PMI: The One Number That Reads the Economy's Pulse

Every month, a quiet little data release hits the wire and barely causes a ripple in the mainstream news cycle. But on trading desks and in Fed conference rooms, it gets read like a doctor reading a heartbeat monitor. That report is the PMI — the Purchasing Managers' Index — and once you understand what it's actually measuring, you'll never look at economic headlines the same way again.

Here's why you should care: PMI data tends to move before the economy does. Jobs reports are backward-looking. GDP is ancient history by the time it's published. PMI is one of the few leading indicators we have — meaning it's trying to tell you where things are going, not just where they've been.

What PMI Actually Is (No Jargon, I Promise)

PMI stands for Purchasing Managers' Index, and the name tells you almost everything. The people being surveyed are purchasing managers — the folks at companies whose literal job is to buy the raw materials, components, and supplies that keep operations running. Think of them as the supply chain's early warning system.

Every month, these managers get a simple survey. Are things better, worse, or the same compared to last month? That's essentially it. The survey covers things like new orders, production levels, employment, supplier delivery times, and inventories. The responses get compiled into a single composite number — the PMI — that runs on a scale from 0 to 100.

The magic number is 50.

• Above 50 = the sector is expanding. More orders coming in, more things being made, more people getting hired.
• Below 50 = the sector is contracting. Orders drying up, output slowing, companies pulling back.
• Right at 50 = flat. Nothing's really getting better or worse.

That's the whole framework. A reading of 55 is meaningfully healthy. A reading of 45 is a warning sign. A reading that's been below 50 for six straight months is the kind of thing that makes Federal Reserve governors start updating their recession risk models.

There are two main PMI reports you'll see referenced: Manufacturing PMI and Services PMI. In the U.S., the most closely watched versions come from ISM — the Institute for Supply Management — though S&P Global also publishes its own PMI surveys. Both matter, but they're measuring different slices of the economy.

Manufacturing vs. Services — Why Both Matter

Manufacturing PMI gets the most historical attention. For most of the 20th century, manufacturing was the backbone of the U.S. economy, so a factory slowdown was basically synonymous with a recession. That relationship has loosened since the 1990s — services now make up about 70% of U.S. GDP — but manufacturing PMI is still a powerful signal because it's so tied to global trade, industrial investment, and supply chains.

Services PMI, tracked as the ISM Services Index, covers everything from healthcare and restaurants to tech and finance. Given how service-dominant the modern economy has become, some economists now argue this reading is actually more telling for the overall economic picture.

Here's the brutally honest version: when manufacturing PMI craters but services PMI holds up, you often get a strange split economy — industrial companies hurting while consumer-facing businesses stay afloat. That's actually a fairly common pattern heading into and out of recessions, which makes watching both numbers side by side so useful.

| PMI Reading | What It Signals | Historical Parallel |

|---|---|---|

| 60+ | Strong expansion — rare, usually near cycle peaks | Mid-2021 post-lockdown surge hit 64.7 |

| 55–59 | Healthy growth, economy firing well | 2017–2018 tax-cut-era manufacturing boom |

| 50–54 | Modest expansion, steady but not exciting | Typical mid-cycle range |

| 45–49 | Contraction — manageable, watch the trend | Late 2015–2016 manufacturing slump |

| Below 45 | Serious contraction, recession risk elevated | COVID collapse hit 41.5 in April 2020 |

| Below 40 | Deep contraction, likely a recession is underway | 2008–2009 financial crisis troughs |

Why PMI Moves Markets — Even If You Don't Own Stocks

This is where it gets personal. PMI isn't just trivia for market nerds. A surprise in the PMI report can genuinely affect your mortgage rate, the interest you earn on savings, and whether your employer decides to hire or freeze headcount.

Here's the chain reaction: a weak PMI reading suggests the economy is slowing. If the economy is slowing, inflation pressures typically ease. If inflation eases, the Federal Reserve has more room to cut interest rates. When rate cuts become more likely, bond prices rise, bond yields fall — and mortgage rates, which track closely with the 10-year Treasury yield, start drifting lower.

Flip that around: a surprisingly hot PMI — especially a Services PMI that shows businesses raising prices and scrambling to hire — can signal that inflation is stickier than expected. The Fed notices. Rate cut bets get pushed out. And the mortgages that were supposed to get cheaper... don't. That dynamic has played out repeatedly in recent years, and it's exactly the kind of environment described in Fed's Schmid Said the Quiet Part Loud — and Mortgages Are Paying for It, where sticky inflation data kept the pressure on borrowing costs far longer than most people expected.

