The yield curve went through its longest period of inversion ever and is now on the cusp of uninverting. Some say this is a recession trigger. Let’s investigate.
Note: I wrote this post last night. The inversion steepened today but that does not change the following analysis.
What Is an Inversion?
In a normal yield curve longer-term Treasuries have higher yields than short-term treasuries.
The above chart shows the opposite. Short-term yields are higher than long-term yields. Numbers below zero means inversion.
People go on recession watch when yields invert. But the wait can be long. Others say the recession signal is when the yield curve uninverts.
Steepening Yield Curve
The Gap Shrinks
Massive Move Today
No Recession?
This won’t lead to a recession because a recession has already started.
Q:Has that ever happened before?
A: Yes, at least twice.
Two Consecutive Recessions Without Uninverting
The trigger is not uninversion. None of these yield curve measures provide a good signal.
Dudley Changes His Mind, Says “Fed Needs to Cut Rate Now” to Avoid Recession
Yesterday I commented Dudley Changes His Mind, Says “Fed Needs to Cut Rate Now” to Avoid Recession
Citing the McKelvey recession indicator, former NY Fed President Bill Dudley, wants the Fed to cut rates now. It’s too late Bill, recession has started.
Although it might already be too late to fend off a recession by cutting rates, dawdling now unnecessarily increases the risk.
The best indicator of recession is a slightly modified version of the McKelvey rule. click on the above link for a series of six charts.
A recession has started but few see it yet.
2 quarters of declining GDP is NOT the definition of a recession, not even close.
We have had many 2 neg quarters of GDP with no recession and 1 recession with only 1 quarter (2 months actually) of negative GDP.
Second, recession would not start with the first quarter anyway. They are pinned to MONTHS not quarters. So one could easily start in June if April and May were strong enough.
Third, the NBER is the official arbiter of recessions and it also looks at GDI.
I am disappointed to see Mish readers who ought to know better by now spreading nonsense about what a recession is.
I resemble that remark.Their is no well defined definition according to the nber. I always understood that the two months of declining gdp was the traditional definition. The nber definition is opaque and may be subject to politization. To be a legitimate arbiter of a recession they need to publish defined parameters.
There are no parameters. Nor do I think there should be. If we had parameters, what happened in 2020 would not have fit them.
But I will offer an observation.
Two quarters of declining Real Final Sales (not GDP), is a sufficient but not necessary condition.
They do not necessarily have to be consecutive if the gap in between is small enough.
I highly doubt that has ever happened without there being a recession.
We recently has 2 quarters of declining GDP and I did state “no recession” because RFS was positive. I was correct on that. The difference between RFS and GDP is inventory adjustment which nets to zero over time. I suggest inventory adjustment be removed from GDP as RFS is the true bottom line estimate.
So instead of negative gdp but just a decline while still positive, could fit into a recession call?
I’m seeing a lot of this uninversion = recession calls by the perma bears. Whether you or I believe we are on the precipice of recession is entirely beside my following point on this matter:
I inverted yield curve signaling recession is usually from the long end falling less fast than the FFR and two year rates. Currently (since about July last year), the long end has been rising, while the FFR has stagnated and the 2 year is little changed. Long end rising definitely is not the bond market telling us recession is imminent. If anything, it’s the opposite.
If you are going to read the tea leaves, take some context with it. Like if UE were to rise from 2% to 3%, trigging Sahm/Beveridge/whatever despite being a healthy increase
Recession quickly following Presidential election, no matter who wins.
One of the traders was wondering if it was not the case this time since the Fed did it vs the economy?
Engage brain before linking yield curves to economic conditions.
Wikipedia credits Campbell Harvey for the inverted yield curve in 1986. His PhD dissertation at U. Chicago studied the term structure of interest rates as predictors of the business cycle. (you can read it here: https://people.duke.edu/~charvey/Research/Thesis/Thesis.pdf) BTW, his dissertation uses the word ‘inverted’ one time, and not in a yield curve context. Instead, he uses the ‘yield spread’ between short and long term rates.
Now, the basis of Harvey’s thesis was Irving Fisher’s consumption-based theory of interest rates (1907). Harvey’s conclusion:
“…While the real interest
rates do not appear to be strong predictors of real consumption growth, the yield
spread variable has some predictive power – especially over the final 20 years of
the sample.”
The yield spread was “…the spread between annualized real rates – a common measure of the term structure…”
Harvey’s graphs of yield spread and consumption growth are revealing with regard to synchronicity–see the last third of the dissertation.
Diverting from defining recessions using GDP, it might be more useful to look at what Harvey used for consumption. Specifically,
“…National Income and Product Accounts (NIPA) quarterly. consumption data..”
For a hoot, you might look at real consumer spending, or a proxy like Real Retail sales.
https://www.macrotrends.net/1371/retail-sales-historical-chart
The last three years are particularly revealing
“People go on recession watch when yields invert. But the wait can be long. Others say the recession signal is when the yield curve uninverts.”
