In 30 seconds
- ROIC divides after-tax operating profit (NOPAT) by the capital that shareholders and lenders have put into the business.
- Unlike ROE, it can't be pumped up with debt: it measures the quality of the business, not how it's financed.
- A company creates value when its ROIC beats its cost of capital (WACC), which for large listed companies usually runs between 7% and 10%.
- A sustained 15% is good in technology and industrials; in consumer staples and pharma the bar sits at 12%.
- Be wary of a ROIC that soars because buybacks have shrunk equity, as Mastercard's did in 2025.
Introduction
In 1972 Warren Buffett and Charlie Munger paid $25 million for See's Candies, a California chocolate maker. The business needed $8 million of capital to run and earned a little under $5 million before tax. Buffett summed it up in his 2007 letter to shareholders: See's was earning 60% pre-tax on the capital it employed. By 2007 it was making $82 million on just $40 million of capital.
That ratio, what a business earns against the money you have to put into it to keep it running, is ROIC. And it explains how a candy company that barely sold more pounds of chocolate each year ended up paying for a long list of other Berkshire acquisitions.
Explanation
ROIC (return on invested capital) measures how much operating profit, after tax, a company squeezes out of every dollar that shareholders and lenders have put to work in the business. A 20% ROIC means every $100 tied up in factories, warehouses, software or inventory brings back $20 a year.
Why isn't ROE enough? Because ROE only looks at the shareholders' money, and a company can pump it up by borrowing. ROIC puts debt and equity together in the denominator, so it doesn't matter how the company is financed: it measures the quality of the business, not its capital structure. That's why it's the favorite metric of anyone hunting for an economic moat. A real moat shows up as a high ROIC that holds for years, not as one good quarter.
Picture two bakeries selling the same bread. One owns its building, runs three ovens and keeps a month of flour in the storeroom. The other rents, bakes in shifts and gets paid in cash. If both earn $50,000 a year, the second is the far better business: it needs less capital to earn the same money. ROIC puts a number on that gap.
How to calculate it
You need two figures. On top goes NOPAT, net operating profit after tax. Take operating income (EBIT) from the income statement and subtract tax at the company's effective rate (income tax expense divided by pre-tax income). You start from EBIT rather than net income so that interest on debt stays out of the picture: ROIC wants to measure the business, not how it's paid for.
On the bottom goes invested capital, the money shareholders and lenders have tied up in the business: shareholders' equity plus total debt. That's the most common definition and the one Kaplio uses, with no cash netted out and goodwill left in. Some analysts subtract cash and strip out goodwill, and their ROIC comes out higher. If you can, use the average of opening and closing capital for the year; if capital barely moved, the year-end figure will do.
There are almost as many versions of ROIC as there are analysts. Aswath Damodaran, a professor at New York University, devoted a whole paper in 2007 to its variants and to the ways it gets dressed up. What matters is comparing companies with the same recipe, and never mixing yours with somebody else's.
The hurdle: the cost of capital
A 9% ROIC is neither good nor bad until you set it against what money costs. That cost is the WACC (weighted average cost of capital): the return that shareholders and lenders, taken together, demand. For a large listed company it usually sits between 7% and 10%. If ROIC beats WACC, every reinvested dollar creates value. If it falls short, the company grows while destroying value, and the more it invests, the worse it gets for shareholders.
Formula
ROIC = NOPAT / Invested capital
NOPAT = EBIT × (1 − effective tax rate)
Effective tax rate = Income tax expense / Pre-tax income
Invested capital = Shareholders' equity + Total debt
ROIC = NOPAT / (Shareholders' equity + Total debt)
Example
The table below compares four payments companies on Kaplio's data, almost all on a trailing-twelve-month basis as of October 4, 2026. Here's how to read it. Mastercard earns about 48 cents of after-tax operating profit a year for every dollar of capital (48.1%). Visa earns about 34 cents (34.2%). PayPal manages about 14 cents (14.5%). American Express comes in at 8.3%, on its 2025 fiscal year. Be careful with that last one: American Express lends money to its cardholders, so it works almost like a bank, and its ROIC doesn't compare with that of a pure payments network.
Step by step: Visa, fiscal 2025 (year ended September 30, 2025; figures from its Form 10-K annual report, in millions of dollars):
1. Operating income (EBIT): 23,994.
2. Effective tax rate: 4,136 of income tax divided by 24,194 of pre-tax income = 17.1%.
3. NOPAT: 23,994 × (1 − 0.171) = 19,892.
4. Invested capital: 37,909 of shareholders' equity + 25,171 of total debt = 63,080.
5. Divide: 19,892 / 63,080 ≈ 31%.
Kaplio shows 28.4% for fiscal 2025 because it normalizes the tax charge on operating income. A three-point gap doesn't change the reading: Visa earns far more than the 15% benchmark.
So the table says 34.2% and the fiscal 2025 figure on Visa's stock page says 28.4%. That isn't a mistake. The table uses the trailing twelve months through October 2026; the walkthrough uses the fiscal year that closed in September 2025. When you compare companies, always use the same period for all of them.
Taking the stock-page figure, Visa earns 28 cents a year for every dollar invested. An industrial company that reaches 15% is already considered good. Why such a gap? Because Visa doesn't lend money or make anything. It collects a fee on every payment that crosses its network, and that network needs almost no fresh capital to process more payments.
