Why the Stock Market Went Up on a Bad Jobs Report
This week’s episode of The Tradeoff, 14 minutes: Listen on Spotify · Apple Podcasts
The headline is that the economy had a bad day. The tradeoff is that the stock market had a good one, for the exact same reason.
On Friday, October 2, the Bureau of Labor Statistics reported that the economy added 29,000 jobs in September. Economists expected 90,000. Unemployment rose to 4.2%. July got revised from a small gain to a loss of 10,000 jobs, and August came down by 29,000. Wages grew 3% over the year, the slowest pace in five years. By any reasonable reading, that’s a bad report.
The S&P 500 closed up 0.7%. The Nasdaq rose 1.1% and touched a record high.
If you watched that and felt like the market lives on a different planet, you’re half right. It lives on the same planet. It’s asking a different question.
What the market was reading
Two and a half weeks before the jobs report, on September 16, the Federal Reserve raised its target range a quarter point to 3.75–4%, its first increase in three years. Markets don’t love rate hikes, because a hike makes capital more expensive. A company that borrows to grow pays more to do it, and a dollar of profit five years from now is worth less today when money costs more.
There’s a flip side. The Fed doesn’t usually raise rates unless it thinks the economy can handle it, so a hike is also a statement of confidence. Once the Fed is hiking, every new piece of data gets read through one lens: will this make another hike more or less likely?
A week before the report, futures markets put the odds of another hike at the October 28 meeting at 64.2%. After the report, those odds were 23.8%.
That’s the whole rally. Fewer jobs meant less pressure on the Fed. Less pressure on the Fed meant cheaper money ahead. Cheaper money is what stocks trade on.
So the stock market isn’t asking whether news is good. It’s asking how news is relevant to the price of money and the path of future profits. That answer is less connected to whether your neighbor has a job than you’d think.
This isn’t new. Ask anyone who started working in 2009.
On March 9, 2009, the S&P 500 closed at 676.53. The Fed had already cut rates to zero the previous December, and it would leave them there for seven years. By the end of 2009, the S&P was up almost 65% from that low.
If you remember 2009, it didn’t feel like a boom. Unemployment kept rising after stocks turned and didn’t peak until October, at 10%, seven months after the market bottomed. Hiring was slow. Wages weren’t moving.
Zero rates did two things at once. They made safe assets pay almost nothing, and they made future profits worth more today. Valuations had been crushed by the recession, and when you start at the bottom, a recovery looks pretty good on paper. Profits recovered on one timeline. Paychecks recovered on another.
I think we’ve badly underestimated what it does to a generation to watch that gap widen in their first years at work. You’re told the market is up and things are getting better, and your own pay says otherwise for years. People don’t come out of that confused about economics. They come out convinced the rules don’t work the way they were promised, and that’s a lesson we taught them.
It’s part of why this show spends so much time on the connection between what you feel about the economy and what happens in Washington. We’ve lived through twenty-odd years of once-in-a-lifetime events, and the systems that are supposed to translate those events into policy haven’t kept up.
The rule runs in reverse, and then it breaks
If a weak economy can lift stocks, a strong one can sink them. In 2022, unemployment hit 3.5%, a five-decade low. The S&P 500 fell 19.4% for the year. A strong job market meant a Fed that would keep raising rates, and that raised the price of capital.
Then there’s where the rule stops working.
On August 2, 2024, the July jobs report showed 114,000 jobs and unemployment at 4.3%. That tripped the Sahm rule, a recession indicator with a long track record. The Fed had held rates steady two days earlier. This time the market didn’t rally. The following Monday, the S&P 500 fell 3%.
The difference was what bad news promised. In 2026, a weak report means the Fed probably won’t hike again. In August 2024, with the Fed already sitting still, a weak report raised the question of whether it had waited too long, and whether the economy was cracking instead of cooling.
The stock market will cheer a cooling labor market right up until that labor market looks broken, and the line between the two is usually only obvious after the fact. That’s the line we’re walking now. One month of 29,000 jobs is a cooldown. A few more months of it starts to look like something else.
The new part of the story: the bond market sends the bill
For most of the last fifteen years, the conversation was the stock market and the Fed. I think the bond market is now the one sending the bill.
