PMI and ISM Manufacturing Index

PMI data—the Purchasing Managers Index—lands on the first business day of every month, before almost every other hard economic release. That timing makes it one of the most market-moving data points on the calendar. The ISM Manufacturing PMI, published since 1948, surveys purchasing managers at roughly 400 US companies about new orders, production, employment, deliveries, and inventories. Their collective responses produce a single number that tells you whether manufacturing is expanding or contracting—and more importantly, where it's headed next. The practical value isn't predicting GDP to the decimal. It's reading direction and momentum before the rest of the data confirms it.
TL;DR: The ISM Manufacturing PMI is a diffusion index where readings above 50 signal expansion and below 50 signal contraction. New orders is the most forward-looking component. Track it alongside the headline to anticipate turns 60–90 days early.
What PMI Data Actually Measures (And Why It Arrives First)
The Purchasing Managers Index captures business conditions through monthly surveys of purchasing managers—the executives responsible for buying materials, equipment, and services. Each respondent answers whether conditions are "better," "same," or "worse" compared to the prior month. The responses get converted into a diffusion index.
The point is: PMI data is released before most hard data (industrial production, factory orders, employment revisions), making it a leading indicator of economic direction. Industrial production data from the Federal Reserve arrives two to three weeks later. Factory orders from the Census Bureau arrive even later than that. PMI fills the gap.
Who publishes PMI data:
- ISM (Institute for Supply Management): The original US manufacturing and services PMIs, published since 1948. Surveys approximately 400 companies weighted toward larger firms.
- S&P Global (formerly IHS Markit): Competing PMIs with different methodology and a larger sample of approximately 800 companies including more mid-sized firms. History only extends to 2007.
The diffusion index methodology is straightforward. If 60% of respondents say conditions improved, 10% say conditions worsened, and 30% say no change, the diffusion index reads (60 + (30 × 0.5)) = 75. That formula—percentage reporting improvement plus half the percentage reporting no change—produces numbers that oscillate around 50.
The 50-Level Threshold (And the Nuance Most People Miss)
PMI readings revolve around a single dividing line:
| Reading | Interpretation |
|---|---|
| Above 50 | Expansion—more respondents reporting improvement than deterioration |
| Exactly 50 | No change from prior month |
| Below 50 | Contraction—more respondents reporting deterioration |
Critical nuance most commentary ignores: A reading of 48 does not mean manufacturing output fell 2%. It means more purchasing managers reported declining activity than improving activity. The rate of change and direction matter more than the absolute level. A move from 44 to 48 is improving (even though it's still below 50), and markets often react positively to improving momentum even in contraction territory.
Historical context for calibration:
- Long-run average: approximately 52–53
- Strong expansion: above 55
- Deep contraction: below 45 (rare outside recessions)
- The index bottomed at 34.5 during the 2008 financial crisis and hit 41.5 during the initial COVID shock in April 2020
Why this matters: Knowing the historical range prevents overreaction. A reading of 49 sounds alarming ("contraction!") but is well within normal oscillation. A reading of 44 is genuinely unusual and warrants attention.
ISM Manufacturing Index Components (The Five Moving Parts)
The headline ISM Manufacturing PMI is a composite of five equally weighted components, each measuring a different dimension of manufacturing activity:
| Component | What It Measures | Why It Matters |
|---|---|---|
| New Orders | Forward-looking demand signal | Best leading indicator of future headline direction |
| Production | Current output levels | Confirms or contradicts new orders |
| Employment | Hiring and layoff trends | Links to payroll data; lags orders |
| Supplier Deliveries | Supply chain conditions | Inverse scale (above 50 = slower deliveries) |
| Inventories | Stock levels relative to demand | Signals future production adjustments |
The calculation:
ISM Manufacturing PMI = (New Orders × 0.20) + (Production × 0.20) + (Employment × 0.20) + (Supplier Deliveries × 0.20) + (Inventories × 0.20)
Equal weighting is simple but important—it means a single component swinging hard (like supplier deliveries during COVID) can move the headline even when other components are stable.
