S&P 500 closes at 7,718.60 on September 4, 2026 — it fell because the jobs number was good

All three US indices fell on September 4.

• S&P 500 7,718.60 (−0.38%)
• Nasdaq Composite 26,506.99 (−0.29%)
• Dow Jones Industrial Average 53,414.25 (−271.86 points, −0.51%)

There was one cause: the August employment report released that morning. And its contents were not bad. They were considerably better than expected.

1. The condition under which good data hurts

The familiar chain runs: strong data, stronger earnings, higher share prices.

It inverts under one condition — when the central bank is weighing tightening.

That is exactly where US policy sits. The Bank of Korea raised its policy rate to 3.00 percent on August 27; in the US, Fed Chair Kevin Warsh leaned toward an increase in his Jackson Hole speech on August 28, and markets began treating the September 15-16 meeting as live.

On September 3 those odds eased. Governor Christopher Waller said the decision depended on the August inflation print, markets read that as "not decided yet", and the S&P rose 1.06 percent to close at 7,747.71 (see "S&P 500 closes at 7,747.71 on September 3, 2026").

Then the September 4 report landed.

Triple the forecast, and the July decline that had rattled markets on August 7 no longer exists in the data (see "US August payrolls 162,000").

If the labour market can hold up under higher rates, there is no reason to wait. That is how the good number became the bad one.

2. A day when stocks and bonds fell together

The defining feature of the session is that equities and short-dated Treasuries sold off at the same time.

Normally, when stocks fall, money moves into government bonds, prices rise and yields fall. But when the cause of the equity decline is the rate outlook itself, both assets move the same way: if rates are going up, the bonds you already hold are worth less too.

The 2-year Treasury yield reached its highest level since January 2025. The 2-year is the maturity that tracks the expected policy path over the next year or two most directly, so a jump there means the market is putting an increase into actual prices rather than commentary (see "What the 2-year Treasury yield is").

Accounts of the 10-year diverged. Some described yields rising broadly after the release; another said the 10-year fell to 4.77 percent from 4.818 percent. Short yields rising faster than long ones is a normal tightening-phase shape, but we did not settle the day's figure.

3. Do not quote a single probability

The September hike odds reported that day:

A 12-point spread. The probability is not a vote count; it is backed out of fed funds futures prices, which move all session, so the timestamp changes the answer. Method matters too — which contract month is used, and whether the move is assumed to be a fixed 25 basis points (see "What a 56% chance of a rate hike actually measures").

Every tally agrees on one thing: the odds are higher after the report than before.

4. Lululemon, down 17 percent

The largest single move of the day happened outside the index story. Lululemon fell about 17 percent after second-quarter sales declined and the company cut both revenue and profit guidance.

On a day when the indices moved 0.3 to 0.5 percent, a 17 percent single-stock drop means a separate signal arrived from the consumer side while attention was on rates. Strong hiring alongside weakening premium apparel demand is worth reading next to the composition of the payrolls report itself: 59,000 of August's 162,000 jobs came from food services and drinking places.

5. What is still open

• The next fork is August CPI, the print Waller named, due before the September 15-16 meeting. Employment has supplied one argument for a hike; inflation supplies the other.
• August payrolls will be revised twice. July moved 44,000 in this release. Only the direction is settled today.
• The week was effectively flat. Against the August 28 close of 7,711.76, September 4's 7,718.60 is a gain of 0.09 percent — containing daily moves of −0.24, +1.06 and −0.38 percent. This is a market whose range is widening without its direction changing.
• Korea gets it next session. The KOSPI closed 1.64 percent higher at 6,687.21 on September 4, on a premise that was overturned hours later (see "KOSPI closes at 6,687.21 on September 4, 2026").