Guide · Markets
PMI Explained: How to Read a Purchasing Managers' Index, Flash vs Final, and What October 2026's 56.0 Said About Prices
A purchasing managers' index is a monthly survey of companies that reads above 50 when activity is growing and below 50 when it is shrinking. Here is how it is built, how the flash release differs from the final one and from the ISM, and why the price components of the October 2026 US flash PMI, with manufacturing at a 53-month high of 56.0, mattered more than the headline.
Most economic statistics count things after the fact: goods shipped, jobs filled, prices paid. A purchasing managers' index asks people instead. Each month a survey firm sends the same short questionnaire to several hundred companies and asks the managers who buy their inputs whether output, new orders, employment, delivery times, inventories and prices are higher, the same or lower than a month earlier. Each answer becomes a diffusion index: the share reporting an increase plus half the share reporting no change. A reading of 50 means as many companies saw improvement as deterioration; above 50, activity is expanding, below 50 it is contracting, and the distance from 50 says how broad the change is, not how big. A PMI of 56 does not mean output rose 6%; it means clearly more companies reported growth than decline.
Two families of PMI matter for US markets. S&P Global publishes manufacturing, services and composite PMIs for the United States and some forty other economies, including the HCOB-branded eurozone surveys, and releases a flash estimate around the third week of the month, based on about 85% of the responses, followed by a final reading about ten days later that rarely moves by more than a few tenths. The Institute for Supply Management publishes its own manufacturing and services indexes on the first and third business days of the following month, with a longer history going back to 1948 and a slightly different method. Because the flash PMI arrives first, it is the earliest broad read on a month's activity, a week before the ISM and three weeks or more before most official data.
Here is a real example. On October 23, 2026, S&P Global's flash US manufacturing PMI rose to 56.0 from 55.9, against 55.5 expected, its highest reading in 53 months. The services index fell to 57.9 from 58.8, just under the 58.0 forecast, and the composite slipped to 57.6 from 58.4. The headlines were mixed, but the components were not: new orders grew at the fastest pace in more than four and a half years, hiring at the fastest since June 2022, and the prices companies charge for their goods and services at the sharpest rate since August 2022. Four days before a Federal Reserve meeting at which policymakers were debating whether to raise rates again in December, the price questions were what investors read first. The 10-year Treasury yield rose about 3.5 basis points that day, and futures continued to price a December increase as more likely than not.
The same morning showed why the surveys of different regions are read side by side. The eurozone composite PMI fell to 51.9 from 53.1, well below the 52.7 expected, with French services back below 50, while the UK composite rose to 53.1 against a 51.8 forecast. On days like that, the euro and sterling tend to move against each other before any official data confirm the difference, which is exactly what a flash survey is for.
A few habits make PMIs more useful. Read the components, not just the headline: prices paid and prices charged are early inflation signals, new orders lead output, and the employment index often turns before payrolls. Compare the reading with the consensus forecast, because markets trade the surprise, not the level; 56.0 against 55.5 is a beat, while 56.0 against 57.0 would have been a miss. Look at the direction over three months rather than one, since monthly survey noise is a point or so. And remember that a PMI measures breadth, not size: a reading of 52 in a large, slow-growing economy and in a small, fast one say different things, which is why investors check them against GDP when it arrives, as US investors would on October 29, 2026, with the first estimate of third-quarter growth.