# Global Macro — U.S. factory demand strengthens as input inflation and long yields raise the cost of expansion

September's U.S. manufacturing data describe a more difficult macro configuration than the headline PMI alone suggests. The ISM Manufacturing PMI was almost unchanged at 54.5, but the internals moved sharply: new orders rose to 55.3, employment to 52.7 and backlogs to 56.4, while the Prices Index jumped 6.8 points to 77.9. At the same time, the 10-year U.S. Treasury yield closed 1 October at 5.24% and the 10-year real yield at 2.88%. Demand is expanding, but the price of inputs and the price of long-duration capital remain high.

That combination matters more than a one-tenth change in the headline PMI. Orders and backlogs indicate that factories still have work coming in; contracting producer inventories and customer inventories classified as too low leave room for replenishment; yet higher input prices, slower supplier deliveries and long yields above 5% make the expansion more expensive to finance and execute. The immediate macro question is therefore not whether U.S. manufacturing is expanding. It is whether demand can continue to translate into production without a renewed squeeze on margins, working capital and investment.

Europe offers a different cyclical signal. Euro-area unemployment remained 6.4% in August, but the number of unemployed people was 267,000 higher than a year earlier. U.S. energy trade adds a structural layer: propane exports averaged 2.0 million barrels per day in the first half of 2026, 11% above a year earlier, with flows shifting toward India and South Asia while exports to China fell.

**Information cutoff: 1 October 2026, 20:58 BRT.**

## What changed on 1 October

| Indicator | Latest | Previous / comparison | Analytical signal |
|---|---:|---:|---|
| ISM Manufacturing PMI | 54.5 | 54.6 in August | Expansion continues, but the headline conceals stronger internal movements |
| ISM New Orders | 55.3 | 53.7 in August | Demand intake accelerated |
| ISM Backlog of Orders | 56.4 | 51.8 in August | Unfilled work rose materially |
| ISM Employment | 52.7 | 51.2 in August | Factory employment diffusion strengthened |
| ISM Prices | 77.9 | 71.1 in August | Input-price pressure intensified sharply |
| ISM Inventories | 48.6 | 50.6 in August | Producer inventories returned to contraction |
| 10-year U.S. Treasury | 5.24% | 5.29% on 30 September | Long-duration financing remains restrictive despite a daily decline |
| 10-year U.S. real yield | 2.88% | 2.93% on 30 September | The real discount rate remains high |
| Euro-area unemployment | 6.4% | 6.4% in July; 6.3% in Aug. 2025 | Stable monthly, slightly softer year on year |
| U.S. propane exports, H1 2026 | 2.0 mb/d | +11% year on year | U.S. energy abundance is being transmitted through external trade |

The table separates three channels that should not be collapsed into a single "growth" signal: factory demand and production, the financial cost of carrying that expansion, and external labour and energy conditions.

## The U.S. manufacturing headline is stable; its internals are not

A PMI of 54.5 versus 54.6 in August looks uneventful. The underlying indexes are more informative. New Orders increased 1.6 points to 55.3 and Employment increased 1.5 points to 52.7. The Backlog of Orders Index rose 4.6 points to 56.4. Production remained strong at 56.7, but fell 1.6 points from August.

The resulting pattern is one of continuing demand with a wider gap between incoming work and some parts of the production response. It does not prove an economy-wide supply constraint, but it is consistent with a manufacturing sector where order intake remains firm while inventories and selected inputs are becoming less comfortable.

```chart
type: bar
title: Change in selected ISM manufacturing components, August to September 2026
unit: percentage points
Prices | 6.8
Backlog of Orders | 4.6
New Orders | 1.6
Employment | 1.5
PMI | -0.1
Supplier Deliveries | -0.3
Production | -1.6
Imports | -1.5
Inventories | -2.0
New Export Orders | -2.3
```

Inventories sharpen the interpretation. The ISM Inventories Index fell to 48.6, back into contraction, while Customers' Inventories fell to 41.6 and remained classified as "too low." Low customer inventories can support future orders because buyers eventually need to replenish stocks. But when that restocking pressure appears alongside a Prices Index of 77.9, the same mechanism can increase competition for materials and working capital rather than produce a clean volume expansion.

