Free Cash Flow Yield

The formula, and what it's actually asking

Free cash flow yield answers one question: for every dollar you pay for this business, how many cents does it generate in real cash after keeping itself running? The formula is direct. Take operating cash flow from the cash flow statement. Subtract capital expenditures — the money spent on property, plant, and equipment. That is free cash flow. Divide by market cap. Multiply by 100. You have FCF yield.

A working example. A company with a $12 billion market cap reports $1.8 billion in operating cash flow and spends $600 million on capex. Free cash flow is $1.2 billion. FCF yield is 10%. For every dollar invested in this stock, the business generates ten cents in real cash each year — not accounting income, not EBITDA, but cash that physically cleared the bank account.

Two variants cause most of the confusion when you start reading research. Levered FCF yield uses cash flows after interest payments, showing what equity holders receive after debt service. Unlevered FCF yield strips out the financing effect, making cross-company comparison cleaner. For most valuation work, unlevered FCF divided by enterprise value is the more honest comparison — it doesn't let a highly leveraged company appear cheap simply because its equity slice is a thin sliver of total capitalization. If you're using market cap in the denominator, debt is hiding in the background. Know that going in.

One thing to skip: "free cash flow" as reported in company press releases. Management routinely adds back stock-based compensation, deal costs, and restructuring charges to arrive at "adjusted free cash flow" — a number that can run 30–50% above the GAAP version in software and technology. Use the cash flow statement from the 10-K, not the earnings release. The GAAP number is the one that reflects real cash.

Why analysts reach for this instead of P/E

Net income is the output of accounting choices, not just business performance. The gap between reported earnings and real economic earnings can be substantial — and every bit of it is legal, disclosed somewhere in the footnotes, and genuinely hard to see without knowing where to look.

Three examples that recur constantly. First: depreciation schedules. A company can extend the assumed useful life of manufacturing equipment from 10 years to 15 years and immediately boost reported earnings, because annual depreciation expense falls. No cash changes hands. Nothing about the factory changed. Earnings went up. Second: revenue recognition timing. A software company selling multi-year contracts has real discretion over how much revenue it records in year one versus later periods, even under ASC 606. Recognition assumption changes create earnings variability with no corresponding shift in underlying economics. Third: recurring one-time charges. A company that takes a restructuring charge every two or three years isn't being one-time — it's running a structurally higher-cost operation and smoothing that reality into periodic lumps.

Cash is harder to manage than income. When a customer pays an invoice, cash lands in the account. When a supplier gets paid, cash leaves. The cash flow statement is not perfectly immune to timing manipulation — companies can delay payables or accelerate receivables collections at quarter-end — but the degrees of freedom are narrow compared to the income statement. Over a full year, the mismatch between accrual accounting and actual cash becomes visible and measurable.

The comparison to earnings yield (the inverse of P/E) is instructive. Both divide "what the business produces" by price. Earnings yield uses net income. FCF yield uses cash. The two numbers converge for simple, stable businesses with predictable asset lives and minimal working capital. They diverge — sometimes dramatically — in capital-intensive industries, companies with heavy acquisition histories, and businesses where revenue recognition is complex. That divergence is itself information. When FCF yield is significantly lower than earnings yield, the gap is worth investigating before you trust either number.

The number that looks the same but means opposite things

Take two companies, both reporting an 8% FCF yield. They look identical on a screener. They are not.

Company A is a consumer staples manufacturer with $40 billion in revenue. It generates $5 billion in operating cash flow and spends $1.6 billion on capex annually — replacing filling lines, upgrading packaging equipment, and maintaining distribution infrastructure. The plants are relatively modern. Depreciation runs approximately $1.7 billion per year, just above capex. That spread is exactly what you would expect from a business replacing assets at roughly the rate they wear out. Free cash flow is $3.4 billion. Market cap is $42.5 billion. FCF yield: 8%. The cash is real, and it will be there next year.