The stock market reacts too — and sometimes in counterintuitive ways. A strong PMI in a rate-sensitive environment can actually spook equity investors, because it implies the Fed stays tight. A weak PMI can briefly rally stocks if it convinces traders that rate cuts are coming. It's maddening, but that's the game.

A Brief History of PMI Getting It Right (and Sometimes Wrong)

PMI has a solid track record as a leading indicator, but it's not a crystal ball. Let's look at a few memorable moments.

The 2008 Warning Shot

ISM Manufacturing PMI started falling below 50 in January 2008 — a full eight months before Lehman Brothers collapsed in September. By October 2008, it had cratered to 38.3, one of the lowest readings on record at the time. Plenty of economists were still debating whether a "soft landing" was achievable while the PMI was already screaming recession.

The 2015–2016 Manufacturing Scare

Manufacturing PMI spent most of 2015 and all of 2016 below 50. This triggered legitimate recession fears, but the services sector kept chugging along above 55. No recession came. The lesson: don't read manufacturing PMI in isolation, especially in a services-heavy economy.

The 2020 COVID Collapse and Rocket Recovery

April 2020 saw manufacturing PMI collapse to 41.5 — a historic drop happening almost overnight. But unlike 2008, the recovery was just as dramatic. By August 2020, manufacturing PMI was back above 56. The speed of that reversal was partly driven by the fiscal stimulus and partly by a demand surge in goods — as everyone stuck at home bought furniture, electronics, and exercise equipment instead of services. By mid-2021, manufacturing PMI hit 64.7, an absolute rocket ship reading that signaled supply chains were overwhelmed trying to keep up.

The 2022–2024 Divergence

As the Fed aggressively hiked rates to fight inflation, manufacturing PMI spent most of 2022 through early 2024 in contraction territory. Industrial companies were pulling back hard. But services PMI stayed remarkably resilient — often printing above 53 or 54 — which gave the Fed cover to keep hiking without triggering an outright recession. That split-screen economy, where factories suffered while restaurants and tech firms kept hiring, explains a lot of the confusion people felt during that period. It felt like a recession in manufacturing; it mostly didn't feel like one if you worked in services.

That industrial softness had real knock-on effects for major companies. When you see a company like Caterpillar — one of the best real-world proxies for industrial and construction demand — threading the needle between weakness in traditional infrastructure and new demand from AI data centers, you're watching the PMI story play out in earnings form. More on that in Caterpillar Just Told Us Something Important About the AI Boom.

The Sub-Indexes Hidden Inside the PMI

The headline PMI number gets all the attention, but the real analysis happens in the sub-components. These are the five main inputs that go into the ISM Manufacturing PMI, along with their typical market significance:

| Sub-Index | What It Measures | Why It Matters |

|---|---|---|

| New Orders | Demand coming in right now | The most forward-looking component — it tells you what production will look like next month |

| Production | Current output levels | Confirms whether demand is actually being met |

| Employment | Hiring and headcount changes | Feeds directly into labor market expectations |

| Supplier Deliveries | How long it takes to get supplies | Counterintuitively, slower deliveries = stronger demand (suppliers are overwhelmed) |

| Inventories | Stockpiles of raw materials | High inventory can signal over-ordering and a coming pullback |

New Orders is the one I watch most closely. It's essentially demand right now, which means production levels next month are going to follow it up or down. When New Orders diverges significantly from the headline PMI — say, the headline is 52 but New Orders dropped to 47 — that's a yellow flag that the overall number might be about to turn south.

Supplier Deliveries deserves a special mention because it reads backwards from everything else. A higher Supplier Deliveries number actually helps the PMI reading, because it means deliveries are slower — and slower deliveries typically mean suppliers are overwhelmed by demand. During the 2021 supply chain crunch, this component shot through the roof and contributed to some of those historic high PMI readings. Once supply chains normalized, Supplier Deliveries came back down and the PMI followed.

How PMI Connects to Tech, AI, and the Sectors You're Watching

You might think PMI is mainly a story about factories and steel mills, but the concept bleeds into tech in interesting ways — especially as AI infrastructure spending becomes a major economic theme.