The September commodity list reinforces the cost signal. ISM respondents reported price increases across metals, electronic components, freight, fuel, memory components, semiconductors, resins and steel products, while no commodities were listed as declining in price. That list is not a consumer-price index and does not establish broad inflation by itself. It does show that purchasing managers were facing a broad input-cost problem inside the manufacturing supply chain.

## Long rates make the industrial expansion more expensive

The financial side of the signal is unusually important. U.S. Treasury data show the 10-year nominal yield at 5.24% on 1 October, after 5.29% on 30 September. The 30-year yield was 5.61%. The 10-year real Treasury yield was 2.88%.

```chart
type: line
title: U.S. 10-year Treasury yield
unit: %
2026-09-24 | 5.18
2026-09-25 | 5.17
2026-09-28 | 5.24
2026-09-29 | 5.26
2026-09-30 | 5.29
2026-10-01 | 5.24
```

The daily decline does not change the level problem. A long nominal yield above 5% and a real yield close to 3% raise hurdle rates for long-lived investment, increase the carrying cost of inventories and make refinancing more expensive. Manufacturers with strong order books can therefore face a paradox: demand justifies capacity or inventory spending, while financing conditions make that spending harder to approve.

This is not evidence that one interest-rate variable is causing the PMI configuration. Treasury yields incorporate several forces and the ISM survey is not a capital-expenditure survey. The analytical point is narrower: the expansion is occurring inside a high-discount-rate environment, so volume growth must clear a higher financing threshold.

```flow
Stronger new orders -> larger backlog and replenishment needs -> greater demand for inputs and working capital -> higher sensitivity to input prices and financing costs -> more selective capacity expansion and margin pressure
```

The transmission is likely to be uneven. Firms with cash, pricing power or strategically scarce capacity can absorb higher financing and input costs more easily. Smaller or highly leveraged producers are more exposed to the same combination. The aggregate PMI can therefore remain expansionary while investment quality, margins and balance-sheet pressure diverge beneath it.

## Europe is not collapsing, but its labour buffer is no longer tightening

Eurostat's August release is a softer signal than the U.S. manufacturing data. The euro-area unemployment rate remained 6.4% for a fourth month, while the EU rate remained 6.1%. There is no evidence in the release of a sudden labour-market break.

The year-on-year comparison is less comfortable. Euro-area unemployment increased from 11.090 million people in August 2025 to 11.357 million in August 2026, an increase of 267,000. EU unemployment increased by 348,000 over the same period. Youth unemployment in the euro area was 15.0%, down slightly from July but above 14.8% a year earlier.

National dispersion remains large. Germany was at 4.0%, France at 8.2%, Spain at 10.0% and Finland at 10.3%. That heterogeneity limits any interpretation that treats the European labour market as a single homogeneous cycle.

For the global comparison, the relevant point is not that Europe is "weak" and the United States is "strong." The data are measuring different things and different reference periods. The useful contrast is that U.S. factories are reporting firmer orders and employment diffusion in September while the European unemployment stock has edged higher over the year. Monetary and corporate transmission therefore need not be synchronized across the Atlantic.

## Propane exports reveal a structural trade shift behind the cyclical data

The EIA's 1 October energy release adds a different kind of evidence. U.S. propane exports averaged 2.0 million barrels per day during the first half of 2026, 11% above the same period of 2025. Exports reached a record 2.1 million b/d in April.

