Energy, metals and technology inputs are putting pressure on production costs, even as several finished consumer-goods categories show much milder price increases. The divergence matters for inflation, corporate margins, investment and household purchasing power.
- The official September picture
- The sector breakdown
- Production costs and import dependence
- Consumer goods and domestic industry
- Pharmaceuticals, biotechnology and life sciences
- Industry and country-of-origin comparisons
- Wages, purchasing power and the 2008 comparison
- Market implications
- NSA, annualization and deflator methodology
- Sources
The official September picture: inflation pressure extends beyond fuel
Import prices help answer two different questions. Are overseas suppliers adding to, or easing, U.S. inflation? And can American businesses still use imported inputs and equipment to control their production costs?
The Bureau of Labor Statistics release for September 2026, published October 16, reported a 1.1% monthly increase in import prices and an 8.1% increase over the year. Nonfuel import prices increased 0.7% during the month and 6.0% over the year. All these import-price changes are not seasonally adjusted (NSA).
That combination matters: an energy shock can spread through transport and manufacturing, while rising nonfuel prices suggest that the pressure is reaching a wider range of purchases.
| Category | MoM | Y/Y | YTD | YTD simple annualized |
|---|---|---|---|---|
| All imports | +1.1% | +8.1% | +7.64% | +10.18% |
| Nonfuel imports | +0.7% | +6.0% | +5.20% | +6.93% |
| Fuels and lubricants | +4.8% | +34.5% | +39.53% | +52.71% |
| Petroleum and petroleum products | +4.6% | +35.4% | +45.85% | +61.13% |
| Natural gas | +1.6% | +96.5% | -22.47% | -29.96% |
MoM and Y/Y are BLS-published changes. YTD is calculated from September 2026 versus December 2025 index levels; the annualized column applies the same simple 12/9 scaling used in the sector breakdown. Both YTD columns are ATN calculations, not separately published BLS annual rates. Calculations use unrounded ratios of published index levels, with rounding only for display.
The broader perspective is important: all import prices are up 7.64% since December, equivalent to a simple annualized pace of 10.18%, while nonfuel imports have risen 5.20%, or 6.93% annualized. Fuel prices are up 39.53% YTD, a 52.71% simple annualized pace. These measures describe the price increases accumulated during 2026 rather than only the latest month.
Natural gas illustrates why the time windows must be read together: its import-price index is 96.5% above September 2025 but 22.47% below December 2025. The positive Y/Y and negative YTD figures use different starting points; they are compatible and tell different parts of the story. The September monthly increase of 1.6% adds the latest direction without replacing either longer comparison.
Tariffs are additional: BLS import-price indexes exclude import duties. They therefore do not measure the full tariff-inclusive acquisition cost faced by U.S. buyers. The eventual burden depends on supplier pricing, exchange rates, substitution and the ability of businesses to absorb costs or pass them on.
The sector breakdown: a sharp divide between inputs and finished goods
The research snapshot below highlights the contrast between upstream costs and several finished-goods categories. It uses the sector calculations supplied to ATN; the detailed YTD index levels and individual series mappings have not been independently reproduced. These figures are presented as a provisional research dataset, separately from the verified BLS headline table above.
Read the columns as Y/Y change | YTD change | simple annualized YTD pace. The annualized column scales the displayed YTD percentage by 12/9, consistent with nine elapsed months. It is neither a compounded annual rate nor a forecast. Values have been recalculated from the displayed YTD figures for consistency.
