At 8:30 p.m. on Friday, three things happened at the same time.
U.S. September nonfarm payrolls increased by only 29,000.
The market’s original expectation was an increase of 90,000 people.
That same day, the Fed’s No. 2 and No. 3 officials both weighed in, saying there’s no need to rush.
By all reasonable accounts, this should be bond investors’ night of celebration.
With jobs falling apart and rate hikes gone, yields should be heading lower.
As a result, the 10-year Treasury yield closed at 5.28%.
It even briefly touched 5.342%, the highest level since the beginning of 2002.
The thirty-year also didn’t lag behind.
On September 29, during the day it briefly touched 5.62%, the highest level since June 2002.
Within a day, it fell first then rose, and closed near the highs.
That’s the real thing to look at today: the Fed can determine the front end, but it can’t determine the back end.
I. This 29,000 is, statistically, indistinguishable from zero
First, let’s make clear how weak this number is.
The unemployment rate rose from 4.1% to 4.2%.
Average hourly earnings rose only 0.1%, up 3.0% year over year.
July was revised down by 31,000: from an increase of 21,000 to a decrease of 10,000.
August was revised down by 29,000—from 162,000 to 133,000.
So over two months combined, it’s 60,000 fewer jobs than previously reported.
The average net increase over the past twelve months is 45,000.
Putting these three things together for September, it’s clearly a cooling down.
But there’s another detail that gets mentioned less.
Based on the sampling methodology from the BLS itself, the monthly payrolls would need to move by about 122,000 people or more for it to be statistically significant.
29,000 is very far from this line.
In other words, this month’s employment cannot be statistically separated from “zero growth.”
Even more telling are two surveys fighting each other.
A company survey said new hiring was 29,000.
The household survey said employment increased by 406,000.
One says the fire is out; the other says it’s fine.
That means there’s only one practical interpretation: single-month data is already insufficient to support any conclusion—you can only look at multi-month trend.
II. The Fed says, “No need to rush,” because it can’t control it
On September 29, New York Fed President Williams said one very key thing.
He said monetary policy can’t move the ship, nor can it open the oil pipeline or the refinery.
This line explains the entire policy logic since September.
The source of this round of inflation is diesel, shipping routes, and refineries—not demand.
Because the Fed has no tools to repair an oil refinery.
So the only thing it can do is prevent this price surge from becoming a long-term expectation.
So when Williams said, “No need to rush,” he also added that another upward adjustment may still be needed later this year.
So he wasn’t turning dovish—he was admitting incompetence.
And on the inflation side, it also hasn’t loosened.
In August, CPI is still up 3.4% year over year.
The vice chair, Jefferson, followed up the next day with an even more direct remark.
He said that inflation above the Fed’s 2% target has been above it for more than five years.
He also said that after the September meeting, the entire yield curve moved higher again—investors are reassessing.
Pay attention to who said this.
The Fed’s No. 2 official, in his own words, admitted the market is repricing the long end.
And the Fed had just raised the federal funds rate in September to 3.75%–4%.
III. That night, the market first bought the notes—two hours later it changed its mind
The market has already played out this disagreement on the screen.
In the first hour after the payrolls report, bonds rose, gold rose, stock index futures rose, and Dow futures briefly gained more than 500 points.
What the market reads is: rate hikes are off the table—buy.
Therefore, in the second hour, the direction flipped.
Gold plunged from its high, and yields turned upward.
By the close, all three maturities were higher.
The two-year yield rose by 5 basis points to 4.83%.
The 10-year yield rose by 4 basis points to 5.28%.
The 30-year yield rose by 2 basis points to 5.63%.
Extend the window to a week, and the divergence becomes even clearer.
From September 25 to October 2, the two-year yield moved only up by 2 basis points.
In the same period, the 10-year yield rose by 11 basis points, and the 30-year rose by 14 basis points.
The front end listens to the Fed; the back end doesn’t.
Payrolls and the dovish tilt can hold down the two-year, but they can’t hold down the ten-year.