Worked Example: Reading an Actual Release
Here's a real worked example using October 2024 data (released November 1, 2024):
Component readings:
- New Orders: 47.1
- Production: 46.2
- Employment: 44.4
- Supplier Deliveries: 52.0
- Inventories: 42.6
Headline PMI: (47.1 × 0.20) + (46.2 × 0.20) + (44.4 × 0.20) + (52.0 × 0.20) + (42.6 × 0.20) = 46.5
How to interpret this release:
The headline at 46.5 sits in contraction territory but well above deep-contraction levels (below 45). The more telling signal is the component breakdown. New orders at 47.1 with employment at 44.4 indicates firms are cutting headcount faster than orders are declining—a common pattern when companies expect weakness to persist.
Supplier deliveries at 52.0 (slightly slow) is unremarkable. Inventories at 42.6 (the weakest component) suggests firms are actively drawing down stock, which could set up a future production bounce once destocking ends.
Revision context: ISM revises prior months' data, though revisions are typically small—usually ±0.2 to 0.5 points on the headline. Large revisions (above 1 point) are rare and worth investigating when they occur. Always check whether the prior month was revised before comparing month-over-month changes.
The point is: The headline number gets the media attention, but the component breakdown tells you the actual story. A 46.5 driven by inventory destocking has very different forward implications than a 46.5 driven by collapsing new orders.
Why New Orders Matters Most (The Leading Edge)
Among the five components, new orders is the most forward-looking. A drop in new orders today signals lower production in coming months. This isn't theory—it's mechanical. Purchasing managers report on orders they've received, and those orders must eventually become (or fail to become) production.
The signal hierarchy in practice:
- New orders decline →
- Production slows (1–2 month lag) →
- Employment adjusts (2–4 month lag) →
- Inventories correct (variable lag, depends on how wrong firms' forecasts were)
This sequence plays out consistently. In the 2022–2024 manufacturing slowdown, new orders dropped below 50 in June 2022. Production followed below 50 by August 2022. Employment didn't cross below 50 until October 2022.
What matters here: Watch new orders to anticipate where the headline will be in 60–90 days. If new orders are rising while the headline is still in contraction, the headline is likely to improve. If new orders are falling while the headline is still above 50, deterioration is coming.
Prices Paid: The Inflation Signal (Not in the Headline)
The ISM also reports a prices paid subindex that is not included in the headline calculation. This measures input cost pressures faced by manufacturers—what they're paying for raw materials, energy, and components.
| Prices Paid Level | Interpretation |
|---|---|
| Above 60 | Significant price pressures (inflationary) |
| 50–60 | Moderate price increases |
| Below 50 | Prices falling (deflationary signal) |
Why this matters for your macro framework: Elevated prices paid combined with slowing new orders signals stagflationary pressure—rising costs with falling demand. That's the worst combination for policymakers (and for equity markets). During 2021–2022, prices paid exceeded 75 while new orders were volatile, foreshadowing the aggressive Fed tightening cycle.
Falling prices paid combined with stable or rising new orders signals the opposite: a goldilocks environment where input costs ease while demand holds. Markets tend to respond well to that combination.
Supplier Deliveries: The Supply Chain Signal (Inverse Scale)
Supplier deliveries works inversely to other components, which trips up many readers:
- Above 50: Slower deliveries (supply constraints, bottlenecks)
- Below 50: Faster deliveries (ample capacity, less strain)
A rising supplier deliveries reading mechanically pushes the headline PMI higher even though it reflects supply problems, not economic strength. This distortion matters during supply shocks.
The pandemic example made this obvious. Supplier deliveries spiked above 70 in 2021–2022 as supply chains seized up. This inflated the headline PMI, masking underlying demand weakness. By late 2024, it normalized to the 50–52 range, removing that distortion.
The point is: When supplier deliveries is unusually high or low, mentally adjust the headline. The "true" demand picture comes from new orders and production, not from supply-chain noise bleeding into the composite.
ISM vs. S&P Global PMI (When They Disagree)
Both surveys measure manufacturing conditions but sample different populations:
| Dimension | ISM | S&P Global |
|---|---|---|
| Sample size | ~400 companies | ~800 companies |
| Company size focus | Skews larger | Broader mix including mid-cap |
| History | Since 1948 | Since 2007 |
| Release timing | First business day | Usually 1–2 days earlier |
| Revision policy | Revises prior month | Also revises; final vs. flash readings |
When they diverge, the divergence itself is information. Large company conditions (ISM) may differ from small and mid-sized company conditions (S&P Global). During 2023, the S&P Global manufacturing PMI occasionally read 1–2 points higher than ISM, suggesting smaller manufacturers were faring better than industrial giants—possibly due to nearshoring orders flowing to agile mid-sized firms.
