ISM Manufacturing Hits 55.6 (July 2026): A Four-Year High Built Partly on Worse Delivery Times — and Prices Eased to 71.1 While Hike Odds Fell
July ISM printed 55.6, a four-year high vs 54.0 expected, while Prices eased to 71.1. Why part of the beat was supply friction — and why the dollar shrugged.
ISM Manufacturing Hits 55.6 (July 2026): A Four-Year High Built Partly on Worse Delivery Times — and Prices Eased to 71.1 While Hike Odds Fell
The July ISM Manufacturing Report On Business landed on Monday 3 August at 55.6%, up 2.3 points from June's 53.3% and the strongest reading since May 2022, against a 54.0% consensus. Production surged 6.3 points to 58.5%, the highest since November 2021, and the Employment Index crossed into expansion at 52.8% for the first time in 33 months. The Prices Index — the number this piece said would be the referee — went the other way, easing to 71.1% from 73.0% for a third consecutive decline. So the S&P Global flash that showed input costs at a 14-month high did not get confirmed; in its own final print, published the same morning, S&P Global revised that away too, reporting input cost inflation at a four-month low. And the dollar did almost nothing: the dollar index sat near 99.82, and September hike odds fell to 64.5% from about 67% on Friday.
Two things in this report are worth more than the headline. The first is that the same morning produced two US manufacturing PMIs for the same month — ISM at 55.6% and S&P Global at 53.9% — and the gap is largely a construction artefact, not a disagreement about the economy. The second is that the Prices Index fell while the panel's own written comments got angrier, one executive telling ISM the pricing volatility was "arguably worse than the pandemic era." Both are cases where a number and the thing it measures point in different directions, which is exactly where a factor-by-factor read earns its keep.
- The print: Manufacturing PMI 55.6% in July, +2.3 from 53.3%, highest since May 2022 (55.9%), versus 54.0% expected. Seventh straight month of expansion; 15 industries grew, only Chemical Products contracted.
- Production 58.5% (+6.3) — highest since November 2021 — did most of the work, with New Orders 56.7% (+0.7) and Backlog of Orders 55.0% (+4.5).
- Employment 52.8% (+3.1) entered expansion for the first time in 33 months, four days before the 7 August jobs report.
- Prices eased to 71.1% (−1.9), a third consecutive decline — but above 50 still means rising costs, and raw materials have risen 22 months in a row.
- Supplier Deliveries 58.9% (+1.5) slowed for an eighth straight month. In ISM's equal-weighted construction, slower deliveries add to the headline; in S&P Global's, they subtract. That is most of the 55.6 vs 53.9 gap.
- The combination that landed was in none of the three scenarios cleanly: strong growth and softer prices, which pushes the growth factor up and the interest-rate factor down at the same time.
- Rate path: the FOMC held at 3.50–3.75% on 29 July in a 9–3 vote; September hike odds were 64.5% at midday Monday (CME FedWatch), down from ~67% Friday and from near 95% before the meeting.
- FX: dollar index ~99.82 and little changed after sliding more than 1.5% the prior week; EUR/USD near $1.1523 after a $1.1559 one-and-a-half-month high; the yen near 157 on joint US-Japan intervention rather than US data.
- See how the interest-rate, growth and commodity factors are scoring all eight majors right now on the live meter.
What actually happened
ISM published the July Manufacturing Report On Business at 10:00 a.m. Eastern on Monday 3 August — the first business day of the month, pushed from Saturday 1 August. The full release is the cleanest source for the internals, and the methodology sits on the ISM website.
| Sub-index | July 2026 | June 2026 | Change | Note |
|---|---|---|---|---|
| Manufacturing PMI | 55.6 | 53.3 | +2.3 | Highest since May 2022; 54.0 expected |
| New Orders | 56.7 | 56.0 | +0.7 | Seventh month of growth |
| Production | 58.5 | 52.2 | +6.3 | Highest since Nov 2021 |
| Employment | 52.8 | 49.7 | +3.1 | First expansion in 33 months |
| Supplier Deliveries | 58.9 | 57.4 | +1.5 | Eighth month slowing (higher = worse) |
| Inventories | 51.2 | 51.4 | −0.2 | Broadly flat |
| Prices | 71.1 | 73.0 | −1.9 | Third consecutive decline |
| Backlog of Orders | 55.0 | 50.5 | +4.5 | Pipeline firmed |
| New Export Orders | 53.0 | 48.5 | +4.5 | Back above breakeven |
| Imports | 55.7 | 52.9 | +2.8 | |
| Customers' Inventories | 40.7 | 42.3 | −1.6 | Too low — supportive of future orders |
Two lines in that table were not on anyone's preview list. New Export Orders returning to 53.0 after months of contraction is a genuine surprise in a tariff-fragmented trade environment — foreign demand for US manufactured goods firmed while the trade barriers stayed up. And Customers' Inventories at 40.7 deserves more attention than its billing: when the customers of manufacturers are holding inventories this thin, restocking demand is stored up rather than spent, which is part of why a Backlog reading of 55.0 has some durability behind it.