Fourth: the long-end notes are not in the Fed’s hands
Since it wasn’t the Fed, then who was it?
First, look at one small but tough move.
On September 24, the U.S. Treasury conducted a repo of 20 to 30-year Treasuries.
It said it would buy up to $6 billion.
The volume reported to the market is $10.468 billion.
In the end, the Treasury only completed $4.078 billion in deals.
The quoted volume is more than two and a half times the traded volume.
This is the second consecutive round of not buying it fully.
The fact that the buyback wasn’t fully completed—plain English: the price wasn’t agreed upon.
The quantity sellers want and the quantity buyers want differ by three times; the middle price line didn’t line up.
So what the long end lacks isn’t demand; it’s the price level that both sides agree on.
Now look at the other side.
This year, companies have already authorized a record $1.3 trillion in buybacks; the execution window opens in stages starting October 15.
In the same period, the positioning of trend-tracking funds swung from extreme over-allocation at the end of August to slightly net short.
The Z value measuring deviation fell from positive 2.35 to negative 0.80.
A one-month move exceeding three standard deviations—almost unprecedented in recent years.
On one side, the long end can’t find a bid price to take; on the other, stocks are waiting with buyback ammunition of more than $1 trillion.
So what should you be watching this round?
First: why isn’t gold rising?
Spot gold closed on Friday at $4,143.10, down 0.82%.
That day, its range was $101: it surged after payrolls, then plunged.
On Binance perpetuals, gold is now at $4,145.11.
For someone holding longs, this move is the most painful.
Payrolls are weak, rate-hike expectations are down, geopolitics is still there—these three things are all good for gold.
And it fell.
The reason isn’t complicated: gold trades based on the dollar’s real return, not the goodness or badness of employment.
As long as the 10-year yield stays above 5%, the counterparty for gold is always there.
Second: why didn’t crypto react?
This week, crypto was the quietest corner in the entire market.
Bitcoin is now trading at $84,856, up 0.25% over the past 24 hours.
Ethereum at $2,682.69, up 0.07%.
SOL at $119.67, up 0.03%.
The probability of a rate hike dropped from 64.2% to 22.7% within a week; Bitcoin moved only a few basis points.
So this isn’t strength—it’s not participating.
This shows crypto is now trading independently—macro data can’t get through.
Third: December is the real meeting.
For the October 27–28 meeting, the market assigns only a 22.7% probability to a rate hike.
But the probability of a December hike is still around 61%.
Goldman’s line is: the second rate hike would be pushed to December, and it’s even possible it won’t be needed after all.
And the September hike was passed by unanimous voting.
The disagreement is still over on the other end.
Dallas Fed President Logan believes the target range still needs to be raised by at least 50 basis points.[6]
So the October meeting isn’t the endpoint—it just pushes the problem to December.
Fourth: the oil price line has already been taken over by policy.
Brent perpetuals are now at $102.19.
Last Friday, the G7 decided to release 100 million barrels of reserves via the IEA, to be completed in four months.
Diesel is concentrated in the first twenty days.
At the same time, Trump also said he would not implement a diesel export ban.
The pricing power along this line has shifted from geopolitics to inventory figures.
VI. Your discipline
Don’t go long bonds just because payrolls came in weak.
Payrolls weigh on the front end; the front end has moved only 2 basis points this year.
What’s really rising is the 10-year and the 30-year—and the basis for pricing on both legs is inflation and supply.
Don’t go long gold just because payrolls were weak.
This week, gold told you with a $101 swing range that it’s tracking real yields—not employment.
Don’t go long crypto just because payrolls were weak.
It didn’t even care about a halving in hike probability—meaning it isn’t listening to macro right now.
The two things truly worth waiting for.
First is the October inflation data.
Only if inflation softens in sync does the issue of rate hikes truly pass.
Second is the December meeting itself.
Before that, every Treasury auction and repo outcome was more worth watching than the Fed’s speeches.
—MK keeps the promise