The practical rule: Use ISM as your primary reference (deeper history, more market impact). Check S&P Global for confirmation or divergence. If they disagree by more than 2 points for two or more consecutive months, investigate the size-segment or sector explanation.
PMI and GDP Correlation (Useful But Imprecise)
Research from ISM and various Federal Reserve banks suggests approximate PMI-to-GDP relationships:
| PMI Level | Approximate GDP Growth |
|---|---|
| 60 | ~5% annualized |
| 55 | ~3% annualized |
| 50 | ~1% annualized |
| 45 | ~-1% annualized |
| 40 | ~-3% annualized |
The caveat that changes everything: Manufacturing represents only about 11% of US GDP. These PMI-to-GDP correlations held more reliably when manufacturing was a larger share of the economy. The ISM Services PMI (covered in the companion article on ISM Services and Composite Measures) often tells you more about overall GDP trajectory.
Manufacturing can diverge from the broader economy for extended periods. In 2015–2016, manufacturing entered recession territory (below 50 for multiple months) while services expanded and the economy grew. In 2022–2024, manufacturing contracted for over a year while services remained resilient and GDP stayed positive. In a services-dominated economy (services account for roughly 70% of GDP), manufacturing weakness does not guarantee recession.
Risks, Limitations, and Common Pitfalls
Pitfalls that cost investors money:
- Treating PMI as GDP. Manufacturing PMI measures one sector. A below-50 reading does not mean the economy is shrinking. This error generates unnecessary panic trades.
- Ignoring duration. A single month below 50 is noise. Six consecutive months below 50 is a pattern worth acting on. The threshold matters less than the persistence.
- Conflating ISM and S&P Global. Different surveys, different respondents, different methodologies. Saying "PMI was 49" without specifying which PMI creates confusion.
- Missing seasonal patterns. Some months have consistent seasonal biases (January often shows rebound effects from holiday-period slowdowns). The data is seasonally adjusted, but adjustments aren't perfect.
- Overweighting a single component. Supplier deliveries distortions (especially during supply shocks) can move the headline without reflecting actual demand changes.
The lesson worth internalizing: PMI is a directional signal and a momentum indicator. It tells you whether conditions are getting better or worse and how quickly. It does not tell you the level of economic output with any precision.
Checklist for ISM Manufacturing Release Day
Essential (before the release, first business day of month)
- Know the consensus expectation for the headline (available from major financial data providers)
- Note the prior month's reading and whether it was revised
- Check for known distortions (strikes, hurricanes, seasonal quirks)
- Note the S&P Global flash reading (if already released 1–2 days earlier) for directional context
After the release (within 30 minutes)
- Compare headline to consensus—the surprise relative to expectations moves markets, not the absolute level
- Check new orders separately—this is your forward signal for the next 60–90 days
- Check employment separately—compare to jobless claims data (see Jobless Claims as a Weekly Signal) for labor market cross-reference
- Note prices paid for inflation signal—watch for divergence from new orders direction
- Compare to S&P Global reading for confirmation or divergence
- If headline was revised from prior month by more than ±0.5 points, investigate why
Ongoing tracking (monthly habit)
- Maintain a simple spreadsheet of headline PMI and new orders for the past 12 months
- Track the new orders minus inventories spread—a positive spread suggests future production acceleration; a negative spread suggests deceleration
- Note months where ISM and S&P Global diverged by more than 2 points
Next Step
Track the ISM Manufacturing PMI new orders subindex alongside the headline for six months. Record both numbers each month in a simple two-column log. Note how often changes in new orders precede changes in the headline by 1–2 months. This exercise builds firsthand intuition for using PMI as a leading indicator—the kind of pattern recognition that turns a data release from noise into a usable signal.
Related Articles

Consumer Confidence and Sentiment Surveys
How the Conference Board and University of Michigan surveys differ, what they predict about consumer spending, and their limitations as economic indicators.

Inventory-to-Sales Ratios
Understanding how inventory-to-sales ratios signal economic turning points, including sector-specific thresholds and interpretation guidelines.

Natural Gas Pricing Hubs and Seasonality
How natural gas pricing works across regional hubs, with seasonal patterns that create predictable price swings of $1-3/MMBtu between winter and summer.