Which says all there is to say about the economic relevance and usefulness of entirely arbitrary nothingburgers like yupvettions, blonips and recessions. Neither of which means ANYTHING real; instead serving solely to facilitate idiots dumb enough to believe picking random numbers is some form of “zkill” that clueless little them are magically better at than any other dice or lottery ticket.
LOL. When a recession hits you or your family, you’ll feel differently about how “real” it was. Talk to any of the millions of people who lost job and their entire net worth in 2007-2010.
Recessions are like a bit like major storm systems. Not everyone gets hurt in a recession, just like sometimes a storm is just a few hours of wind and rain. But both recessions and storms can do a lot of damage to a significant fraction of the people in their path.
Seems like the Yield Curve inverted when there was a liquidity crunch(massive rush to cash) and now the uninversion is a sign that the economy is now trying to get cheaper money into the system . Low rates= economic slowdown. It makes sense. The FED elongated these low rate phases to allow corporations to juice the stock market and borrow to buyback stocks-but really burned the candle at both ends. Either to help incumbents stay in power or keep the big boys flush with cheap money to buy assets for cash
Approx time from last un-inversion (2/10 spread) to recession start…. 2001 – 2 months, 2007 – 8 months, 2019/2020 – 5 months. If the 2/10 curve uninverts soon and we use history as our guide, then recession would start Q4 2024 or Q1 2025.
All of these inversion indicators are fraught with error.
1980 was 3 months late and 1981 was 2 months late
By any chance is this a stagflationary recession like those?
also 2+8+5-3-2= 10
10/5= 2 months on average
It’s heads I win tails you lose – didn’t we just get done discussing how the yield curve inverting meant recession last year? Now it’s uninverting that’s the problem?
So confusing!
“So confusing!”
Attempting to shoehorn some veneer of supposed “insight” onto random series always is.
#overfittingwithlimiteddata
We are in depression coz I don’t have job.
The definition of a recession, or when it starts, is purely an academic exercise. Predictive value for capital markets is all that matters, and the r-squareds here are lacking. Thoughts Mish?
The predictive value for labor markets is more important to more people than the predictive value for capital markets.
I said it before, I’ll say it again. The bond market has been broken since 2009 and continues to be broken. The yield curve is meaningless. Just buy stocks. Bonds are the new toxic asset that have zero chance of keeping up with “real” inflation.
you can buy short-dated corporate bonds that are investment grade. Thats pretty safe, unlike the US Bond market these days
Are we in recession : CASY made a new all time high last month.
Investors bet on recession. They crowd the front end. In Germany the middle caved in. The 7Y is the lowest. In the US the 5Y is the lowest. Gravity with Germany pulls the middle down. US 10Y – DET 10Y = 4.24 – 2.42 ==> almost 2%
The pig is wearing lipstick. The lipstick is starting to wear off.
https://www.conference-board.org/publications/how-are-US-recessions-defined
How Are US Recessions Defined?
Thanks for the link. The article as I interpret it basically states we have a lot of moving parts and somehow we will figure it out. Trust us. I am wondering if politics factor into their decision making sausage factory. I live in the metro detroit area and no indication of a slowdown at all. Two homes went up for sale in my sub and were gone in one day.
Lets see what Mr. Market does today with the silly GDP numbers. Does this mean that the Fed can hold the line until September – until next year? Heck all of the numbers are screwy because of the goofy inflation calculations but hey people just react to headlines anyway.
Figure the spike in the yield curve will fall back today and we stay nice and inverted.
No chance of a cut this month and no meeting in August. So, yeah unless an emergency meeting, September it is.
A recession is defined as two consecutive quarters of negative GDP That has not occurred. The economy may be slowing but that’s about all
NO!!!!!
That is NOT the definition of a recession, not even close.
We have hand many times 2 neg quarters of GDP with no recession and 1 recession with only 1 quarter (2 months actually) of negative GDP.
Well – in a textbook sense it is. Recessions are “called” by a group of self appointed boffins at the NBER. They can pretty much call anything they want a recession, but from a practical standpoint if my neighbor looses his job its a recession.
The classic:
If your neighbor doesn’t have a job, it’s a recession.
If you don’t have a job, it’s a depression.
But… what if you and your neighbor lose your jobs and take a different one at lower pay and benefits?
This was the traditional definition. I think this is why it’s important to ask what people mean when they forecast a recession. My definition is at least 1 quarter of negative GDP. That would mean 3 consecutive months of economic shrinkage. I say we aren’t there and won’t be there this summer. If we keep making it easier by changing the definition to suit some narrative then this becomes too political. I would suspect most here would be saying the opposite if a Republican were president. Likewise with hard-core dems but those people don’t even come here. Even middle of the road independents like me recognize that there is too much politics already here. In no world is positive GDP a recession unless you have let politics play a role in your definition.
GDP isn’t an ideal metric because it can swing due to import and export and inventory shenanigans that have little to do with the overall health of the domestic economy.
Long rates always go up in war…
Not sure