Guide

Purchasing Managers Index (PMI) explained

When factory floors and service firms feel demand shift, they adjust orders and hiring weeks before government statisticians publish GDP or payrolls. The Purchasing Managers Index (PMI) captures that early signal by surveying purchasing and supply executives each month: Are new orders rising or falling? Are suppliers delivering faster or slower? Are input prices climbing? The headline number is a diffusion index — 50 separates expansion from contraction — and it is one of the most-watched leading indicators on the economic calendar. Bond traders react to sub-50 manufacturing prints as recession warnings; equity analysts use Services PMI to gauge consumer-facing demand before earnings guidance shifts. This guide explains how PMI is constructed, the difference between ISM and S&P Global surveys, manufacturing versus services components, how PMI relates to recession dating and labor market data, a Harbor Logistics monthly demand read worked example, an indicator decision table, common pitfalls, and an investor checklist.

What PMI measures — and how the diffusion index works

PMI is not a price level and not a dollar measure of output. It is a sentiment and activity diffusion index built from monthly questionnaires sent to hundreds of purchasing managers at manufacturing plants or service companies. Each question offers three choices: better, same, or worse than the prior month (wording varies by survey). The formula assigns 100% of weight to “better,” 0% to “worse,” and 50% to “same,” then averages across questions.

The result is an index where 50 is the breakeven line: readings above 50 mean a majority of respondents report improving conditions; below 50 means deterioration. A PMI of 55 is not “55% growth” — it means the balance of firms is clearly expansionary. The distance from 50 indicates breadth and momentum, not magnitude of GDP change.

Why purchasing managers?

Purchasing managers sit at the choke point between sales forecasts and production schedules. They cut or raise material orders before headcount changes show up in payrolls, and they see supplier lead times before retail shelves empty. That makes PMI a leading indicator: New Orders sub-indexes often turn before industrial production and GDP growth rates confirm the direction.

ISM vs S&P Global — two U.S. PMI publishers

In the United States, two competing surveys dominate headlines:

ISM Manufacturing PMI — published by the Institute for Supply Management on the first business day of each month (10:00 a.m. ET). Covers ~300 manufacturing firms; the official U.S. manufacturing PMI in many calendar apps.
ISM Services PMI — same publisher, third business day of the month; covers non-manufacturing (services, construction, mining). Often larger share of U.S. GDP than factories.
S&P Global PMI (formerly Markit) — separate panel, slightly different methodology and sector weights. Often releases a “flash” estimate mid-month and a final reading at month-end.

ISM and S&P Global can diverge in the same month — different sample sizes, question wording, and seasonal adjustments. Markets usually react to ISM first for U.S. manufacturing because of its long history and fixed release slot. For global comparisons, S&P Global publishes harmonized PMIs across the eurozone, UK, China, and Japan on the same framework.

Key PMI components

Headline PMI is a weighted composite. Sub-indexes tell you why the headline moved — critical for sector trades and macro forecasting.

New Orders — the most forward-looking component; leads production and employment by one to three months. A headline PMI below 50 with New Orders above 50 suggests a shallow dip, not a deep contraction.
Production / Output — current activity level; coincident with industrial production data.
Employment — hiring intentions; leads official payrolls but is noisier in services.
Supplier Deliveries — inverted in ISM math: slower deliveries (longer lead times) boost the index, signaling demand pressure. Fast deliveries can mean weak demand, not efficiency.
Inventories — rising inventories with falling New Orders is a warning of future production cuts.
Prices Paid — input cost pressure; directionally related to PPI but measures breadth of firms seeing higher prices, not actual price levels.
Backlog of Orders — pipeline depth; shrinking backlogs with weak New Orders foreshadow layoffs.

Manufacturing PMI weights New Orders heavily; Services PMI emphasizes business activity and new export orders for trade-sensitive firms. Always read the component table in the release PDF — headline surprises driven solely by Supplier Deliveries are less durable than New Orders shocks.

Manufacturing vs Services PMI

U.S. manufacturing is roughly 10% of GDP; services exceed 70%. A sub-50 Manufacturing PMI with a strong Services PMI above 55 paints a different macro picture than both indexes contracting together.

Manufacturing PMI — cyclical and trade-sensitive

Tied to capital goods, autos, chemicals, and exports. Reacts quickly to interest rates, dollar strength, and global demand. Extended periods below 50 often coincide with industrial earnings recessions even when consumer spending holds up.

Services PMI — domestic demand and labor

Covers healthcare, finance, retail, logistics, and technology services. Employment sub-index here foreshadows service-sector hiring. Prices Paid in services feeds into core inflation narratives watched by the Fed — pair with monetary policy expectations when Prices Paid stays elevated while New Orders cool.

A manufacturing–services divergence (mfg weak, services strong) is common in mid-cycle slowdowns. Simultaneous sub-50 readings in both raise recession probability faster than either alone.

PMI vs GDP, payrolls, and other leading indicators

PMI and GDP measure different things at different speeds. PMI is monthly and survey-based; GDP is quarterly and accounts-based. Economists use PMI-to-GDP regression models — a rule of thumb (not a law) is that ISM Manufacturing near 50 aligns with flat industrial production, while sustained readings below 48 historically overlap with broader slowdowns.

PMI vs GDP — PMI leads by one to two quarters on average; GDP revisions can erase apparent contradictions months later.
PMI vs payrolls — Employment sub-index leads nonfarm payrolls directionally but not in magnitude; services hiring is stickier than PMI implies.
PMI vs consumer confidence — PMI reflects business purchasing decisions; consumer surveys reflect household sentiment. Businesses can cut orders while consumers still spend (and vice versa).
PMI vs yield curve — inverted curves predict recessions on a longer horizon; PMI tells you if the real economy is slowing now.

