US PPI July 2026: Headline Flat at −0.03% Against +0.2% Expected — but Core Ex-Trade Services Is Running at 5.09% Annualised
July PPI came in at −0.03% m/m against +0.2% expected and annual PPI fell to 4.66%. But both core measures firmed, and the ex-trade-services core runs 5.09% annualised.
US PPI July 2026: Headline Flat at −0.03% Against +0.2% Expected — but Core Ex-Trade Services Is Running at 5.09% Annualised
US producer prices were essentially unchanged in July, falling 0.03% month-on-month against a Dow Jones consensus of +0.2%, and the annual rate dropped to 4.66% from a revised 5.54%. That is the soft headline, and it is the least informative line in the release: both core measures firmed, and the measure that strips out distributor margins is running at 5.09% annualised over three months against 2.21% for the measure that includes them. Of the three shapes this report could have taken, it printed the second — headline soft, core firm — which is the one markets read worst in the first minute. And the pricing-date trap held exactly as expected: these are 14 July prices, collected ten days before the new Section 301 duties took effect.
The annual rate falling almost nine-tenths of a point on a flat month is not a disinflation signal. It is a base effect, and a large one: July 2025 was itself a +0.81% month, and when a figure that size leaves the twelve-month window the annual rate falls by roughly that much whatever the new data does. August 2025 was −0.18%, which means the same arithmetic runs the other way in the report published days before the Federal Open Market Committee meets in September.
- The print: final demand −0.03% m/m against +0.2% expected, with the annual rate at +4.66%, down from a revised +5.54%. Goods fell 0.73%; services rose 0.20%.
- The cores firmed: less foods and energy +0.24% m/m, less foods, energy and trade services +0.35% m/m — the second consecutive month in which the ex-margin measure ran hotter.
- The three-month run-rate: annualised, headline is +1.25%, ex-food-and-energy +2.21%, and ex-food-energy-and-trade-services +5.09%. Distributor margins are doing all of the disinflating.
- The annual drop is arithmetic: July 2025 was +0.81% m/m. A flat August 2026 would lift the annual rate back to about 4.85%; a +0.2% month would put it near 5.06%.
- June was revised up: headline from −0.28% to −0.11%, services from +0.21% to +0.47%, core from +0.20% to +0.39%. The revision risk flagged before this release arrived immediately.
- The tariffs are still not in it: reference prices were collected 14 July; the rebuilt Section 301 duties took effect 24 July.
- See how the interest-rate factor is scoring the dollar against the other seven majors right now on the live currency strength meter.
What actually happened
Six numbers carry the release. Here they are against the month before — using June as it now stands after revision, not as it was first published — and against what was expected.
| Measure (final demand) | June 2026 (revised) | July 2026 (actual) | Consensus | July y/y |
|---|---|---|---|---|
| Headline, m/m | −0.11% | −0.03% | +0.2% | +4.66% |
| Goods, m/m | −1.42% | −0.73% | — | +6.35% |
| Services, m/m | +0.47% | +0.20% | — | +3.92% |
| Less foods and energy, m/m | +0.39% | +0.24% | — | +4.16% |
| Less foods, energy, trade services, m/m | +0.11% | +0.35% | — | +4.65% |
Every figure is computed from the Bureau of Labor Statistics published seasonally adjusted index values rather than rounded off a headline, which is why they carry two decimals. The headline that crossed the wires as "unchanged" was −0.03%. A Dow Jones survey of economists had the consensus at +0.2%; no settled consensus figure for the core measures had emerged across reputable surveys ahead of the release, which is one reason the composition matters more than the miss.
Read the bottom two rows against each other. Excluding food and energy, producer prices rose 0.24% on the month. Excluding food, energy and trade services, they rose 0.35%. The ex-margin measure ran hotter for the second month running, and it accelerated while the other decelerated.
The soft headline is a base effect, and August takes it back
The most-quoted number in this release will be the fall in the annual rate, from 5.54% to 4.66% — a drop of 0.88 percentage points in a single month on a monthly print of −0.03%. That combination is only possible through the base.
July 2025 was a +0.81% month for producer prices. Rolling a figure that large out of the twelve-month window mechanically pulls the annual rate down by close to the same amount, and it would have done so even if July 2026 had printed at +0.2% as expected.
The revision the preview flagged arrived first
Ahead of this release, the first of three things that could change the picture was revision: PPI is preliminary for four months after publication, and the spread that flipped in May could move without any new information about July. That risk did not wait.
| June 2026, m/m | As first published | As revised |
|---|---|---|
| Headline | −0.28% | −0.11% |
| Goods | −1.38% | −1.42% |
| Services | +0.21% | +0.47% |
| Less foods and energy | +0.20% | +0.39% |
| Less foods, energy, trade services | +0.13% | +0.11% |
June read at the time as a clean downside surprise driven by goods. It now reads as a much softer one, with the services side revised up by 26 basis points and the core by 19. Everything in the year-on-year comparisons above rests on data that can still move in the same way — including July's, for the next four months. That is not a reason to ignore the release; it is a reason to weight the direction of travel over any single decimal.
Strip the margins and the pipeline is running at 5.09%
Trade services in the PPI measure the margins received by wholesalers and retailers — the difference between what they pay for goods and what they sell them for. That margin applies to imported goods as readily as domestic ones, which is what makes it the closest thing the official statistics have to a live gauge of who is absorbing a cost increase.