When manufacturing PMI is weak but specific industrial sub-sectors are getting hammered with orders for data center components, cooling systems, and power infrastructure, you start seeing earnings diverge in ways that confuse people. A company can be operating in a broad manufacturing contraction while its specific niche is booming. That kind of divergence — between macro PMI signals and company-level demand — creates both the confusion and the opportunity you see in markets.

The memory chip sector is a great example of this tension. Broad tech demand can look soft by PMI measures while a specific application — say, AI training workloads — drives massive orders for high-bandwidth memory. The PMI captures the aggregate; it misses the rotation happening underneath. That's part of what made earnings from companies like SanDisk and Western Digital so confusing to outside observers — see AI Memory Stocks Just Did Something Really Strange for how that played out in real time. Similarly, AMD Grew Revenue 50% and Still Got Sold. Make That Make Sense. shows how even strong revenue growth doesn't read cleanly on a macro indicator like PMI — the aggregate masks the bifurcation happening within the sector.

How to Actually Use PMI Data in 2026

Okay, so what do you actually do with this information? Here's my practical take.

Watch the trend, not just the number. A reading of 49.5 isn't scary if it follows readings of 46, 47, and 48 — that's a recovery trend. A reading of 51.5 is worth worrying about if it follows 56, 54, and 53 — that's a deteriorating trend. Direction matters as much as the absolute number.

Compare manufacturing and services. A split economy — where one is expanding and one is contracting — is genuinely uncertain territory. Don't assume the expanding side will pull up the contracting one. Sometimes it does. Sometimes the manufacturing weakness spreads into services with a lag.

Pay attention to New Orders. If you only have time to read one sub-component, make it New Orders. It's the most forward-looking piece of the puzzle.

Don't overreact to a single print. One bad PMI reading proves nothing. Two or three consecutive readings below 50 is when you start taking the signal seriously. The Fed certainly does.

Connect it to rate expectations. PMI doesn't exist in a vacuum. Weak PMI data tends to push rate cut expectations forward in time — which affects mortgage rates, savings yields, and bond prices. Strong PMI keeps rate cuts on the back burner. Getting your head around that relationship makes a lot of other financial news suddenly click into place.

FAQ

What does a PMI above 50 mean?

A PMI reading above 50 means the sector being measured — manufacturing or services — is expanding compared to the previous month. More specifically, it means that a majority of the purchasing managers surveyed reported conditions getting better: more new orders, more production, more hiring, or some combination of those things. It doesn't mean the economy is great — a reading of 51 is technically expansion but barely — it just means the direction of travel is positive. The further above 50 you go, the stronger the growth signal.

What does a PMI below 50 mean for the stock market?

A PMI below 50 signals contraction, but the stock market's reaction depends heavily on context. If the PMI is dropping fast and investors fear a hard recession, stocks tend to sell off. But if a weak PMI reading convinces the market that the Fed will cut rates sooner, stocks can actually rally on bad economic news — because cheaper money tends to boost stock valuations. This is the frustrating reality of trying to trade macro data. The number doesn't tell you the market reaction; you also need to know where rate expectations stand.

How often is the PMI report released?

Both the ISM Manufacturing PMI and ISM Services PMI are released monthly. Manufacturing PMI comes out on the first business day of each month, covering the prior month. Services PMI follows a few days later, typically on the third business day. S&P Global also releases its own "flash" PMI estimates toward the end of the current month — so you get a preview of where things are headed before the official ISM release hits.

What's the difference between ISM PMI and S&P Global PMI?

Both measure the same general thing — business activity in manufacturing and services — but their methodologies and sample sizes differ. ISM surveys roughly 300–400 purchasing managers and has been running since the 1940s, giving it a long historical track record. S&P Global (formerly IHS Markit) surveys a larger panel of about 800 companies and publishes preliminary "flash" estimates before the final number. The two readings often track closely, but divergences between them can generate interesting debates about which is capturing the real picture. Most institutional investors watch both, especially when they're sending different signals.

Is PMI a reliable recession predictor?

It's one of the better ones we have, but not foolproof. Manufacturing PMI has dropped below 50 before several recessions — including 2001, 2008, and 2020 — with enough lead time to be useful. But it also spent extended periods below 50 in 2015–2016 without a recession following. Services PMI tends to be stickier and harder to shake below 50, which is partly why economists pay so much attention when it does fall — it doesn't do so casually. Think of PMI as a smoke detector, not a fire alarm. It goes off when there's something worth investigating, but you still have to look around to figure out whether it's an actual fire or someone burning toast.