No single PMI print “calls” a recession. The NBER dates recessions using multiple coincident indicators. PMI contributes early warning, not official classification.

How investors use PMI release days

ISM Manufacturing hits at 10:00 a.m. ET on the first business day of the month — often the same week as jobs data. Markets trade the surprise versus consensus (published by Bloomberg/Reuters surveys) and the component mix.

Typical asset reactions

Equities — weak Manufacturing PMI hurts cyclicals (industrials, materials, transports); strong Services PMI supports consumer discretionary and domestic tech.
Bonds — sub-50 PMI with weak Prices Paid supports duration rallies (growth scare); hot Prices Paid with strong New Orders can sell off Treasuries (inflation + growth).
U.S. dollar — strong relative PMI vs eurozone or China supports USD; weak U.S. PMI with strong global peers can weaken DXY.
Commodities — manufacturing weakness weighs on copper and crude demand expectations; Prices Paid spikes can lift near-term inflation hedges.

Position sizing matters: PMI revisions are rare but methodology changes are not. Flash vs final S&P Global readings can whipsaw intraday traders. Long-term investors log trends — three-month averages of New Orders — rather than trading every 0.2-point miss.

Worked example: Harbor Logistics monthly demand read

Harbor Logistics runs regional freight and warehouse contracts. Before updating their quarterly revenue outlook, the strategy team runs a ten-minute PMI checklist on ISM release morning:

Check ISM Services headline and New Orders — Harbor’s clients are retailers and wholesalers. Services New Orders below 50 for two consecutive months triggers a volume-downside flag on their internal forecast model.
Read Manufacturing New Orders and Inventories — if manufacturing New Orders are weak but customer inventories are falling, destocking may be ending (bullish for freight). If both New Orders and Inventories are weak, expect fewer inbound shipments.
Note Prices Paid in both surveys — rising Prices Paid with soft New Orders signals margin pressure for clients, not volume growth. Harbor delays warehouse expansion capex.
Compare U.S. ISM to China Caixin Manufacturing PMI — Harbor moves trans-Pacific containers; China PMI below 50 with U.S. above 50 suggests import volume may soften in six to eight weeks.
Write one paragraph for the ops journal — e.g. “ISM Mfg 48.2 (New Orders 47.1), Services 53.4 (New Orders 54.0). Manufacturing contraction deepening but services demand intact. Hold headcount; cut discretionary capex 10% until New Orders mfg prints two months above 50.”

Harbor does not trade PMI releases — they use the survey to time hiring, capacity, and client conversations before hard revenue data arrives.

Indicator decision table

Question you have	Best indicator	Why
Is the economy slowing before GDP confirms?	ISM New Orders (mfg + services)	Earliest monthly business-demand signal in the survey
Are factories specifically in recession?	ISM Manufacturing PMI	Direct read on industrial cycle; can diverge from services
Will payrolls weaken next month?	PMI Employment sub-index	Leads hiring intentions; confirm with JOLTS and claims
Are input costs rising at producers?	PMI Prices Paid + PPI	PMI = breadth of firms; PPI = actual price levels
Official recession dating?	NBER coincident indicators (payrolls, IP, sales)	PMI is not used for official recession calls
Consumer spending outlook?	Retail sales + Services PMI	Household spending data beats business surveys for B2C

Common pitfalls

Treating 49.9 vs 50.1 as economically huge — both are essentially flat; trend and components matter more than the decimal.
Ignoring seasonal adjustment quirks — January and July PMIs often distort; compare year-over-year or three-month moving averages.
Confusing PMI with output growth rates — PMI can fall while GDP still grows (slowing expansion, not contraction).
Using manufacturing PMI alone for U.S. macro — services dominate; always check ISM Services the same week.
Equating Prices Paid with CPI — Prices Paid is a diffusion index of cost pressure, not consumer inflation; see PPI/CPI guides for price levels.
Overreacting to Supplier Deliveries — supply-chain normalization can boost this component without demand strength.
Mixing ISM and S&P Global without labeling — they are different surveys; comparing flash to final across publishers creates false signals.
Single-country PMI for multinationals — S&P Global composite PMIs or regional breakdowns matter for exporters.

Investor checklist

On ISM Manufacturing day (first business day, 10:00 a.m. ET), read headline, New Orders, Employment, and Prices Paid before trading the surprise.
Three days later, repeat for ISM Services — compare manufacturing–services divergence.
Track three-month moving average of New Orders for both surveys to filter noise.
Map PMI trends to sector exposures (industrials vs consumer services vs imports).
Cross-check weak PMI with GDP growth, unemployment, and yield curve — convergence strengthens the signal.
Log surprises in your macro journal; note whether bond or equity reaction matched the component story.
For global portfolios, add eurozone and China PMI prints the same week for relative growth trades.

Key takeaways

PMI is a monthly diffusion index from purchasing manager surveys — 50 separates expansion from contraction.
ISM Manufacturing (first business day) and ISM Services (third business day) are the primary U.S. releases; S&P Global offers an alternative methodology.
New Orders is the most forward-looking sub-index; Prices Paid signals input inflation breadth, not CPI directly.
Manufacturing and services PMIs can diverge — read both before drawing macro conclusions.
PMI leads GDP and payrolls directionally but does not officially date recessions; use it as an early warning alongside hard data.