Why ROIC matters more than ROE

Here's a number that gets thrown around constantly: return on equity. Buffett cites it. Screeners filter on it. It's on every stock's summary page.

And it's the easier number to fool.

Quick definitions, no jargon. Return on equity asks: how much profit does this company generate per dollar of shareholders' money? Return on invested capital asks: how much profit per dollar of all the money in the business — shareholders' AND lenders'?

The difference is debt. And that difference is the whole story.

The leverage illusion

Imagine two companies, each earning $10M a year on $100M of total capital. Company A funded itself entirely with shareholder money: ROE = 10%. Company B funded itself with $20M of equity and $80M of debt: same business, same profit — ROE = 50%.

Company B looks five times better on the metric everyone screens for. It's the same business. It just borrowed more.

That's not a technicality. Leverage flatters ROE in exactly the environments where you least want to be fooled — late in a cycle, when borrowing is cheap and everything looks great. Then rates rise or revenue dips, and the same leverage that made ROE beautiful makes the equity go to zero first.

Why ROIC can't be flattered this way

All capital counts, however it was raised. A company earning 15% on its total invested capital is creating real value — it's earning far more than that capital would make sitting in Treasuries. A company earning 8% while Treasuries pay 4-5% is barely clearing the hurdle, whatever its ROE says.

This is why our quality screen is built on returns on invested capital sustained across a decade — not ROE, not margins, not growth. A decade, because one good year proves nothing, and because a business that clears its cost of capital for ten straight years has demonstrated something structural: a durable reason it earns more than money costs.

A worked example

Take Deckers (DECK), the company behind Hoka and Ugg — a current Quality Universe member. Over the past decade its return on invested capital has run in the 20–33% range, averaging roughly 24%, and clearing the cost of capital comfortably in every one of those years. Now look at how it's funded: debt-to-equity of about 0.2, on a balance sheet that holds more cash than debt.

That combination is the whole point. Deckers' returns aren't a leverage story — there's barely any leverage to tell. You cannot borrow your way to a decade of 24% ROIC; that number is the business earning far more than its capital costs, year after year, on its own steam. When ROIC is high and debt is low, the two facts reinforce each other: the returns are real, and the balance sheet can survive a bad year without those returns evaporating. That is exactly the profile the screen is built to find — and exactly what a headline ROE, flattered or not by leverage, would never have told you on its own.

The takeaway

When you look at any company, look at both numbers. If ROE is dramatically higher than ROIC, the difference is leverage — and leverage is a loan against exactly the bad year you're supposed to be able to survive.

Our screener runs this test — a decade of ROIC above the cost of capital — across roughly 4,500 US stocks every month. About 150 pass. If you want to see the full framework applied end to end, the complete Microsoft and Nvidia sample analyses are free to read.

---

Editorial analysis, not investment advice. Past performance doesn't predict future results.