CapEx to-revenue ratio by industry: comparing capital intensity

Comparing a utility’s capital spending ratio directly against a machinery maker’s or a software vendor’s without checking the numerator will misclassify reinvestment needs. A company or industry comparison requires matching the CapEx formula, the accounting perimeter and the asset base before reading any percentage.

To compare the CapEx to-revenue ratio by industry, first verify whether the source reports gross capital expenditures divided by revenue or net capital expenditures (after deducting depreciation) divided by sales, and whether sector figures are cumulated dollar totals or simple firm averages. In NYU Stern’s January 2026 U.S. dataset across 4,822 nonfinancial listed firms, cumulated Net CapEx to sales is 4.29% and Sales to invested capital is 1.43, spanning from 43.72% Net CapEx to sales in water utilities (Sales to invested capital of 0.24) to 4.14% in machinery (Sales to invested capital of 1.95). Place these comparisons inside the wider capital spending research guide before moving to company filings.

How do the CapEx formula and capital intensity metrics differ?

Capital intensity can be measured either as an annual cash flow relative to sales (CapEx to revenue) or as an installed balance-sheet stock relative to sales (PP&E to revenue or Sales to invested capital). Using one without the other can confuse a temporary construction pause in a heavy industry with an asset-light business model.

When reading a U.S. public company’s annual report on SEC Form 10-K, the cash-flow and balance-sheet inputs sit in Part II, Item 8 (Financial Statements and Supplementary Data), while forward commitments and liquidity discussion appear in Part II, Item 7 (MD&A). The SEC’s guide to financial statements explains that cash outflows to buy property, plant and equipment appear under investing activities on the statement of cash flows, whereas depreciation allocates historical cost on the income statement.

Researchers encounter three distinct formulas labeled as a “CapEx formula” or capital intensity ratio:

1. Direct cash-flow CapEx formula (preferred for issuer filings): Extract cash paid for additions to property, plant and equipment directly from the investing section of the consolidated statement of cash flows. Exclude business acquisitions and deduct any government grants or noncontrolling-interest contributions only if you record the exact company reconciliation, as illustrated in our oil company profits and capex guide.
2. Accrual balance-sheet approximation: Textbooks often write CapEx = Ending Net PP&E - Beginning Net PP&E + Depreciation. In real company filings, that shortcut sweeps in property acquired through mergers, asset write-downs, disposals, foreign-currency translation and newly recognized lease assets. Use the cash-flow statement line first, and read the PP&E footnote to explain balance-sheet changes.
3. Net CapEx and Invested Capital turnover (sector datasets): In Aswath Damodaran’s variable definitions at NYU Stern, Capital Expenditures is the cumulated cash-flow capital spending for the sector (generally excluding acquisitions), Net Capital Expenditures is defined as capital expenditures minus depreciation, and Invested Capital is book value of debt plus book value of equity minus cash.

Pairing a flow ratio (Gross CapEx ÷ Revenue or Net CapEx ÷ Sales) with a stock ratio (Sales ÷ Invested Capital or Net PP&E ÷ Revenue) separates two different questions: how much capital the business requires to exist at all, and how fast it is adding assets in the current year.

How does the CapEx to-revenue ratio vary by industry in U.S. data?

Across U.S. listed companies, heavy infrastructure, utilities, extractive energy and rail transport tie up several dollars of invested capital per dollar of sales and spend heavily on physical plant, whereas machinery assembly, packaging and commercial services turn invested capital faster and require smaller net additions per revenue dollar.

The table below extracts selected industrial, resource, utility and service sectors from NYU Stern’s Capital Expenditures by Sector (US) dataset, dated January 2026. Every ratio in this source is computed from cumulated sector totals (summing the numerator across all listed firms in the sector and dividing by the summed denominator), not as an unweighted average of individual firms.

Three structural patterns stand out when you read the columns together rather than looking at Net CapEx / Sales in isolation:

• Gross spending versus net spending after depreciation: In Oil/Gas (Integrated), four large companies spent $52,222.10 million in gross capital expenditures, yet Net CapEx / Sales is only 0.39% because depreciation and amortization ($51,102.80 million) nearly matched gross additions (CapEx / Deprecn of 102.19%). Treating 0.39% as a “low-capex industry” would confuse net expansion above depreciation with total capital required to maintain production.
• Why Net CapEx / Sales can be negative while gross CapEx is positive: Advertising spent $705.57 million on capital expenditures, but its Depreciation & Amort was $1,885.26 million (CapEx / Deprecn of 37.43%), yielding -1.52% in Net CapEx / Sales. Conversely, where Damodaran’s full reinvestment table incorporates net R&D or acquisition adjustments, Net CapEx / Sales can stay positive even when CapEx / Deprecn is below 100% (as seen in Packaging & Container at 89.84% and Machinery at 67.86%). Always check whether a database reports gross PP&E additions or net reinvestment.
• Flow intensity versus balance-sheet turnover: Transportation (Railroads) has a moderate Net CapEx / Sales of 6.47%, not far from Steel (5.69%) or Packaging & Container (5.14%). Yet railroads generate only 0.46 in Sales / Invested Capital (LTM), compared with 1.80 in steel and 1.84 in packaging, because track networks and rolling stock sit on the balance sheet for decades.

