The ISM Manufacturing PMI is normally released on the first business day of the month at 10:00 a.m. Eastern. The published headline is conventionally given to one decimal place, not two. It summarizes respondents' reported changes and is not a measured dollar change in production or a prediction of a fixed currency move.

What the ISM Manufacturing PMI actually is

The Institute for Supply Management surveys US manufacturing purchasing and supply executives about changes from the previous month. PMI means Purchasing Managers’ Index. The response categories depend on the component; Supplier Deliveries reports slower, unchanged or faster, so a higher delivery index does not simply mean improved business.

A raw diffusion index adds the percentage reporting an increase to half the unchanged share; Supplier Deliveries uses slower rather than increase. Hypothetical shares 24% higher, 60% unchanged and 16% lower give 24 + 0.5 × 60 = 54. This is not 54% of firms reporting higher or 4% output growth. Even raw 50 can describe different groups: 1%/98%/1% and 40%/20%/40% both produce 50 before seasonal adjustment.

Half the firms increase output. What happens to the total?

In two fictional cases, half produce more and half less. Response directions are identical; firm-level change sizes differ.

Scroll sideways to see the full diagram

Diffusion index versus output quantityEach case starts with 100 firms producing 100 units each, total 10,000. In A, 50 increase 1% and 50 decrease 50%, totaling 7,550. In B, a 100% increase and 1% decrease total 14,950. Raw DI is 50 in both.DI = 50, DIFFERENT OUTPUT TOTALSEach case: 50 firms up, 50 downA: +1% / −50%B: +100% / −1%5,050 + 2,50010,000 + 4,9507,55014,950−24.5% quantity+49.5% quantityReference line: 10,000 units before changeDI: 50% “higher” + ½ × 0% “same” = 50

Each fictional case has 100 valid, equally weighted responses. Every firm initially produces 100 units of the same good. We calculate raw Production DI without seasonal adjustment, not the full five-component Manufacturing PMI or US economic output. The documentation ISM: Manufacturing PMI methodology explains that raw DI combines positive responses and half unchanged responses, while headline PMI comprises five components.

Bars show physical-unit totals in the model. Identical raw diffusion indices do not determine this total.

DI retains response direction but not quantity-change size. Thus model DI 50 can accompany either a 24.5% aggregate decline or a 49.5% increase. In actual reports, distinguish a component, its seasonal adjustment and headline PMI.

How to read the 50 line

For a conventional change diffusion index, 50 means equal shares reporting increase and decrease; above and below 50 describe the balance of responses. The headline Manufacturing PMI is a composite whose supplier-deliveries component rises for slower deliveries. Distance from 50 is not a percentage change in measured output, and slower deliveries can also reflect disruption.

How to interpret an ISM Manufacturing PMI reading
Above 55 pointsA higher composite reading; not proof of overheating or a required rate rise
Between 50 and 55 pointsAbove the conventional sector threshold; inspect the five components
Exactly 50 pointsThe conventional composite threshold; not a count of all respondents
Between 45 and 50 pointsBelow the conventional sector threshold; not a measured output-growth rate
Below 45 pointsA lower composite reading; not an automatic recession designation

A move from 54 to 51 lowers the headline diffusion index but does not quantify the change in measured output. A reading of 50 versus a forecast of 48 is an upward surprise and sits at the conventional neutral line, not inside contraction. An exchange-rate response remains conditional on other information and expectations.

The sub-indices that matter most

The five headline components are New Orders, Production, Employment, Supplier Deliveries and Inventories, each weighted 20%. Prices is separate; Customers’ Inventories is also distinct from respondents’ own Inventories. In a hypothetical calculation, raw 54 divided by an invented seasonal factor 1.020 gives 52.941…, rounded to 52.9. The following inputs are ready for averaging after the appropriate adjustments.

Hypothetical Manufacturing PMI components
ComponentIndexContribution at 20%
New Orders52.910.58
Production51.110.22
Employment48.09.60
Supplier Deliveries61.012.20
Inventories47.09.40

The five levels sum to 260:260/5 = 52.0. Changing only Supplier Deliveries from 61 to 51 lowers the result to 50.0, a 2-point headline change, without changing orders, production or employment. Slower deliveries can have different causes; inspect commentary rather than declaring them good news. Seasonal factors and historical series can be updated; use the factors for the relevant vintage.

Another hypothetical pair swaps New Orders 60/Production 40 for New Orders 40/Production 60, holding the other three components at 50. Both headlines are 50, but orders minus production is +20 or −20 index points. That subtraction is our comparison, not an official ISM backlog measure or a dollar value of orders.

New Orders describes reported order changes; Employment describes respondents' employment changes; Prices records price-change breadth. None guarantees a later production, payroll or CPI outcome. Prices is separate from the headline PMI, and the Employment Situation can be published before the ISM release in some months.

Why this indicator arrives ahead of the rest

Check the ISM publication schedule and convert its Eastern time for the specific date. It provides an early survey of the preceding month, but does not invariably precede every employment release. Publication speed is different from an indicator leading business-cycle turning points.

A series is economically leading when its turning points tend to precede the chosen reference cycle. Early publication alone does not establish that property or forecasting ability. Examine the component, reference series, revisions and information available at each date.

How the ISM moves the dollar

Separate level, monthly change and surprise. In a hypothetical example, new PMI 52.0 versus previous 51.5 and consensus 50.8 gives a level above 50, a monthly increase of 0.5 index points and surprise +1.2 points. None is a percentage output change or a dollar forecast. A higher headline can reflect deliveries as well as demand; Employment is not an NFP count and Prices is not consumer inflation.

Do not infer a fixed currency-impact ranking from a sector’s size. The ISM Services indicator is a separate survey with a different composite; a simple average of the two headlines is not an official national PMI. Compare matching measures with the gross domestic product release, preserving each definition and period. Neither a sector threshold nor an estimated historical GDP relationship determines the next Fed decision.

What to do after closing this article

  1. Put the next release date in your calendar. Open the economic calendar at your broker or on a macro portal and check the official release date and set a reminder for 10:00 US Eastern Time converted to your own zone. Flag the event as a high-impact read so a spike in volatility on dollar pairs does not catch you off guard.
  2. Read the composition and separate series. Check the five headline components and then Prices separately. Record what each one measures; agreement between them does not establish statistical independence or future macro outcomes.
  3. Record three comparisons. Save consensus before publication, then calculate the surprise and monthly change separately. Compare the headline with its sector threshold as a third question. A result above 50 but below consensus is coherent; these differences do not determine a price reaction.
  4. Match definitions across the calendar. Check the actual ordering and periods of labour-market and inflation releases. Do not assume employment data always follow ISM or convert its Employment or Prices indices into payroll jobs or CPI. See fundamental analysis in our library for the wider framework.