Every company claims a competitive advantage. Strategy decks are full of ecosystems, flywheels, and "unique positioning." Return on invested capital is how you check the claim against the accounts — because a real economic moat must eventually show up as one specific, measurable thing: the ability to earn high returns on capital and keep earning them while competitors try to take them away.
What ROIC actually measures
ROIC answers a blunt question: for every dollar the business has tied up, how many cents of after-tax operating profit does it produce each year?
The standard formula is NOPAT ÷ invested capital — net operating profit after tax, divided by the capital invested in operations (roughly: equity plus debt, minus cash the business doesn't need). A company with $10 billion invested that generates $2.5 billion of NOPAT runs a 25% ROIC.
Three things make ROIC the right lens for moat analysis, where more familiar metrics mislead:
- It ignores financing tricks. Return on equity can be pumped up with leverage — borrow more, shrink the equity base, ROE rises while the business itself hasn't improved. ROIC charges the company for all the capital it uses.
- It ignores margin illusions. A 40% operating margin sounds great until you learn the business needs enormous capital to produce it. Grocery chains run thin margins with fast capital turns; software runs fat margins on almost no capital. ROIC nets the two effects into one comparable number.
- It has a natural benchmark. Capital has a cost — the weighted average cost of capital (WACC), typically somewhere in the 7–10% range for large companies. ROIC above WACC creates value; ROIC below it destroys value regardless of how fast revenue grows. Growth is only worth paying for when the spread is positive.
Why a sustained ROIC is moat evidence
Here's the economic logic. High returns on capital are an open invitation: any industry earning 25% on capital will attract entrants and imitators until returns get competed back toward the cost of capital. Economists count on this mean reversion, and for the average company it works exactly as advertised.
Which makes the exceptions informative. McKinsey's long-run studies of US corporate returns found that high-ROIC companies tend to hold their advantage: a firm earning above 20% ROIC had roughly even odds of still being in the high-return group a decade later, and top-tier performers sustained mid-teens or better returns over very long horizons. Growth rates, by contrast, decayed toward the average far more reliably. Persistent excess returns are rare enough that when you find a decade of them, something structural has to be blocking the competition.
That "something" is the moat. This is why ROIC is the first of the financial fingerprints in our moat identification checklist: one great year proves nothing — cyclical peaks and one-off gains produce those routinely — but ten consecutive years of 20%+ returns force an explanation.
Reading the number in practice
A workable screen, applied to at least a decade of history:
- Level — is ROIC comfortably above WACC? A 2-point spread is noise; a 10-point spread is a signal.
- Persistence — has the spread survived recessions, competitor product cycles, and management changes?
- Trend — is it stable or widening? A drifting-down ROIC can be the earliest visible symptom of moat erosion, showing up years before market share does.
- Incremental returns — what does the company earn on newly invested capital? A legacy business can coast on old high-return assets while reinvesting today's cash at mediocre rates. The incremental number is what your future returns as a shareholder actually compound at.
Companies like Visa (V) and Apple (AAPL) are canonical illustrations of the pattern — decades of returns far above any plausible cost of capital, each traceable to identifiable moat sources rather than luck.
Where ROIC breaks
The metric has failure modes worth knowing before you trust it:
- Acquisitions muddy the base. Goodwill from deals inflates invested capital and depresses reported ROIC. Whether to include it depends on your question: include goodwill to judge management's capital allocation, exclude it to judge the underlying business quality. The two answers can differ dramatically for serial acquirers.
- Asset-light extremes make it meaningless. When a business runs on almost no invested capital, ROIC explodes to three-digit percentages that no longer distinguish good from great. At that point margins, pricing power, and retention tell you more.
- Financials don't fit the formula. For banks and insurers, capital is the raw material, and ROIC as defined above isn't meaningful — return on equity against the cost of equity is the right frame instead. (This is also why MoatScan's methodology values financials with dividend-based models rather than the standard DCF — balance-sheet-driven businesses get different treatment throughout.)
- Peak-cycle flattery. Commodity producers and semiconductors can print spectacular ROIC at the top of a cycle. The decade-long view exists precisely to average this out.
The symptom, not the cause
The most important discipline: ROIC tells you a moat probably exists. It cannot tell you what the moat is — and without naming the mechanism, you can't judge whether it will last. High returns with no identifiable source among the five pillars (network effects, switching costs, intangibles, cost advantages, efficient scale) should make you more suspicious, not less: unexplained excess returns are the ones competition tends to solve.
So use the number as a filter, then do the qualitative work. MoatScan's approach runs both layers on any US-listed stock — the financial analysis surfaces return metrics like ROE and ROA in its Key Ratios section, while the moat analysis argues the five pillars with evidence and grades durability separately. Pick a long-term compounder from the wide moat stocks list and check whether its return history matches its moat story — or browse the full analysis database and hunt for the rare disagreements. Those are usually where the interesting work is.