instead of 0.0568. Mr. Durden has been in this business long enough that he can afford to have three understandable graphs instead of one that creates total confusion.
The two jagged lines are the bid-to-cover (red) and the highest yield (yellow) for leach auction. The y-axis for those two lines is on the right side. The bar graphs are measured on the left axis.
Possible Causes
- Inflation Expectations
What caused this? One possibility is recent inflation numbers. (Remember, the nominal interest rate equals the real interest rate plus expected inflation. Unexpected higher inflation can lead to higher inflation expectations.)
[The graphs below are linked directly from FRED. They are dynamic and interactive. I see the bottom two lines have overwritten each other on several graphs. This is a mystery, as the code seems identical for all of them. Will try to fix that when I have time. If you spot other bugs, please let me know.]
First, let’s look at our old pal, the CPI. Here’s the CPI level:
Fig03
And here’s the year over year growth rate:
Fig04
The PPI (Producer Price Index) is the average price of intermediate products, those used in the creation of final goods. It’s no less forgiving.
Fig05
And here’s the growth rate:
Fig06
Well, that’s bad news. Inflation is a contributing factor. What else might it be?
- High government debt
We’ve talked about this before (here and here among many other articles). Government debt is a stock. The debt increases when the government runs budget deficits. Debt decreases when the government runs surpluses. For you youngsters, the U.S. government has, in fact, run surpluses in the past. Since 2010 there have been quarterly surpluses in 2013:I, 2014:I, 2016:I and 2017:I. But for nine long years we have seen nothing but deficits.
Fig07
Even those few surpluses did little to stop the explosive growth of debt:
Fig08
Deficit spending promotes nominal GDP growth. So naturally, the debt-to-GDP ratio has stabilized around 120%.
Fig09
Addendum: The Pattern of Surpluses
FRED>Keen-eyed observers will have noticed that all four of the surpluses since 2010 were in the first quarter. That’s caused by the U.S. income tax code. April 15 is the deadline for tax filing. Those who expect to owe the government will incur interest and (perhaps) a penalty for filing after that date. Therefore, they file early, including payment of what they think they’ll owe. Those payments fall mostly in the first quarter. This seasonal pattern has been around for as long as the magical date April 15 has been in effect
Addendum 2: FRED Problem
Regular readers will notice that I’m using graphs from the FRED database at the St. Louis Fed. Working on this article, I’ve discovered an interesting problem. Sometimes when you should see this:
Fig10 Caption Correct
you see this instead:
Fig11 caption messed up
I’ll let the St. Louis Fed people know about this. I’m pretty sure it’s a mistake.