Company B is a capital-intensive industrial manufacturer, also showing 8% FCF yield on a screener. But the internals are different. The company runs plants that are 20–25 years old. Depreciation is $2.8 billion per year — the original asset base was enormous and the accounting clock has been running for decades. Actual capex is only $1.4 billion. On the cash flow statement, that gap inflates free cash flow. But the company is replacing assets at half the rate they are depreciating. The factories are consuming their economic life faster than they are being renewed. At some point — two years out, maybe five — a major overhaul cycle arrives and capex doubles or triples for several years. The "8% FCF yield" compresses to 3%, or goes negative, while that bill gets paid.

This is the central deception in FCF yield for capital-intensive businesses. Depreciation is an accounting proxy for asset consumption. Capex is actual cash spent on renewal. When D&A persistently exceeds capex by a wide margin, the business is living off its asset base. The spread between those two numbers is deferred capital spending. It will show up on the cash flow statement eventually. The only question is the timing.

Working capital adds a second layer of distortion. Operating cash flow begins with net income, then adjusts for non-cash items and changes in working capital. When a company grows revenue while allowing receivables to expand faster — collecting cash more slowly from customers — the working capital change is a use of cash that pulls operating cash flow below what the income statement implies. The inverse also occurs: a company can boost operating cash flow in a single year by aggressively collecting receivables or delaying payables, making FCF look temporarily strong without any underlying improvement in the business. Neither distortion is apparent without reading the working capital section of the cash flow statement. The calculator on this page makes the adjustment explicit and shows you how large it is.

The two-company example is why a single FCF yield figure — pulled from a financial database and screened for "value" — is a starting point, not a conclusion. The same 8% can mean twenty years of durable cash generation or eighteen months before a capex supercycle consumes it. The formula cannot tell you which. The capex line can.

Three signals that tell you the FCF is real

Each of these takes under ten minutes to check using three years of cash flow statements from any company's 10-K filings. None requires a data subscription.

• Capex growing faster than revenue. Pull three years of capital expenditures and three years of revenue from the cash flow statement and income statement. Calculate the compound growth rate of each. If capex is growing at 14% annually while revenue is growing at 6%, the business is getting progressively more expensive to operate — more capital required for each dollar of incremental sales. That is a structural return problem that will compress free cash flow margins over time. The exception is a clearly defined, front-loaded expansion with an articulated payoff timeline — a new facility ramping to production, or a capacity addition tied to a specific contract. That is different from capex creep with no coherent explanation. Ask management to describe the largest capex projects on the last earnings call. If the answer is vague, the creep is structural.
• D&A significantly higher than capex, sustained over two or more years. Depreciation and amortization represents the accounting consumption of assets. Capital expenditures represent actual cash spent to replace or extend them. When D&A exceeds capex by 20% or more for multiple consecutive years, the company is not fully replacing what it is consuming. The industries where this pattern is most common: airlines, which defer heavy maintenance in cash-constrained periods; refining and chemicals, where turnaround cycles create lumpy spending; and older industrials with largely depreciated plant. A sustained 25–30% D&A-to-capex spread is a signal to investigate. Look at the 10-K footnotes for average asset age. If the answer is "fully depreciated" or above 20 years, the deferred spending is real and the FCF yield is overstated.
• Working capital expanding alongside revenue. The operating cash flow section of the cash flow statement includes a line showing changes in operating assets and liabilities. If accounts receivable grows faster than revenue — say, revenue up 9% but receivables up 20% — customers are paying more slowly, and that cash never reached operations. If inventory grows without a corresponding revenue increase, the company is building stock that hasn't sold. Across a full business cycle, companies that consistently expand working capital relative to revenue are not generating as much real cash as the income statement implies. The working capital section of the cash flow statement is where this shows up. Most investors read the headline operating cash flow number and stop there. The working capital detail is three lines down.

None of these signals is automatically a reason to sell or skip a position. Rising capex can reflect a genuine competitive investment. Temporarily elevated receivables can reflect expansion into distribution channels with different payment terms. The point of running these checks is to know whether the FCF yield you are seeing is the real yield or an accounting artifact — and to demand a higher yield, or a compelling explanation, when you can see the gap.