| Sector or product group | Y/Y | YTD | YTD simple annualized |
|---|---|---|---|
| Foods, feeds and beverages | +2.88% | +1.69% | +2.25% |
| Agricultural foods, feeds and beverages, excluding distilled beverages | +2.53% | +1.13% | +1.51% |
| Industrial supplies and materials | +20.40% | +18.87% | +25.16% |
| Petroleum and petroleum products | +35.36% | +45.85% | +61.13% |
| Unfinished metals related to durable goods | +25.26% | +12.49% | +16.65% |
| Iron and steel mill products | +6.94% | +9.59% | +12.79% |
| Copper | +38.58% | +27.65% | +36.87% |
| Nickel | +6.49% | +8.02% | +10.69% |
| Tin | +60.82% | +48.91% | +65.21% |
| Zinc | +33.78% | +18.94% | +25.25% |
| Iron and steel products, excluding advanced manufacturing | +4.89% | +6.77% | +9.03% |
| Iron and steel advanced manufacturing | +13.50% | +15.51% | +20.68% |
| Computer accessories, peripherals and parts | +22.92% | +21.90% | +29.20% |
| Semiconductors | +15.82% | +15.07% | +20.09% |
| Automotive vehicles, parts and engines | +1.08% | +1.33% | +1.77% |
| Motorcycles and parts | -1.00% | -0.86% | -1.15% |
| Apparel, footwear and household goods | +2.72% | +2.63% | +3.51% |
| Medicinal, dental and pharmaceutical preparatory materials | +1.49% | +0.95% | +1.27% |
| Furniture and household items | +2.77% | +1.90% | +2.53% |
| Biotechnology | -4.59% | -4.01% | -5.35% |
| Life sciences | -4.75% | -4.75% | -6.33% |
The 20.40% Y/Y figure corresponds to the broad industrial supplies and materials category in the BLS September end-use table, rather than fuels and lubricants. Natural gas is shown only in the verified headline table because the supplied research series could not be matched to the BLS end-use series. Biotechnology and life sciences are specialist research groupings; they should not be treated as BEA end-use aggregates or as stock-market returns.
Energy, metals and technology inputs can squeeze the whole production system
The central concern is the breadth of the cost exposure. Energy affects freight, processing and distribution. Metals affect construction, machinery, electrical equipment and manufactured components. Computers and semiconductors affect the equipment businesses use to produce goods and deliver services.
This creates a potential squeeze between rising input costs and limited pricing power. A manufacturer facing higher metal and energy bills may struggle to raise selling prices if customers are already stretched. A service business can face a similar problem when replacing computers, communications equipment or other technology infrastructure.
Contracts, inventories and equipment replacement cycles can delay the effect. Higher acquisition costs may reach margins over several quarters, while more expensive capital equipment can affect depreciation and required investment returns for years. However, price increases alone do not establish how large or persistent that impact will be.
Imports can still be cheaper than domestic alternatives even when their prices rise. The relevant comparison is the relative delivered cost, including duties, logistics, quality and availability. Where domestic substitutes require new capacity, specialist skills or lengthy qualification, substitution may be difficult in the short term. That is a sector-specific constraint, rather than evidence that the economy has no alternatives to imports.
How the pressure reaches CPI, PPI and GDP
Imported consumer goods can influence CPI through retail prices. Imported business inputs can influence the prices domestic producers charge and therefore affect PPI indirectly. PPI itself measures prices received by domestic producers; imported goods are not directly included as domestic producer output.
For GDP, the question is what happens to domestic value added, productivity and investment. Costlier imported equipment may discourage expansion, while productive imported machinery can also support domestic output. Imports are subtracted in GDP accounting to remove foreign production already included in spending; a larger import bill does not by itself establish an equivalent loss of U.S. production.
Moderate consumer-goods inflation can coexist with domestic industrial weakness
The research snapshot shows much milder price changes in vehicles, motorcycles, apparel, furniture and several household-goods categories than in metals or technology inputs. For consumers, this can provide relief relative to the stronger cost increases elsewhere.
For domestic producers, the same pattern can create competitive pressure: imported finished goods may remain relatively affordable while their own energy, materials and equipment bills increase. If domestic firms cannot pass those costs on, margins and investment can weaken.
However, moderate or falling import prices do not measure production volumes. They can reflect competition, exchange rates, supplier margins, productivity or weaker demand. Establishing a disappearance of domestic activity would require output, employment, capacity utilization, sales and import-volume evidence.