On October 7, the 10-year Treasury yield hit about 5.35%, the highest since 2002. The 30-year reached about 5.72%.
A long-term yield has two parts. One is what investors expect short-term rates to average over the next ten years, which is mostly a bet on the Fed. The other is the term premium, the extra return investors demand for locking their money up that long.
Since mid-September, the New York Fed’s estimate of the term premium has risen about 40 basis points, to roughly 1%, its highest level since 2014. Over the same stretch, the 10-year yield rose about 30 basis points. The premium explains more than the whole move. When the price of risk rises faster than the price of the debt, the market is telling you about supply, not about the Fed.
The supply is easy to find. Federal debt held by the public is about $32.3 trillion, roughly 100% of GDP. The only years it was higher were 1945 and 1946. Interest on that debt ran about $1 trillion this fiscal year, a record share of the economy.
Washington isn’t the only one borrowing. Amazon, Alphabet, Meta, Microsoft and Oracle issued $121 billion in high-grade bonds in 2025 to fund the AI buildout. By the end of June this year, the biggest AI builders had issued roughly $170 billion.
To put $170 billion somewhere you can see it: the Department of Transportation’s outlays in fiscal 2025 were $127.6 billion. That’s every federal dollar for highways, rail, transit and air traffic control in a year. A handful of tech companies borrowed more than that in six months.
To be fair to the other side, research so far hasn’t found that AI bond deals move Treasury yields directly, and that’s not my argument. The question buyers of that debt will ask is when the AI investment starts producing returns. The longer the wait, the riskier the debt looks, and an asset class the size of a federal department getting repriced is the kind of thing that creates tension in the bond market.
A forecast vs. an invoice
For a long time, bond yields were read as a forecast. The classic signal was the inverted yield curve, when a short-term government bond pays more than a long-term one. That’s backward from normal, because investors usually demand more to hold debt for 30 years than for two (think of everything that could go sideways in 30 years). When the near term looks riskier than the long horizon, it’s usually a loud recession warning.
From the back half of 2022 through 2024, the curve stayed inverted, the longest stretch on record. No recession came.
For fifteen years we read the bond market like a forecast. Now it reads more like an invoice.
It doesn’t matter much whether the 10-year is right about growth. It’s the price you start from for every other kind of borrowing: your mortgage, your car payment, your credit card. Each is the 10-year plus a premium for how risky the lender thinks you are. If that number is the price investors charge to absorb the debt that Washington and a handful of companies are carrying, it won’t fall just because the economy cools.
What that means for you
If you’ve been waiting for rates to fall before buying a house, that’s a riskier bet than it was. For fifteen years it was simply assumed that rates followed the business cycle down. An economy that cools while rates stay put breaks that assumption.
It reaches the federal budget too. Budgets are built over a 10-year window, on assumptions that don’t price debt this high. When interest costs outrun the baseline, the squeeze lands on things that potentially are essential.
I was saying this back around 2010. Borrowing wasn’t going to be free forever, and the largest borrower in the world needed a plan for when it got expensive. It’s expensive now. There isn’t a plan.
There’s one honest bright spot. If you have savings, your savings account and CDs, which paid close to nothing for fifteen years, finally pay something. That helps right up until you’re drawing those savings down to cover the cost of living, which is where a lot of households were in last week’s episode.
The better question
The next time someone tells you the economy must be fine because the stock market is fine, there’s a better question to ask: relevant to whom?
The market is a real signal. It’s answering a different question than most of us are asking. Most of us want to know whether there’s work, whether our income keeps up with prices, and whether a house is still within reach for us or our kids. The market wants to know what money will cost, and over what horizon.
You don’t need a trading desk to tell the difference. You need to know which question is being answered.
Listen to the full episode of The Tradeoff, 14 minutes: Listen on Spotify · Apple Podcasts
One thing to do today. Pick one headline this week, a jobs number, a market record, a rate decision. Find the one number underneath it that moved, and reply with it. I read every one, and the best ones will turn up in future episodes (no names unless you say so).
The rules you live under are readable. That’s what this show is for.