At 9:45 a.m. Chicago time on Friday, MNI released the Chicago Business Barometer for August. The consensus called for 59.0. The print was 47.1 — down 10.5 points from July, the weakest reading of the year, the first drop below the 50 line since April and the lowest level since December 2025.
In an ordinary cycle a number like that ends a rate debate. This one started one. In the same hour the survey crossed the terminals, futures markets lifted the probability of a rate hike at the September meeting from 35% to 59%. By the close the Dow Jones Industrial Average was 464 points or 0.9% lower at 53,885.10 and the Nasdaq Composite 0.5% weaker at 26,402.42, ending a five-session winning streak — not because the economy is slowing, but because it is getting more expensive.
Two numbers sharing a digit, produced in the same trading session, pointing in opposite directions: 47.1 against a forecast of 59, and 59% odds of a hike. To understand why the market effectively discarded one and priced the other within minutes, you have to understand what a purchasing managers index actually measures. It is not what most investors assume.
What actually happened on Friday
Friday delivered three data points in quick succession. First the Chicago index, collapsing from 57.6 to 47.1. Then the final University of Michigan consumer sentiment reading at 51.7, revised up from a preliminary 51.0 but still about 6% below July’s 55.2 and roughly 11% below its year-ago level. And finally, from Jackson Hole, Federal Reserve Chair Kevin Warsh.
Warsh said inflation is too high, the labor market is effectively at full employment, and financial conditions may not be restraining the economy much at all. He put numbers on it: PCE inflation running at 3.7% over the past year and at a 4.1% annualized pace over the past six months. The better summer prints, in his phrasing, did not tell him that underlying trends had meaningfully improved.
The market took that seriously. Fed funds futures moved the probability of a hike on September 15 and 16 from roughly 35% to 59%, and on prediction markets the odds of any hike during 2026 climbed toward 69%. The weakest business survey of the year and the highest hike odds of the cycle were produced within two hours of each other. For the month as a whole, equities still finished green: the S&P 500 gained 2.5% in August, the Nasdaq 100 3.8%, the Dow 1.4%.
A diffusion index measures direction, not quantity
This is the heart of it, and it is almost never explained in the coverage. The Chicago Business Barometer is not a measure of output. It is a diffusion index. Respondents are not asked about units, revenue or capacity utilization. They are asked a direction question: better, unchanged or worse than last month. The share of answers is converted into a number between 0 and 100, where 50 means improvements and deteriorations exactly offset.
That has an uncomfortable consequence. A reading of 47.1 does not mean production fell 2.9%. It means slightly more purchasing managers reported a deterioration than an improvement. A company that halves its output counts exactly as much as one that trims it by a percent. And a region running sideways at a high level can print below 50 while nothing whatsoever is happening.
Then there is the sample. The survey draws on roughly 200 purchasing professionals in the Chicago area with a monthly response rate of about 50%, so a typical month rests on something like a hundred completed questionnaires. The headline is a weighted composite of five subindexes: new orders at 0.35, production at 0.25, order backlogs at 0.15, supplier deliveries at 0.15 and employment at 0.10, seasonally adjusted afterwards.
Do the arithmetic and a 10.5-point swing corresponds to a change of mind among a single-digit to low-double-digit number of respondents. It does not take a recession to produce that. It takes a few large plants on August shutdown, one delayed shipment, a single postponed capital order — plus a seasonal factor that is genuinely hard to estimate for a vacation month. The index is explicitly known for this: among the regularly watched sentiment gauges, it is the most volatile, and it has been for years.
Why Chicago is the outlier this time, not the early warning
A single survey cannot be refuted. A contradiction between several can be — and this contradiction is unusually broad.
The national ISM manufacturing PMI printed 55.6 in July, the highest since May 2022 and comfortably above the 54.0 consensus. Forecasts for the August report, out Tuesday at 10:00 a.m. Eastern, still cluster around 55. The distance between 47.1 and 55 is not a nuance; it is a different state of the economy.
Europe offers no confirmation either. The euro area manufacturing PMI rose to 52.8 in August from 51.9, beating the 51.8 expectation and marking the fastest factory expansion in four years. The composite output index edged to 52.1 from 52.0, its highest since November, with hiring returning for the first time this year. The strength was led, of all places, by Germany, where production rose the most since January 2022.
It is possible that greater Chicago — heavy on machinery, auto supply and food processing — turns before the rest of the country. That is precisely the historical case for watching this index. But an early indicator landing in the same week that European industry posts a four-year high and the national survey sits at 55 carries the burden of proof. The resolution arrives Tuesday. If ISM confirms a cooling, Chicago was the signal. If ISM prints 55 again, it was noise.
The one number in the report that is not noise
There is a part of the Chicago report that survives the volatility critique, and it was barely quoted on Friday: prices paid accelerated while the headline fell into contraction.