The motorcycle category deserves attention in that framework. Its price pattern may be relevant to the competitive environment faced by U.S. manufacturers, including Harley-Davidson. It cannot, on its own, establish the company’s financial condition or outlook. Company deliveries, dealer inventories, financing conditions and margins would be needed for that assessment.
Pharmaceuticals, biotechnology and life sciences: focus on economic value
Moderate pharmaceutical import-price growth and falling prices in the supplied biotechnology and life-sciences groupings point to a different environment from industrial commodities. Lower acquisition prices can benefit purchasers while making it harder for some suppliers to sustain margins.
One investment interpretation is that smaller companies may need to demonstrate stronger clinical and economic value to maintain pricing power. Relevant evidence includes treatment outcomes, reimbursement, manufacturing costs, intellectual property and the ability to reach commercial scale.
That is an analytical hypothesis, rather than a conclusion established by import prices. Broad price indexes do not identify which individual therapies or businesses will succeed. Nor does a price decline necessarily indicate deteriorating demand: competition, quality adjustment and product changes can matter.
Production location also needs careful treatment. A product made abroad by a U.S.-owned company can be an import. Import-price data track the traded product, not the nationality of the ultimate parent company, and do not establish how much of a category is produced by offshore U.S. operations.
Three lenses: end use, industry and locality of origin
These classifications answer different questions and should be compared rather than combined into a single total.
- BEA end-use categories: organize imported goods by broad economic use, helping distinguish consumer purchases, capital equipment and industrial supplies.
- NAICS industry indexes: classify imports by the industry producing the goods, rather than the industry buying them. They help locate the source of price pressure.
- BLS locality-of-origin indexes: show price changes for imports from different countries or regions. They are affected by each source’s product mix as well as pricing and currency conditions.
September’s official NAICS import indexes showed primary metals up 23.6% Y/Y and computer and electronic products up 12.4%, compared with transportation equipment up 1.4%. Those broader industry categories support the picture of uneven price pressure, although they are not interchangeable with the detailed product rows above.
| Origin | Y/Y change |
|---|---|
| Canada | +16.1% |
| Taiwan | +8.8% |
| Germany | +5.8% |
| European Union | +3.7% |
| Mexico | +3.3% |
| China | +3.1% |
| Japan | +1.1% |
These changes do not rank countries by their absolute price level or prove that one country is the cheapest supplier. Germany is part of the EU aggregate, so those rows overlap. Product-level delivered-cost comparisons are needed before drawing sourcing conclusions.
Wages and weekly earnings determine whether the squeeze becomes a demand problem
Persistently high prices become more damaging when household earnings fail to keep pace. Average weekly earnings matter alongside hourly pay because a reduction in hours can weaken purchasing power even when the hourly wage increases.
If real weekly earnings deteriorate while essential costs and financing expenses stay elevated, households may reduce discretionary spending, draw down savings or rely more heavily on credit. That can turn an upstream cost problem into weaker sales and a broader earnings problem.
The 2008 comparison: a warning mechanism, not a forecast
The concern is a possible interaction between squeezed incomes, debt-service burdens, weakening demand and tighter credit. That mechanism warrants monitoring, but import inflation and wages alone cannot establish that a 2008-style crisis is approaching.
The 2007–2009 crisis involved a housing downturn, mortgage losses, leverage and severe financial-system stress. A serious comparison therefore also needs evidence on delinquencies, lending standards, housing, bank exposures and funding conditions. See ATN’s related analysis of long-term Treasury yields and the parallels with 2007.
What the divergence means for markets
- Equities: watch the difference between input-cost increases and selling-price growth. Strong revenue alone may conceal margin pressure.
- Technology and NQ: rising technology investment can support suppliers while higher equipment costs and interest rates weigh on buyers and valuations. The effect varies by company.
