How to Analyze a Company's Capital Expenditure Decisions

If you want to know what a management team actually believes about its business, skip the earnings call script and go read the capex line. Capital expenditure is cash committed to future capacity, and it is where value gets created or destroyed. The test is old and unforgiving: a company that earns returns on invested capital above its cost of capital creates value, and one that earns below it destroys value, however impressive the revenue growth looks along the way.

In practice, capex tends to get a quick glance. Check that free cash flow is positive, move on. That leaves a lot on the table, because the size, mix, and productivity of capital spending will tell you more about the next five years than the current income statement will.

Where the Numbers Live

Everything you need is in the 10-K, mostly in three places. The cash flow statement reports capex in the investing section, usually labeled purchases of property, plant and equipment. The segment footnote breaks out capital spending by business unit, which is where you catch a company pouring money into a division that gets two sentences in the shareholder letter. And the liquidity and capital resources section of the MD&A usually includes management's expected capex for the coming year.

Two more spots are worth the click on EDGAR. The commitments footnote discloses contractual obligations to spend, which tells you how locked in the program is. The property footnote shows construction in progress, and a large, growing construction in progress balance means cash is going out the door for assets that are not earning anything yet.

Maintenance vs. Growth Capex

The most useful cut of the number is the split between maintenance capex, meaning what it costs to keep the existing business running, and growth capex, meaning new capacity, new locations, new product lines. A dollar of maintenance spending keeps you where you are. A dollar of growth spending is a bet on getting bigger. The cash flow statement lumps them together, which hides the difference between a company investing aggressively and a company stuck on an expensive treadmill.

Companies rarely disclose the split, so you estimate it. The quick proxy is depreciation. Assume maintenance capex roughly equals depreciation expense, since depreciation is the accounting measure of existing assets wearing out, and treat anything spent above that as growth. The proxy is crude, and it helps to know how it breaks. Inflation means replacing an asset usually costs more than the original did, while in technology the replacement often costs less. As a directional tool, though, it holds up fine.

For a sharper estimate, Bruce Greenwald's method from his value investing work at Columbia scales things off revenue: average the ratio of gross plant to sales over several years, multiply that by the year's revenue growth to approximate growth capex, and call the remainder maintenance. My favorite shortcut is to check whether management simply tells you. Miners and oil and gas producers routinely split sustaining capital from growth capital in their disclosures, and plenty of industrial companies will answer the question on a call if an analyst bothers to ask.

Does the Spending Earn Its Keep?

The core question for any capex program is whether the invested dollars return more than they cost. At the company level, track ROIC through time and set it against the capex trend. Stable or rising ROIC while capex grows is what a healthy reinvestment engine looks like. ROIC drifting down while capex climbs usually means the best projects are done and management is now funding second-tier ideas, which is a normal condition of maturing companies but worth pricing in.

You can get more surgical with incremental returns. Take the change in after-tax operating profit over, say, five years and divide it by cumulative capex (or the change in invested capital) over the same stretch. Suppose a company spent a cumulative $5 billion on capex across five years and grew after-tax operating profit by $400 million. That works out to roughly an 8 percent incremental return, and if the cost of capital is around 10 percent, the growth destroyed value even though every headline number went up.

Watch announced project returns too. When management green-lights a major facility, they often cite an expected return or a payback period. Write it down and check back in three years. A team that habitually promises mid-teens returns and delivers single digits is telling you how to read every future announcement.

Capex Intensity and Its Trend

Capex as a percentage of revenue varies enormously by industry. Telecom carriers, utilities, and energy producers routinely reinvest a mid-teens percentage of revenue or more, while software companies often spend a low single-digit share. Absolute intensity mostly tells you what industry you are in, so I pay more attention to the trend and to the gap versus peers.

Within one company, rising intensity is ambiguous on its own. It can mean a deliberate growth push, or it can mean the asset base is aging and costs more just to keep alive. The maintenance-versus-growth estimate is how you tell those apart. Falling intensity is just as ambiguous, since it can mean a big build-out finished or it can mean the company is starving its assets to dress up free cash flow.

Against peers, the standard is simple. A company outspending its competitors should show better growth or better unit economics with a lag. If it spends more and grows at the same rate, its capital efficiency is worse, whatever the strategy deck claims. A useful companion metric is capex divided by depreciation. Sustained readings well above one mean the asset base is expanding. Sustained readings below one mean the company is consuming assets faster than it replaces them, which works for a while and then abruptly does not.