Why ISM said 55.6 and S&P Global said 53.9 on the same factories
The S&P Global final US manufacturing PMI for July, published the same morning, came in at 53.9% — revised up from the 53.8% flash but unchanged from June, with output growth at its weakest since March and new orders softening for a third consecutive month. On the surface that is a flat contradiction of a 2.3-point ISM jump. Mostly it is not.
The ISM headline is an equal-weighted average of five sub-indexes: New Orders, Production, Employment, Supplier Deliveries and Inventories, 20% each. Run July's numbers: (56.7 + 58.5 + 52.8 + 58.9 + 51.2) ÷ 5 = 55.6. Supplier Deliveries at 58.9 is the second-largest contributor in that average — and in ISM's construction a higher Supplier Deliveries reading means deliveries got slower. Worsening supply chains pushed the ISM headline up.
S&P Global weights its components differently — New Orders 30%, Output 25%, Employment 20%, Suppliers' Delivery Times 15%, Stocks of Purchases 10% — and inverts the delivery-times series so that it moves in the same direction as the others. Longer delivery times pull the S&P Global PMI down. S&P Global reported July supply chains deteriorating at the second-sharpest rate in four years, with the Middle East conflict causing delivery delays and material shortages. The identical real-world fact was additive in one index and subtractive in the other.
Prices fell, and the panel said it got worse
The Prices Index at 71.1% was a third consecutive monthly decline, following June's 9.1-point collapse to 73.0% that ISM attributed to cheaper energy. Read as a trend, that is disinflation in the goods pipeline — and S&P Global independently agreed, putting input cost inflation at a four-month low while still naming energy prices and tariffs as the drivers.
Read as a level, it is nothing of the sort. Anything above 50 means more panelists paid more than paid less, and raw materials costs have now risen for 22 consecutive months. Deceleration from an extreme is not relief; it is a slower rate of a cost that keeps compounding. The panel commentary made the gap between the number and the experience unusually explicit. Roughly 62% of July comments were negative. Pricing volatility appeared in 57% of those, the Iran conflict in 43%, lengthening lead times in 22% and tariffs in 18%. An electrical-equipment executive told ISM that pricing volatility and lead-time extensions were "arguably worse than the pandemic era," and a primary-metals respondent described "no normalcy in sight in the world of metals" — both reported by CNBC alongside the release.
This is the mechanism worth taking from the print, and it generalises well beyond ISM. A diffusion index answers how many firms saw prices rise, never how much. When the level has been extreme for nearly two years, the index can fall for three straight months while the cost base a manufacturer actually faces is still climbing.
The other thing the print settled was a timing question. Both July surveys that had shown hot input costs were answered during crude's spike — Brent closed at $100.69 on 23 July, then settled 8.7% lower at $88.36 on 27 July, the round trip covered in our oil and commodity-currency analysis. The concern was that a hot ISM Prices Index would be faithfully recording a cost that had already evaporated. It did not happen, and for that reason: the energy component was stale by the time ISM's survey closed, and S&P Global's own final revision walked its flash reading back. What did not go away is tariff pass-through — which shows up in the panel comments rather than in the index level, exactly where a monthly diffusion series is worst at capturing it. The same timing trap caught the July consumer-confidence survey, which closed on 22 July, a day before oil peaked.
Employment crossed 50 after 33 months
The Employment Index at 52.8% is the most under-covered line in the report. Thirty-three months below breakeven covers essentially the whole post-2023 factory downturn, and a 3.1-point move into expansion is a turn rather than noise. ISM Chair Susan Spence had noted in June that 64% of panelists were hiring against 36% managing headcount — a pattern that has now arrived in the index itself.