- Industrials and consumer discretionary: firms with high materials exposure and limited pricing power may face a more difficult earnings environment.
- Treasuries: persistent import inflation could complicate disinflation, while weaker real income and demand could pull in the other direction.
- Futures: follow oil, metals, the dollar and yields together. An inflation shock and a growth shock can produce different reactions even when both begin with higher input costs.
The next tests are whether supplier-price increases reach domestic selling prices, whether businesses absorb the pressure in margins, and whether household purchasing power remains resilient.
Methodology: NSA, annualization and inflation adjustment
NSA and SA are separate from nominal and real
NSA means not seasonally adjusted. SA means recurring seasonal patterns have been statistically adjusted. BLS import/export price indexes are NSA. Any comparison with an SA CPI, PPI or output series must identify that different basis explicitly. An NSA observation is not automatically a “real,” inflation-adjusted observation.
Percentage changes and annualization
MoM = (current index / previous-month index − 1) × 100. The general table adds MoM alongside the Y/Y, YTD and annualized YTD measures used throughout the sector breakdown.
Y/Y = (current index / same-month previous-year index − 1) × 100.
YTD = (current index / previous December index − 1) × 100. The provisional sector table uses a nine-month interpretation of the supplied YTD figures.
Simple annualized YTD = YTD percentage × 12/9. For example, petroleum’s displayed +45.85% becomes +61.13%. This is a linear pace comparison. The compounded alternative is [(1 + YTD/100)^(12/9) − 1] × 100, which would be approximately +65.4%. Neither calculation predicts the year-end result. Annualizing volatile NSA commodity movements can magnify seasonal and temporary effects.
Where the deflator belongs
The import-price table already measures price changes; no CPI deflator has been applied to it. Applying CPI would produce a relative-price comparison, rather than an import-volume measure.
To assess wages in real terms, use a consistent CPI series over the same interval. The exact calculation is real earnings growth = [(1 + nominal earnings growth)/(1 + CPI inflation) − 1] × 100, with the growth rates expressed as decimals inside the ratio. Subtracting the inflation percentage from nominal growth is an approximation.
To express historical earnings in current-month purchasing power, use historical nominal earnings × (current-month CPI / historical-month CPI). The current CPI is held fixed across the historical series, and its ratio equals one for the current month. Keep the earnings and CPI series on a consistent seasonal basis.
For real imports or production, use an appropriate product or industry deflator and a compatible value series. A household CPI basket is not a universal deflator for industrial inputs, equipment or trade.
Outlook: elevated prices, uneven pressure and a critical income test
Our assessment is that the import-cost picture creates a risk of persistent price pressure and further increases in selected categories over the medium term. The strongest concern is the combination of expensive energy, industrial materials and technology inputs with more restrained finished-goods pricing.
That does not establish a fixed path for overall CPI or PPI. Exchange rates, commodity reversals, productivity, tariffs, demand and business margins will determine how much pressure reaches the domestic economy.
The decisive test is whether wages, hours worked and average weekly earnings can preserve purchasing power. If they cannot, the economy may face a more difficult combination of high costs, weakening demand and credit stress. Those are conditions to monitor closely; they are not sufficient evidence of an inevitable repeat of 2008.
Sources and further reading
- BLS: September 2026 Import and Export Price Indexes, October 16 release.
- BLS index history via FRED: all imports; nonfuel imports; fuels and lubricants; BLS archived December 2025 index levels. Used for the general-picture YTD calculations.
- BLS Table 1: import prices by BEA end use.
- BLS Table 3: import prices by NAICS industry.
- BLS Table 7: import prices by locality of origin.
- BLS: import-price methodology, seasonal adjustment and exclusion of duties.
- BLS: Producer Price Index overview.
- BEA: NIPA Handbook, net exports.
- Federal Reserve History: the Great Recession and its aftermath.
The official figures above refer to the September release available October 18, 2026, and may be revised later. Detailed sector research calculations are identified separately.