What Is the ISM Manufacturing PMI, and What Does It Actually Measure?
The ISM Manufacturing PMI is a monthly survey of purchasing managers at US factories, boiled down to one number: above 50 means more of them report manufacturing activity expanding than contracting that month; below 50 means the reverse. It's a sentiment read on the factory floor, not a measure of total output, and markets watch it because its subcomponents — new orders especially — tend to turn before broader growth data does.
That's the answer to the search query. The rest of this page covers how the number is actually built, why 50 is the line everyone watches, why a survey of purchasing managers moves markets at all, and the honest limits of reading one monthly print.
What is the ISM Manufacturing PMI built from?
The Institute for Supply Management surveys purchasing and supply executives at manufacturing firms across the country every month, asking each one whether conditions in several categories got better, stayed the same, or got worse than the prior month. Those answers are converted into a diffusion index — the share reporting improvement, plus half the share reporting no change, giving a single number centered on 50.
The headline PMI blends five subcomponents, not weighted equally:
| Subcomponent | Rough weight in the headline number | What it's actually asking |
|---|---|---|
| New orders | Largest single weight | Is fresh demand coming in, before it shows up as shipped output |
| Production | Second-largest | Is the factory floor currently running more or less |
| Employment | Meaningful | Are factories hiring or cutting shifts and headcount |
| Supplier deliveries | Smaller | Are deliveries from suppliers coming in slower (a strain signal) or faster |
| Inventories | Smallest | Are stockpiles building (demand running behind restocking) or depleting |
New orders gets read closest, by a wide margin, because of what it actually measures: not what factories produced last month, but what they've just been asked to produce. That's the subcomponent most likely to turn first when demand genuinely shifts, which is also why "new orders diverging from the headline number" is one of the first things an experienced reader checks before trusting the overall print.
Why does 50 matter?
Fifty is the breakeven line in a diffusion index by construction — it's the point where exactly as many respondents report improvement as report deterioration. Above 50, the manufacturing sector is reporting net expansion; below 50, net contraction. It isn't a growth-rate number like GDP, and a reading of 52 doesn't mean "the economy grew 2%" — it means a clear majority of surveyed purchasing managers are seeing better conditions than a month ago.
| PMI range | What it's generally read as |
|---|---|
| Above 55 | Strong, broad-based expansion |
| 50 to 55 | Modest expansion, usually unremarkable on its own |
| 45 to 50 | Contraction, not yet severe |
| Below 45 | Sharp contraction — historically rare and has clustered around recessions |
Those bands are rough historical pattern-matching, not fixed thresholds with forecasting power on their own — a single month at 48 after a long run above 55 reads very differently than a fourth straight month below 45.
Why does a survey of purchasing managers move markets at all?
Three separate reasons, not one:
- It's genuinely early. A purchasing manager places an order for raw materials and components before anything gets produced, so a shift in ordering intentions tends to show up in this survey before it shows up in official production, shipment, or employment data a month or more later. Among the regularly scheduled economic releases, the new-orders subcomponent is one of the more forward-looking reads on where manufacturing is headed.
- It arrives before the heavier, slower data it tends to foreshadow. Monthly GDP-adjacent manufacturing data and some labor-market detail land with more of a lag and get revised more; the PMI is released on the first business day of the month, covering the month just finished, long before most other data on the same stretch is available.
- Manufacturing has historically swung before the broader economy does. It's a smaller share of the economy than it once was, but the sector's turns have tended to lead the broader growth cycle — the same reason the business-cycle sector-rotation framework treats early-cycle manufacturing strength as one of its tells. A survey that catches that turn first earns outsized attention relative to its slice of GDP.
How does a PMI surprise actually move price?
The mechanism runs through growth expectations, the same channel Treasury yields and the dollar react to on other scheduled data. A PMI print that comes in meaningfully stronger than expected reads as "the growth outlook is better than priced," which can push yields higher on upgraded growth-and-rate expectations and lift cyclical, economically sensitive sectors harder than defensive ones. A weak surprise runs the other way — softer yields, pressure concentrated in cyclicals, relative support for defensives.
The size of the surprise relative to what was already expected matters more than the raw level. A PMI of 51 can be a disappointment if the market was positioned for 54, and a PMI of 47 can spark relief if the fear going in was a number starting with 4 and a 3.
How do traders actually use it?
As one leading-indicator input, read for the surprise and the subcomponent detail rather than the headline alone, in three practical ways:
- Checking new orders against the headline. A headline print held up by inventories or supplier deliveries while new orders slips is a different, weaker signal than the same headline number driven by genuinely strong fresh demand — the detail underneath the number is where the real read lives.
- Reading it alongside the services-sector counterpart. ISM publishes a separate Services PMI; manufacturing and services can diverge for stretches (manufacturing has been a smaller, more cyclical slice of a services-heavy economy for decades), so a manufacturing-only read can overstate or understate the broader picture on its own.
- Using it as one leg of a business-cycle read, alongside jobs data, inflation prints, and the slower, backward-looking confirmation GDP gives one to three months later, rather than as a standalone trigger — the same context-stacking approach that applies to every macro reading on this site.
What the ISM Manufacturing PMI doesn't do
- It doesn't measure output directly. It's a diffusion index of reported improvement versus deterioration, not a count of units produced or dollars shipped — a reading above 50 means breadth of improvement, not a specific growth rate.
- It doesn't cover the whole economy. Manufacturing is a meaningfully smaller share of US GDP than services; a strong or weak manufacturing print doesn't automatically tell you what's happening in the much larger services sector.
- It isn't immune to revision and noise. Survey-based data moves around month to month on sample composition and seasonal factors; one print, especially one near the 50 line, isn't a trend by itself.
- It doesn't time anything. Manufacturing turns have led broader cycles with inconsistent lead times historically — sometimes months, sometimes closer to a year — which makes it a directional tell, not a calendar.
Questions traders ask
What's the difference between ISM and S&P Global's PMI?
Both are manufacturing purchasing-manager surveys, built on a similar diffusion-index idea, but they're run by different organizations with different respondent panels, survey methodologies, and release timing. They usually move in the same direction but can disagree meaningfully in a given month — when the two diverge, that disagreement itself is often the more interesting read than either headline number alone.
Does a weak PMI mean a recession is starting?
Not on its own. A single sub-50 print is common and has shown up plenty of times without a recession following. A sustained run well below 45, especially alongside weakening in jobs data and other business-cycle indicators, is the pattern that has historically carried more weight — one month is a data point, not a verdict.
Why does the market sometimes rally on a weak PMI?
Because a weak enough print can shift rate expectations — if softer growth data makes future policy easing look more likely, that expectation can offset or outweigh the growth concern itself, especially in a market already positioned for softness. It's the same "bad news is good news for policy expectations" dynamic that shows up around other scheduled data, including FOMC decision days.
Where this fits in the bigger picture: the ISM Manufacturing PMI is one more macro context layer, the same way Treasury yields and the dollar index are context — each describes a different slice of the growth backdrop without telling you which stock to trade or when to trade it. How these context layers stack alongside a cycle framework that adds a time dimension is on the methodology page.
Keep reading
Get them free