It is also narrow. Manufacturing is roughly 8% of US payrolls, so factory hiring can inflect without moving the headline print that arrives on Friday 7 August, where June's +57K is the baseline and consensus sits near 83K. The useful application is compositional: if factory payrolls turn positive while the headline stays soft, the softness is in services, which is a different problem for the Fed than a broad labour-market break. Our July jobs report preview sets out how that split routes into the dollar.
Which scenario landed, and how the dollar traded it
Three outcomes were on the board before the print. One was a price rebound with a firm headline — the hawkish case, and what the S&P Global flash implied. Two was an in-line, low-information print. Three was a cracking headline with falling prices.
What arrived was none of them cleanly: a much stronger headline with falling prices. In factor terms that is the growth factor up and the interest-rate factor down, simultaneously — and it explains a dollar reaction that otherwise looks perverse. The dollar did firm intraday on the beat, but the dollar index sat around 99.82, essentially unchanged, after sliding more than 1.5% the previous week. EUR/USD traded near $1.1523 having touched $1.1559, a one-and-a-half-month high. And the odds of a hike at the 15–16 September FOMC, per CME FedWatch, were 64.5% at midday Monday — down from roughly 67% on Friday.
The reason those odds fell on a four-year high is sequencing. The FOMC held the funds rate at 3.50–3.75% on 29 July in a 9–3 vote, with Beth Hammack, Neel Kashkari and Lorie Logan each dissenting in favour of a 25bp hike, as covered in our FOMC July preview. Chair Kevin Warsh called the hold "especially prudent at these uncertain times," reasserted a pledge to "deliver price stability," and said the committee "will not hesitate to act" — without indicating how the split resolves. That took a September hike from near 95% priced before the meeting to the mid-60s, and it is that repricing which governs the dollar now, not a factory survey. A strong-growth, easing-prices print gives a committee whose stated problem is inflation a reason to keep waiting.
A third force was larger than either. USD/JPY had been 163.745 on 28 July; the yen traded near 157 on 4 August after Tokyo and Washington intervened jointly, with Bank of Japan data showing roughly ¥5.33trn spent on the Friday after a record ¥8.45trn the day before — our analysis of the intervention covers the mechanics. When an official operation of that size is live, a US data surprise competes for the tape rather than owning it.
How Pip Theory reads it across the five factors
Pip Theory scores the eight majors — USD, EUR, GBP, JPY, CHF, CAD, AUD, NZD — from five fundamental inputs: interest rates, growth, positioning, risk sentiment and commodities, refreshed every four hours. The July ISM touched four of them at once and in conflicting directions, which is why decomposing it beats summarising it.
The headline, Production, New Orders and Backlog feed the growth factor, and specifically the relative growth gap: a US factory sector printing its best month in four years while eurozone and UK manufacturing lag supports the dollar through a channel that has nothing to do with rates. The Prices Index feeds the interest-rate factor, because it is a forward read on the inflation that determines the policy path rather than the policy rate itself — and in July it pushed that factor down, not up. Supplier Deliveries at 58.9 and the tariff commentary feed the supply side of the growth read, where they lift costs while capping output. The whole report feeds risk sentiment and commodities through the global-cycle channel, which is where AUD, NZD and CAD pick it up.
Score those separately and the flat dollar stops being a puzzle. Growth up, rates down, positioning stretched after a 1.5% weekly slide, and an official yen operation dominating flow — that nets to approximately nothing, which is what the dollar index did. A composite "strength" number would have shown a modest tick and told you nothing about why, or about which of those four channels is most likely to reverse first.
Bottom line
July's ISM was a genuine upside surprise — 55.6%, a four-year high, with Production at a 2021 high, Employment out of a 33-month contraction and Export Orders back above breakeven. It was also, in part, a measurement of congestion: Supplier Deliveries slowing for an eighth month added to a headline that S&P Global's inverted treatment of the same fact subtracted from, which is most of why the two July PMIs printed 1.7 points apart. And the Prices Index that was set up as the referee ruled against the flash — 71.1%, a third decline — while the panel writing the survey described conditions worse than the pandemic. Deceleration from an extreme level is not the same as relief, and a diffusion index cannot tell you the difference.
For the dollar, none of that was the main event. A 9–3 hold and an unresolved committee took September from near-certain to a coin flip with a lean, and joint intervention in the yen took the FX tape. The next real test is Friday's payroll report, where the ISM's employment turn is a narrow but early clue about where any softness is actually located.
Educational macro context only — not investment advice.