Why that deserves different treatment than new orders is methodological. Price diffusions are close to monotonic. Firms rarely report input prices falling; they report increases when suppliers push them through and report no change when nothing happens. A rising price index therefore reflects cost changes that have already occurred rather than sentiment about the future. Which is exactly why the combination of falling activity and rising prices is the only reading of the report that matters to a central bank — and it points toward tightening, not easing.
The Michigan survey lands on the same side. Year-ahead inflation expectations eased from 4.2% to 4.0%, the lowest since March but still historically elevated. Long-run expectations for five to ten years held at 3.3% for a third consecutive month. That is the series a central bank watches, because it measures the credibility of the target rather than the price of gasoline. And one detail from the same survey says more about household conditions than any sentiment index: only 8% of respondents expect their income to grow faster than prices in 2026, down from 18% in December 2024. Year-ahead business expectations fell 10%; the five-year outlook dropped 13%.
Why the Fed can hike anyway
On August 26, two days before Jackson Hole, the July PCE report landed: 0.2% month over month and 3.7% year over year on the headline, with core also up 0.2% on the month and unchanged at 3.3% on the year. That is not a stumble. That is a plateau well above the 2% target.
Then there is energy. Brent traded near $90.69 a barrel on Monday, roughly 3% higher, after the U.S. military struck Iranian rocket launchers apparently preparing to lay mines in the Strait of Hormuz — the first such exchange in more than a month. Gulf exports at an estimated 15 to 16 million barrels a day sit far above the March trough of 5 to 6 million but still well below the 22 to 24 million that flowed before the conflict. An oil supply shock hits headline inflation immediately and the core rate with a lag, through freight and intermediate input costs.
Finally, an institutional point that carries more weight than usual in this configuration. The Fed has pulled back its forward guidance and cut the number of regular meetings. Fewer dates mean more weight per date: skipping a hike in September does not defer it by six weeks, it defers it considerably longer. For a committee that sees long-run inflation expectations anchored at 3.3%, that is an argument for acting — even against a weak regional survey.
What it means for equities
For U.S. investors the distinction between Chicago and ISM is not academic. The industrial complex trades on national capital spending, not on Midwest sentiment. Caterpillar, Deere, Eaton, Parker Hannifin, Nucor and Fastenal earn against order books measured in quarters; a monthly survey of a hundred purchasing managers moves nothing there, whereas an actual decline in U.S. equipment investment moves everything.
The second layer is the discount rate. If the policy rate rises while the European Central Bank makes its own near-fully-priced decision on September 10, the repricing lands hardest on long-duration cash flows. Growth names whose earnings sit far out in the future take more damage than steady compounders — which is why an index at record valuations behaves very differently at the sector level than the headline suggests. Energy is the mechanical hedge in this configuration, and the same Hormuz risk that lifts oil producers is what makes the hike more likely in the first place.
Practically: an investor who rebalances on a single regional survey is trading noise. One who watches the combination of oil, core inflation and the rate path is trading the actual valuation question. The tax side is unchanged by any of it — gains held under a year are taxed as ordinary income, long-term gains at 0%, 15% or 20% plus the 3.8% net investment income tax where applicable — but it is a reminder that headline-driven repositioning carries its own cost, and that the wash sale rule punishes the round trip.
The counterarguments
The other side deserves a fair hearing. First, the Chicago index has genuinely been early on multiple occasions, precisely because the region is industrial and exposed. Dismissing it as noise on principle means eventually missing a turn. Second, the Michigan data are soft data too, and the finding that only 8% expect real income gains measures expectations, not incomes. Third, an oil spike is a supply shock. A central bank answering it with a rate hike is fighting a price increase it did not cause and can only damp through demand destruction.
And fourth, the asymmetry. If the Fed tightens into a genuine slowdown, the error is expensive and hard to undo, because a reduced meeting calendar takes away the fast reverse. That is exactly why the question of whether 47.1 is signal or statistics is more than a footnote in this cycle.
What settles it this week
The calendar resolves the argument unusually fast. On Tuesday, September 1, the national ISM manufacturing index arrives at 10:00 a.m. Eastern, and a few hours earlier the euro area flash inflation estimate — expected near 3.3%, the fastest since 2023, driven by energy. The ECB decides on September 10, the Fed on September 15 and 16.
Two subindexes of the ISM report matter more than the headline. New orders will show whether the Chicago weakness exists nationally. Prices paid will show whether the one unnoisy part of the Chicago report has a counterpart. If both move the way Chicago did, the market traded the wrong data point on Friday. If ISM holds near 55 with prices elevated, the Fed gets precisely the environment in which a hike can be justified: an economy that is running and prices that are not yielding.
Either way, Friday remains instructive. The market saw the weakest business reading of the year, classified it as what it methodologically is — a direction report from roughly a hundred purchasing managers in one region — and traded the number that came out of a speech instead. That is not irrationality. That is correct weighting of data quality, and it happens less often than one would hope.
Try TradingView Free for 30 Days
Plus get a $15 discount on your first subscription through this link.
