- NFP showed a gain of 272k jobs, while the UR survey showed a loss of 456k jobs
- Healthcare (86k) and govt (43k) made up 46% of the gains in NFP, while the UR survey showed a loss of 625k full time jobs
- Birth/death adjustment estimates accounted for 231k of the 272k gains (85%)
And that estimate in 2023 ultimately turned out to overstate job gains by nearly 800k. But let’s keep treating headline NFP like it’s high-quality data.
If you wonder whether the data can stay fishy irrational longer than I can stay committed, you underestimate my dedication to confirmation bias. I benefit greatly from three things.
- I have a Dunning-Kruger bicep tattoo
- I am so stubborn my wife says I have a grudge box where I keep things without an expiration date
- I have no pride…and I gotta tell you, it feels amazing
- 10T: 4.43%
- German Bund: 2.62%
- 2T: 4.89%
- SOFR 5.32%
- Term SOFR 5.33%
- JOLTs Job Openings 8.059MM vs. 8.34MM
- ADP Private Payrolls 152k vs. 175k expected
- ECB cuts from 4.00% to 3.75%
- Lagarde: Fight against high inflation isn’t over yet
- Non-Farm Payrolls 272k vs. 185k
- Unemployment rate 4.0% vs. 3.9%
- Average Hourly Earnings m/m 0.4% vs. 0.3%
- Average Hourly Earnings y/y 4.1% vs. 3.9%
- ISM Manufacturing PMI 48.7 vs. 49.6 expected
“Jobs are Jobs” - JMo
That’s the text response I got from JMo after I suggested the headline report wasn’t as strong as it seemed. Using that logic….
- Properties are properties
- Lumber is lumber
- Rates are rates
- Owners are owners…right JMo?
It’s such an insane statement I assume he’s laying the groundwork to plead insanity on our Core PCE wager.
In economics, there’s a concept of a typical person called, “homo economicus”. Assumptions about this persona include things like perfect information, utility maximization, rational, etc.
I am introducing a concept for my analysis using JMo as a proxy. This is the typical real estate professional. Bright, hard-working, Type A, and perhaps most importantly, busy. That workload (and definitely positively not cognitive laziness) prevents them from reading past the headline. While there is a near-perfect play on homo economicus, to avoid being canceled I will use JMo Proprius. Proprius is latin for “one’s own”.
JMo Proprius is on a call, telling the other caller how smart he is, how dumb JP is, and lamenting the general lack of deals. He’s an office owner but is making the call from his mountain home.
He sees a ticker alert that NFP added 272k.
Looks at T10, up 13bps
“Dang, this job market is nuts. The Fed is never gonna cut.” He hangs up, texts me and tells me how wrong I am.
Last week I highlighted the Quarterly Census Employment and Wages (QCEW). This is the far more accurate measure of jobs that comes with a huge lag. From September 2022 to September 2023, it showed nearly 800k fewer jobs than headline NFP showed.
We decided to create a graph tracking the delta between the two. We used a 12 month rolling average to smooth out the noise. The variance has swung wildly since covid, likely reflecting the huge drop in response rates to employment surveys (70% to 30%).
In other words, data based on surveys and estimates suggest the job gains are much higher than data based on…actual data.
My favorite economist, Dr. Anna Wong (Bloomberg’s Chief US Economist), believes the true NFP is sub-100k. Not 272k. Less than 100k. Too bad we won’t know for nearly two years.
“A true picture of nonfarm payrolls won’t emerge until long after the fact — the BLS will only release final benchmark revisions to the data for most of 2024 in February 2026. When the final revisions are released, we expect 2024 payrolls to be revised down by about 1 million.”
Powell has not once, not twice, but thrice mentioned the potential for “unexpected weakening” of the labor market. I think he might be seeing what I’m seeing, but he’s stuck between a JMo and a JP.
- if he totally discounts the headline numbers, he gets accused of ignoring the data and raises an alarm
- if he focuses exclusively on the 272k, he limits his options as other data confirms a slow down
Most of my beef is with the NFP headline number, which comes from the Establishment Survey. The Household Survey is the one the Unemployment Rate is derived from. The UR increased to 4.0% despite the participation rate falling. The UR climbed because this survey, unlike the NFP survey, showed a loss of more than 450k jobs.
Why does this matter so much to me? Because the Fed is keeping rates higher for longer based on faulty data. This increases the odds of a hard landing.
“But JP, average hourly earnings came in at 0.4%, which will keep inflation up” – JMo Proprius
Yes, it did. But without the seasonal adjustment (another estimate), average hourly earnings fell.
I don’t think the labor market is weak, just weaker than the headlines suggest. It’s still doing ok, but we are at risk of breaking it. Here’s what I concluded last week’s newsletter with. My NFP estimate was based on reading a lot about how government hiring had likely rebounded last month.
I’ve got a trifecta laid on Friday’s report:
- NFP 200k+
- Unemployment 4%
- Fed-speak overemphasizing the NFP
After Wednesday’s FOMC meeting, I find out if this wager hit.
FOMC Meeting
Five months into the year, the unemployment rate is where the Fed projected it to finish the year. Now what? Remember, I think inflation will also be at their year-end projection by the July 31st meeting.
I think Powell will come across as more dovish than most on Wednesday. I think he’s looking past the headline NFP prints and is on alert for sudden weakness. Inflation is on track, forward looking indicators are weak, and the jobs data is noisy. He’ll probably use the word “balanced” a lot. I think it’s a reasonable approach. I can scream into the abyss about labor market weakness, but he can’t.
The main event might be the updated projections, which come out every other meeting. The last one was in March.
Summary of Economic Projections (SEP, aka Blue Dots)
Unemployment Rate – the UR has to be raised since we’re already at 4.0%. I suspect it ticks up to 4.1%.
Core PCE – this one is interesting. Do they overreact to Q1 CPI data that had everyone freaking out? The last SEP showed a year end forecast of 2.6%, but we’re already at 2.75%. If they revise higher and we are there next month, what happens?
GDP – 2.1% last time and revisions have pushed GDP lower. If their growth forecasts are lowered, and their inflation forecasts are raised, they will get more questions about stagflation.
Fed Funds – the only one that matters. Three cuts is likely to be reduced, but to two or one? I think it will be two cuts because they don’t want to appear overly volatile with such mixed data.
Also, expect Powell to commit considerable energy to reminding us that the SEP is not a forecast, but rather a median of the anonymous 17 Fed members’ forecasts. He will remind everyone not to put too much emphasis on these projections.
Lastly, expect some questions about the divergence between the Fed and other central banks. I would expect Powell to suggest that a temporary divergence will look like a minor blip over a long enough timeline. Right now, he doesn’t care if a couple of the G7 central banks have started cutting rates.
CPI
The Fed should have the CPI data on Tuesday when their meeting begins, although it seems unlikely to have much impact on their forecasts. Consensus forecasts call for the headline number to come in at 3.4%, and the monthly to come in at 0.3%. Both would be the same as last month.
The Week Ahead
Duh.
The ideal outcome for Powell is a non-response from the market.
T10: 4.35%-4.51% current range
I can't imagine he says anything that would cause rates to break out of this range, but the low side would be 4.18% and the high side would be 4.64%.