Here is the spread between the two core measures through 2026, in percentage points, recalculated on the revised data:
| Month 2026 | Ex food/energy y/y | Ex food/energy/trade y/y | Spread |
|---|---|---|---|
| January | +3.71% | +3.49% | −0.22 |
| February | +3.83% | +3.53% | −0.30 |
| March | +3.94% | +3.64% | −0.30 |
| April | +4.93% | +4.36% | −0.57 |
| May | +4.44% | +5.04% | +0.60 |
| June | +4.73% | +5.00% | +0.27 |
| July | +4.16% | +4.65% | +0.49 |
The flip that happened in May has held for three months, and July widened it again after June narrowed it. Note also that the revised history is not the history published a month ago — April's spread is now −0.57 rather than −0.49, and May's +0.60 rather than +0.48. The shape survives the revisions; the decimals do not.
The three-month annualised rates sharpen it further, because they strip out the month-to-month noise the monthly prints are full of:
| Measure | 3-month annualised, to July 2026 | July y/y |
|---|---|---|
| Headline final demand | +1.25% | +4.66% |
| Less foods and energy | +2.21% | +4.16% |
| Less foods, energy, trade services | +5.09% | +4.65% |
A pipeline running at 5.09% annualised once distributor margins are removed, against 2.21% with them included, is a distribution chain that has stopped expanding its spread. Read alongside the consumer data, it forms a coherent picture: July CPI showed core goods turning positive for the first time in three months, at +0.20%, while headline CPI rose just 0.07%. Costs are moving upstream and beginning to appear in consumer goods prices, and the firms in between are currently taking the difference out of their own margins rather than passing it on in full.
Be careful how much that carries. The two core measures are different baskets with different weights, so the spread is a heuristic rather than a clean decomposition, and retail fuel margins — volatile, and heavily weighted inside trade services — can move it on their own in a month when energy swings. Intermediate demand shows the same tension: processed goods for intermediate demand fell 0.59% on the month but are still up 9.87% over the year, and intermediate services rose 0.47% and are up 5.11%. The queue upstream is not empty; it is just not moving quickly this month.
The pricing-date trap held: these are 14 July prices
Nothing in this report can tell you about tariff pass-through under the current schedule, and it is worth being precise about why rather than treating the soft headline as evidence either way.
The BLS asks most establishments in the PPI sample to report the selling price for a single day: the Tuesday of the week containing the 13th. In July 2026 the 13th fell on a Monday, so the reference date was 14 July. The rebuilt Section 301 tariff program — duties of 10% or 12.5% imposed after the Supreme Court invalidated the earlier IEEPA tariffs in February, on record in the Federal Register notice — took effect on 24 July. The Section 232 pharmaceutical duties set out in the 2 April proclamation began applying from 31 July for companies named in Annex III, with the broader date still ahead.
Thursday's report describes the pipeline as it stood on 14 July. That makes it the clean baseline the August report will be measured against — and the August report is the one that lands in mid-September, days before the Committee meets.
What it means for the dollar
The transmission runs through the interest-rate factor — one of the five the meter scores across the eight majors — and this release does not resolve the September question in either direction.
The 29 July FOMC statement held the target range at 3.50–3.75% by a 9–3 vote, with Beth M. Hammack, Neel Kashkari and Lorie K. Logan all preferring a quarter-point increase. On 11 August the Treasury par yield curve put the 2-year at 4.22% and the 10-year at 4.70%, with the 3-month at 3.89%. A two-year note yielding roughly half a point above the top of the current target range is a market carrying real probability of tightening rather than easing, and that is the part of the curve an exchange rate trades off.
What this week's two inflation reports delivered is the same shape twice: headline measures cooling, core measures holding. CPI headline rose 0.07% with the annual rate at 3.36%; PPI headline fell 0.03% with the annual rate at 4.66%. Underneath, CPI core goods turned positive and PPI's ex-margin core accelerated. Neither camp on the Committee gets a decisive argument from that pair of reports, which is the substantive result here: a large headline miss that does not actually move the underlying question.
For a fuller read of how these factors combine into a currency score, see the dollar's factor page, the July CPI report that landed the day before this one, and last month's PPI note, which is the direct comparison. The July FOMC decision note covers the meeting that produced the three dissents. The methodology page explains what the five factors are and how the meter is put together.
What would change the picture
Three things would materially alter the read set out above.
The first is further revision, which has now demonstrated it is not a hypothetical. June moved 17 basis points on the headline and 26 on services within a month. July's figures carry the same exposure for the next four months, and the three-month annualised rates above are the most revision-sensitive numbers in this note precisely because they lean on three preliminary months at once.
The second is energy. Crude has been trading above $83 with the Strait of Hormuz still disrupted, and energy is both a large direct weight in producer goods and the input to the retail fuel margins inside trade services. Goods have now fallen for two consecutive months, by 1.42% and 0.73%, after monthly gains near 2% in March, April and May. A large enough energy move dominates both sides of the absorption comparison and makes the margin signal unreadable for a month or two.
The third is the August report itself, published in mid-September. It is the first to price the post-24 July tariff schedule, it lands within days of the FOMC decision, and it faces the easy 2025 base described above. A firm month there would arrive with the annual rate rising rather than falling, which is a materially different backdrop for the same debate.
Educational macro context only — not investment advice.