The driver is not simply stronger global consumption. Record U.S. propane production has widened the price advantage of U.S. supply relative to Asia, while trade flows have been reallocated. Shipments to India and other South Asian markets increased, whereas U.S. propane exports to China fell 19% year on year in the first half of 2026.

```map
title: Reallocation of U.S. propane trade
United States -> India and South Asia | Higher flows | Record U.S. production and favorable relative prices support exports
United States -> China | Lower flows | H1 2026 propane exports were 19% below the same period of 2025
United States -> Asia via Panama Canal | Logistics constraint | Low water levels raise transit costs and restrict effective shipping capacity
```

This creates a structural chain that is partly independent of the monthly manufacturing cycle. Natural-gas production supports propane output; propane becomes an exportable petrochemical and heating feedstock; terminal and shipping capacity determine how much of the price advantage can reach overseas buyers.

EIA identifies two constraints to further expansion. Export capacity can become binding, although an Enterprise project at the Houston Ship Channel and Nederland terminals is expected to add 300,000 b/d of capacity. Panama Canal drought conditions also reduce transit capacity and increase effective shipping costs. U.S. supply abundance therefore creates an external advantage, but infrastructure and geography determine how much of it can be monetized.

## The combined regime: demand is holding, but the expansion is expensive

The 1 October evidence is better described as **costly resilience** than as simple macro dispersion.

U.S. manufacturing demand is still expanding and order books strengthened. At the same time, the purchasing-price signal is worsening and long nominal and real yields remain high. Europe has not experienced a labour-market break, but unemployment is no longer improving on a year-on-year basis. U.S. energy supply continues to create export capacity, but the trade gains are being redirected across destinations and constrained by terminals and shipping routes.

The second-order implication is that headline growth can remain positive while the distribution of gains becomes narrower. Firms with scarce capacity, short payback periods, pricing power or internal cash generation can continue investing. Firms dependent on cheap long-term finance or imported inputs face a less favorable equation. Stronger orders do not automatically produce proportionally stronger output when input prices, inventories, logistics and capital costs are all binding at the margin.

This regime also complicates a clean disinflation narrative. The ISM Prices Index is not CPI, PPI or PCE, and it should not be treated as a forecast of those measures. But a jump from 71.1 to 77.9 means manufacturing purchasing managers are reporting faster price increases at the same time that demand subindexes remain expansionary. The next inflation releases therefore matter for distinguishing a sector-specific cost shock from broader price persistence.

## What would invalidate this reading

A rapid reversal in new orders and backlogs would weaken the demand side of the thesis. A fall in the ISM Prices Index back toward neutral territory, combined with faster supplier deliveries and rebuilding inventories, would reduce the evidence of a manufacturing cost squeeze. A sustained decline in nominal and real long-term Treasury yields would lower the financing constraint.

On the European side, a renewed decline in unemployment would show that the year-on-year deterioration was temporary rather than an emerging softening trend. In energy, a reversal in propane exports or persistent infrastructure critical constraints that prevent new terminal capacity from translating into shipments would weaken the structural export argument.

The next near-term test is the U.S. Employment Situation for September, scheduled for 2 October at 08:30 ET. ISM Services follows on 5 October. U.S. CPI for September is scheduled for 14 October, and Eurostat's next unemployment release is scheduled for 30 October.

## Sources

- Institute for Supply Management, September 2026 Manufacturing PMI Report, 1 October 2026: https://www.ismworld.org/supply-management-news-and-reports/reports/ism-pmi-reports/pmi/september/
- U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates, 1 October 2026: https://home.treasury.gov/resource-center/data-chart-center/interest-rates/
- Board of Governors of the Federal Reserve System, H.15 Selected Interest Rates, release dated 1 October 2026: https://www.federalreserve.gov/releases/h15/
- Eurostat, Euro area unemployment at 6.4%, 1 October 2026: https://ec.europa.eu/eurostat/en/web/products-euro-indicators/w/3-01102026-ap
- U.S. Energy Information Administration, U.S. exports of propane reached records in the first half of 2026, 1 October 2026: https://www.eia.gov/todayinenergy/detail.php?id=68244
- U.S. Bureau of Labor Statistics, October 2026 release calendar: https://www.bls.gov/schedule/2026/10_sched_list.htm