Capex as Competitive Behavior

In capital-intensive industries, the capex line is where competitive strategy becomes visible. A manufacturer that builds newer, more efficient plants earns a structural cost advantage over rivals running older ones. A carrier that lays fiber first forces everyone else to choose between matching the spend or ceding the market. When one player outspends the group for several years running, the effects tend to surface in market share and margins over the following three to five years, long after the spending depressed reported earnings.

Underspending tells you just as much. A company persistently spending less than its peers may be harvesting the business. Current margins look great because the reinvestment bill is being deferred, while the asset base slowly falls behind. This can run for years before it turns into downtime, reliability problems, or lost share. When you see peer-lagging capex and peer-leading margins together, ask how much of the margin is borrowed from the future.

Commitments, Flexibility, and Projects in Flight

Not every capex budget can bend. Some spending is discretionary and can be cut in a bad quarter. Some is contracted, regulated, or half-built and proceeds no matter what the economy does. That distinction determines how well free cash flow holds up in a downturn.

Two places tell you most of it. The commitments footnote quantifies contractual purchase obligations. The MD&A shows you management's posture: language about ranges, priorities, and flexibility to adjust signals discretion, while a list of named projects with firm budgets and dates signals the money goes out whether or not conditions cooperate.

For projects in flight, look for expected completion dates and remaining cost to complete, in the footnotes or on earnings calls. A large project mid-construction sits in an awkward phase where capex is elevated, assets are growing, and no revenue has arrived against them yet. Knowing the commissioning date tells you when the earning phase should begin, and a date that keeps slipping is its own data point.

What It Means for Free Cash Flow

Capex is the bridge between operating cash flow and free cash flow, so the same dollars read differently depending on their quality. A company with strong operating cash flow and heavy capex can be perfectly healthy if the spending earns above the cost of capital and is tied to a defined growth phase. The bad version is structural, where operating cash flow grows, capex grows proportionally forever, and free cash flow never expands. And when a heavy investment cycle genuinely ends while operating cash flow keeps growing, free cash flow can expand quickly, which is often where the stock story gets interesting.

Four Distortions Worth Checking

The headline capex number can mislead in a few specific ways, so run these checks before trusting it.

Capitalization policy comes first. Costs that get capitalized flow into capex and depreciate slowly, while costs that get expensed hit earnings immediately, so a company that aggressively capitalizes software development or other internal costs can flatter current earnings while inflating its asset base. The accounting policies footnote spells out what gets capitalized, and a big jump in capitalized costs deserves a question.

Acquisitions come second. A company can look capital-light on the capex line while buying all of its growth through deals, which sit in a different part of the investing section. If you are judging how much a business reinvests, read the whole investing section, because M&A is often just capex wearing a different label.

Leases come third. Before the lease accounting change (ASC 842 for US filers), a retailer or airline could build its entire footprint through operating leases with nothing appearing in capex and nothing on the balance sheet. The newer standard at least puts lease assets and liabilities on the balance sheet, but leased growth still never touches the capex line, so intensity comparisons between lease-heavy and asset-owning peers need adjustment.

Write-downs come last. An impairment charge on assets built five years ago is a delayed confession that past capex earned less than promised. A pattern of periodic impairments next to consistently high capex is one of the clearer signs of a weak capital allocation process.

A Working Checklist

1. How does total capex compare to depreciation, and what has the trend been over the last five or so years?
2. What is the rough maintenance-versus-growth split, using depreciation as the floor?
3. Is capex intensity rising or falling relative to revenue, and does management's explanation hold up?
4. How does intensity compare to direct peers, and is any extra spending producing extra growth?
5. Is ROIC above the cost of capital, and what do incremental returns over the last five years look like?
6. What major projects are in flight, when do they complete, and how much is left to spend?
7. How much of the budget is contractually committed versus discretionary?
8. Is free cash flow expanding after capex, or does operating cash flow growth keep getting absorbed?
9. Any aggressive capitalization, lease-heavy growth, or a history of impairments?

Working through that list takes maybe fifteen minutes with a 10-K open, and it pays for itself quickly. Capex is one of the few places where you can grade management on decisions made years ago with cash you can actually trace. Run the exercise across a few holdings and you will start noticing which teams treat shareholder capital carefully and which treat it as a budget to be spent.