What counts as a good yield — and good compared to what

FCF yield has no universal "good" threshold. It is a relative metric in three directions: relative to the current risk-free rate, relative to industry peers, and relative to the company's own history. A number that looks attractive in one context looks ordinary — or alarming — in another.

The risk-free rate is the most important anchor. If the 10-year Treasury yields 4.3% — roughly where it sat in early 2026 — then an equity with a 4.5% FCF yield is offering a 20-basis-point premium above a government bond with no credit risk, no earnings volatility, and no possibility of permanent capital loss. That is not a compelling proposition. Historically, value investors have looked for FCF yields of 6–10% or more to justify the equity risk premium — that spread varies by sector, leverage, and business quality, but the logic is the same: you want to be paid meaningfully above Treasuries for the additional risk. The same 4.5% FCF yield in January 2021, when the 10-year was at 1.0%, offered 350 basis points of spread. A very different proposition. This is why blanket statements like "a 5% FCF yield is cheap" are context-free and therefore useless.

Industry context narrows the range further. Consumer staples companies — Procter & Gamble, Colgate, Church & Dwight — typically trade at FCF yields of 3–5%, reflecting the market's confidence in the durability and growth of their cash flows. A 3.5% FCF yield on Colgate is the price of certainty. A 3.5% FCF yield on a regional steel producer or a contract manufacturer is alarming — the cash flow is cyclical, the balance sheet may carry significant debt, and the next economic downturn could cut that yield to zero. The same number carries completely different risk profiles depending on the business behind it. Check what the industry's historical FCF yield range has been before deciding where a current number sits in that distribution.

Company history is the third reference point. A business that has traded at an FCF yield of 6–8% for a decade and now trades at 4% is relatively expensive versus its own history, even if 4% seems fine in isolation. The reverse is equally useful: a company historically priced at 5% FCF yield now showing 9% is either genuinely cheap, or experiencing fundamental deterioration the market is pricing ahead of you. History sets the baseline. Divergence from it is a question that demands an answer before you act on the number.

Industries where FCF yield is least reliable

FCF yield is useful in the right context. In the wrong context, it produces confident-sounding numbers that mean close to nothing.

• Lumpy-capex businesses: airlines, utilities, mining. A major airline replacing its narrowbody fleet will spend $3–6 billion in capex over two to three years, then almost nothing on that fleet for a decade. A single year's FCF number — high or low — is noise. A mining company in a five-year development phase on a new open-pit mine shows deeply negative FCF during construction, then cash generation that looks extraordinary at full production. For these industries, normalized FCF over a full capex cycle is the relevant figure. One year's yield is not.
• High-growth companies reinvesting aggressively. Amazon from 2012 to 2016 reported minimal or negative free cash flow while building out AWS and fulfillment infrastructure. Netflix spent years with deeply negative FCF as it built its content library. Low or negative FCF yield in these situations was not a warning — it was evidence of reinvestment into a competitive position that would eventually generate substantial cash. The question for high-growth companies is not "what is the FCF yield today" but "what will normalized FCF yield be when investment spending moderates, and how long until that happens." That is a forecasting problem, not a current-multiple problem. Applying FCF yield to a business in heavy reinvestment mode produces a number that misstates both the risk and the opportunity.
• Financial companies: banks, insurers, asset managers. Standard enterprise value calculations break down for banks, and so does the FCF framework. Banks do not have capex in any meaningful sense — their reinvestment is lending, and their "free cash flow" is bounded by regulatory capital requirements. Insurers have float that creates cash flows structurally disconnected from underwriting operations. For financial companies, the relevant metrics are tangible book value per share growth, return on equity, efficiency ratio, and — for insurance — combined ratio. Running FCF yield on a large bank produces a technically calculable number that is analytically useless.

The common thread across these exceptions: FCF yield works when the relationship between current cash generation and future cash generation is relatively stable and visible. When that relationship is obscured by capex cycles, growth reinvestment, or regulatory capital constraints, the metric needs to be set aside for something appropriate to the actual business model. Using it anyway is not conservatism — it is precision in the wrong unit.