Yield Curve Gets Serious: 10-Year to 7-Year Treasury Spread Collapses to 4 BPs

Mike Mish Shedlock

JP Morgan is following yield spreads on the GBI broad bond index. Things are getting serious says one analyst in a Tweet

That Tweet by Holger Zschaepitz got me thinking more about the flattening US treasury yield curve. Most analysts follow the 10-year to 2-year spread. The Tweet mentioned 10-year to 7-year and 1-year to 3-year spreads. Here are some charts I created in Fred.

10-Year Minus 7-Year Spread

This portion of the yield curve looks highly likely to invert soon. That said, predicting recessions off such divergences is clearly problematic.

3-Year Minus 1-Year Spread

This is another unusual spread to watch. It's currently at 34 basis points. Assuming the 1-year yield rises 25 basis points on a Fed hike (probable), but the 3-year yield doesn't (questionable), the Fed could get in no more than one more hike before this portion of the yield curve inverts.

10-Year Minus 2-Year Spread

This chart is so popular that Fred has it precalculated. I created the first two charts and the following chart using Fred tools.

The last three recessions all started with 10-2 spreads higher than 35 basis points. Thus we are already well into recession territory. The one missing ingredient is a prior 10-2 inversion.

10-Year Minus 1-Year Spread

As long as we are investigating unusual ranges I thought I would post one of my own. The "Great Recession" started with a spread 14 point higher. The Dot-Com recession started with this spread 11 points lower, and the July 1990 recession started with the spread 17 points lower.

This spread went negative at some point before each of the last five recession. Unlike the 10-2, 3-1, and 10-7 this indicator gave no false positives an it also gave plenty of warning.

Missing Warning Signal

Will we have a recession without the yield curve inverting?

I don't know, nor does anyone else.

Fundamentally, there is no reason to believe an inversion is a necessary ingredient for a recession. Japan offers proof enough unless we are to assume Japan is different.

One strong reason to suspect we may be similar to Japan is the length of time the Fed held rates close to zero. That certainly was different. Tests of zero-bound constraints are different as well.

Everyone Guessing

Everyone is guessing. But interestingly, nearly everyone seems to be guessing the same way: There will be a warning signal.

What if there isn't?

We may find out soon enough. If the Fed gets in two more hikes, I suspect that some portions of the curve will invert.

But even so, I also suspect the vast majority of investors will ignore that signal too.

In fact, I know they will. Mathematically they must. For everyone trying to time the top, there will be a dip buyer buying the dip. Mathematically, there cannot be a stock market escape in aggregate.

Mike "Mish" Shedlock

Comments (10)
No. 1-10

I recently bought a Certificate of Deposit. It seems the sweat spot being offered by almost everyone is 1 year. The 1 year had a better APR than the 5 year at many banks. I'm assuming it's because many think the FED will raise rates for a period of roughly a year before loosening again.


So “if” history repeats, the next recession is 2-3 years away because the “spread” will have to dip below the dotted line for 1-1.5 years, then turn back up for 1-1.5 years before touching the dotted line again. All very interesting analysis till you look at the first two charts you created 10vs7 and 3vs1, which show no discernible pattern.


The longest U.S. economic expansion is 10 years. This month we are sitting at year 9. The clock is ticking. Time is slowly running out. In 2008, 2018, was a decade away. The future is now. They decided not to fix the mess in 2008. They decided to create a bigger mess, going forward. The bigger mess is now near the doorstep.


Central bank strategists have this (everything)under control, BY JUST PRINTING MOAR MONEY!!!


The stock market tends to peak about the time of the inversion, and then falls harder as the recession hits. The market then bottoms about at the time these charts peak. The best time for the market is the time when the rate differentials are falling.


We are no longer in business cycles but credit cycles (quote from Peter Boockvar). Anything who thinks the economy can handle a couple of interest rate increases is kidding themselves. Once credit starts drying up, good assets will have to be sold to pay for losses in bad assets. The trade situation isn't helping the certainty that businesses need to operate smoothly either. I suspect when we look back on late 2018 sometime in 2019, experts will say the recession started in late 2018. The bottom will drop out when tax receipts decline in earnest b/c government won't have money for anything other than interest payments, social security, medicare and defense and even defense will likely get cut. Right now it is impossible to see from what corner or where things will go south. This isn't like 2000 or 2007 in that sense. I see something more like the Asian financial crisis of the late 90s where problems in one country caused contagion to most of Asia. This time it will be on a global scale as all assets are linked globally.


Mish said: "Will we have a recession without the yield curve inverting?"

As I understand it, yield curve inversion blows up short-long carry trades. I think that is the financial tide alluded to in Warren Buffett's quotation, "You never know who's swimming naked until the tide goes out." Until then, rolling over short term debt to finance long term obligations can continue.


We've all been discussing this inverted yield curve signal for years. Can't dispute its correlation to previous recessions but nothing is a lock or as simple as "watch this indicator that everyone else is watching." And the mid 90s really cause a lot of questions, don't they? We inverted then quite a few years before a recession. Also, what about the lead time to the inversion? We've been creeping toward this narrowing and if we invert, it will be through that slow "creep". Previous moves have been much more abrupt. Also, the 90s were coupled with extreme valuations (above current) and 07-09 was called the Great Financial Crisis for many reasons - none of them being that we had experienced an inverted yield curve. I'm definitely not suggesting that it doesn't matter - not by a long shot. I just don't know that it's as simple as many are making it to be. Hope to get your thoughts!

Global Economics