What is a good CapEx ratio, and why does no single benchmark work?

There is no universal “good” CapEx-to-revenue ratio across industries; a defensible ratio is one that covers the sector’s physical replacement requirements while earning a return above the cost of capital on any expansion spending.

Rules of thumb that label any CapEx-to-revenue ratio below 5% or 10% as “efficient” fail as soon as you cross sector boundaries or project phases:

1. Physical asset structure sets the baseline: A water or electric utility (Sales / Invested Capital of 0.24 to 0.37 in the January 2026 U.S. table) must fund treatment plants, substations and distribution lines upfront and recover them over multi-decade regulatory lives. An engineering contractor (Sales / Invested Capital of 4.36) bills labor and project execution on customer sites and needs far less owned plant per dollar of revenue.
2. Maintenance versus expansion stage: As explained in types of CapEx: maintenance and growth, two companies in the same industry can report very different ratios in a single year because one is building a new plant while the other is running an existing line. A low ratio during a multiyear deferral of maintenance can precede a catch-up spending spike rather than signal durable efficiency.
3. Own-versus-lease and outsourcing choices: Under CapEx vs OpEx in manufacturing, a company that uses contract packaging or operating leases shifts cash outflows out of the PP&E additions line on the cash-flow statement. Its gross CapEx-to-revenue ratio will look lower than a vertically integrated peer’s even if both run the same packaging machinery on the shop floor.
4. Denominator volatility in cyclical industries: In mining, steel and oil production, revenue swings with realized commodity prices while committed project spending moves on multiyear engineering schedules. When commodity prices drop, CapEx ÷ Revenue rises mechanically even if management cuts the dollar capital budget. This is why EBITDA valuation in capital-intensive sectors and valuation multiples require multiyear cycle checks rather than single-year snapshots.

Why do aggregate sector ratios and company averages diverge?

An industry’s cumulated ratio (Σ CapEx ÷ Σ Revenue) weights each company by its dollar denominator, whereas a simple average of company ratios weights a $200 million regional supplier equally with a $20 billion producer.

Damodaran’s variable guide notes that sector ratios are calculated by summing the numerator across all firms in the sector and dividing by the summed denominator. The hypothetical two-company example below shows how the same underlying numbers produce different benchmarks depending on the aggregation method and gross-versus-net definition.

Illustrative example only (hypothetical figures for two industrial companies in the same fiscal year and currency, not real company disclosures):

In this hypothetical pair, both companies have nearly identical balance-sheet asset intensity (Net PP&E / Revenue of 60.00% and 65.00%). Yet Company B’s expansion project pushes its single-year Gross CapEx / Revenue to 22.00%. Because Company A is ten times larger by revenue, the cumulated sector gross ratio is 6.55%, while the simple average of the two firms is 13.50%, and the cumulated net ratio is 1.55%. Before citing an “industry average CapEx ratio,” record whether the source is value-weighted (cumulated), equal-weighted (mean) or median, and whether depreciation has been deducted.

How do Census ACES/AIES, NYU Stern sector tables and 10-K filings differ in scope?

Never divide U.S. Census survey capital expenditures by a public company’s 10-K revenue, and never treat the historical Annual Capital Expenditures Survey (ACES) as a current standalone collection.

According to the U.S. Census Bureau’s About ACES documentation (last revised July 24, 2026) and AIES program page (last revised October 1, 2026), ACES transitioned into the Annual Integrated Economic Survey (AIES), which began data collection in March 2024 and consolidated seven annual business surveys (including ACES and the Annual Survey of Manufactures, ASM) into a single survey covering revenues, expenses, payroll, employment and capital spending. See our data sources directory for primary links.

Use the scope comparison table below to verify which dataset matches your research question:

A comparison worksheet you can copy into your research notes

Before comparing capital intensity across two companies or benchmarking a company against a sector table, complete one column per subject and leave any unverified cell blank until you open the primary filing.

Method and limitations

This guide synthesizes primary statistical and regulatory documentation opened on October 17, 2026: NYU Stern’s Capital Expenditures by Sector (US) dataset (dated January 2026) and Variables used in Data Set, the U.S. Census Bureau’s About ACES (revised July 24, 2026) and AIES program pages (revised October 1, 2026), and the SEC’s How to Read a 10-K and Beginners’ Guide to Financial Statements.

Sector ratios from NYU Stern reflect listed U.S. companies as of January 2026 and will shift as commodity prices, depreciation schedules, accounting rules and sample membership change. The two-company comparison is a constructed hypothetical example designed solely to demonstrate weighted versus unweighted aggregation and gross versus net formulas. Nothing on this page ranks “best” high-CapEx stocks or provides investment advice; read our terms and research disclaimer for full usage boundaries.