Your turn. Mastercard, 2025 accounts (10-K, millions of dollars): EBIT of 18,897; income tax of 3,610 on pre-tax income of 18,578. Work out its effective tax rate and its NOPAT. Hint: NOPAT comes to roughly $15,200 million.
Before you look: do you think Mastercard's ROIC beats Visa's? And would that make it a much better business? The answer is under "Pitfalls and limitations."
Real-data example
| Company | Ticker | ROIC |
|---|---|---|
| Visa Inc. | V | 34.2% |
| Mastercard Incorporated | MA | 48.1% |
| American Express Company | AXP | 8.3% |
| PayPal Holdings, Inc. | PYPL | 14.5% |
How to read it
How to read it
These are rough thresholds, the same ones used in the Summary on every Kaplio stock page. In cyclical sectors, look at the five-year average rather than a single year.
| Sector | Weak | Normal | Good |
|---|---|---|---|
| Technology and software | below 10% | 10-15% | 15% or more |
| Industrials | below 10% | 10-15% | 15% or more |
| Pharma | below 8% | 8-12% | 12% or more |
| Consumer staples (5-year average) | below 8% | 8-12% | 12% or more |
| Energy (5-year average) | below 6% | 6-10% | 10% or more |
ROIC doesn't work for banks and insurers: debt is their raw material, so you look at ROE instead. Regulated utilities are a special case. They live with a low single-digit ROIC because the regulator sets their return, and their cost of capital is low too.
Pitfalls and limitations
1. Invested capital has more than one definition. Kaplio uses the most common one: shareholders' equity plus total debt. A popular variant calculates invested capital without cash and without goodwill, to isolate the quality of the business; it produces the highest figures. Neither is wrong. What's wrong is comparing one company's ROIC built with one recipe against another's built with the opposite.
2. The shrinking denominator. If you did the exercise, you already know the answer to the pre-test: Kaplio shows Mastercard with a 48.9% ROIC in 2025, well above Visa's. Now look at Mastercard's shareholders' equity at the end of 2025: $7,737 million, a fifth of its rival's. Mastercard has been buying back stock for years ($11,727 million in 2025 alone), and every buyback reduces equity. The business is excellent, but part of the edge is accounting: when invested capital shrinks, ROIC goes up even though the business hasn't changed.
3. Goodwill. A company that grows by acquisition carries on its balance sheet whatever it paid above the fair value of what it bought (goodwill). Leave goodwill in and ROIC drops; take it out and ROIC rises. The two figures answer different questions: without goodwill you see the quality of the business; with it, whether management paid sensibly for its deals.
4. One year proves nothing. A 30% ROIC in the best year of the cycle isn't worth the same as 18% sustained for ten years.
5. The tax rate in an odd year. A one-off tax refund or charge moves NOPAT without any change in the business. If the effective rate drifts far from normal, use the usual rate instead.
Case in point: Microsoft and the AI bill
Kaplio shows Microsoft with a 21.6% ROIC in fiscal 2025 (year ended June 30, 2025). On a trailing-twelve-month basis, as of October 4, 2026, its ROIC is 20.6%. Operating income, meanwhile, climbed from $128,528 to $155,237 million between fiscal 2025 and fiscal 2026, a 21% jump. So why didn't ROIC follow? Because capital grew just as fast or faster: shareholders' equity went from $343,479 to $442,387 million between the June 2025 and June 2026 year-ends, and Microsoft is pouring that money into data centers for artificial intelligence.
None of this means Microsoft is in trouble. A 20.6% ROIC still clears the 15% bar Kaplio sets for a tech company. It means the investor's question has changed. It's no longer whether today's business is good, but whether every new dollar it invests will earn as much as the old ones. In his 2007 letter Buffett draws the same line between See's and a business that needs a lot of capital to grow: both can be good investments, but they aren't the same investment.
To check whether the price of a high-ROIC company is justified, the next step is enterprise value (EV), which measures what you pay for all that capital, debt included.
Practice on Kaplio
Frequently asked questions
What is a good ROIC ratio?
It depends on the sector. Kaplio treats 15% or more as good in technology and industrials, and 12% or more in pharma and consumer staples, where a five-year average works better. The real test is the cost of capital: a ROIC above the company's WACC, usually 7% to 10% for large listed companies, creates value.
What is the difference between ROI and ROIC?
ROI measures the return on one specific investment, such as a marketing campaign or a new machine: what it earned against what it cost. ROIC measures the return on all of a company's capital, using after-tax operating profit. Use ROI to decide on a project; use ROIC to judge the whole business and compare it with others.
What is the difference between ROIC and WACC?
ROIC measures what the capital invested in a business earns; WACC measures what that capital costs, the average return shareholders and lenders demand. The comparison settles everything. If ROIC beats WACC, every reinvested dollar creates value. If it falls short, the company destroys value even while its sales and profits keep growing.
Does Warren Buffett use ROIC?
The idea runs through his thinking. In his 2007 letter to Berkshire shareholders he praised See's Candies for earning 60% pre-tax on the capital it needed, and contrasted it with businesses that need a lot of capital to grow. Profit measured against the capital a business ties up is exactly what ROIC captures.
Is 30% ROIC good?
On its face, yes: it's double the 15% bar Kaplio sets for tech and industrial companies. But check two things. Is it sustained? A 30% ROIC in the best year of the cycle is worth less than 18% held for ten years. And has the denominator shrunk? Buybacks can inflate ROIC, as with Mastercard in 2025.