Return on equity is the number investors reach for first — one figure that seems to say “how good is this business at making money for its owners?” It's useful, but it hides a trap: a company can post a gorgeous ROE not because it's a great business, but because it borrowed a lot. Learning to see which one you're looking at is one of the highest-return habits in fundamental analysis.
The two ratios, plainly
Return on equity (ROE) = net income ÷ shareholders' equity. Profit earned per dollar the owners have put in. A 15% ROE means 15 cents of annual profit for every equity dollar.
Return on assets (ROA) = net income ÷ total assets. Profit earned per dollar of everything the company uses — equity and borrowed money alike. Because it counts debt-funded assets too, ROA can't be flattered by leverage. That single difference is why you always read the two together.
The DuPont breakdown — where ROE actually comes from
The DuPont identity splits ROE into three things multiplied together:
ROE = net margin × asset turnover × equity multiplier
- Net margin (net income ÷ revenue) — is the company profitable on each sale?
- Asset turnover (revenue ÷ assets) — is it efficient, generating lots of sales from its asset base?
- Equity multiplier (assets ÷ equity) — how much leverage is it using? More debt, higher multiplier, higher ROE.
Two companies can both show a 20% ROE and be completely different animals. One earns it with a 20% margin and no debt — a genuinely excellent business. The other earns it with a 4% margin and five turns of leverage — a thin-margin operation levered up. DuPont is how you tell them apart in thirty seconds, and it's why “high ROE” alone is a headline, not an answer.
Why leverage flatters ROE — and bites back
Debt is the cheat code for ROE. Borrow, buy assets, earn a spread on them, and the return lands on a smaller equity base, so ROE climbs. In good years it looks like brilliance. But leverage is symmetric: the same multiplier that magnifies gains magnifies losses, and a levered company with a soft year can see ROE crater or go negative. This is the whole reason banks — which are leveraged by design — report ROEs around 10–15% while their ROA sits near 1%: the gap between the two is the leverage. For a bank, ROA is the more honest read of the underlying franchise.
How to read them together
A quick, reliable routine: look at ROE for the headline, then ROA to see how much of it survives once leverage is stripped out. If ROE is high and ROA is healthy too, the returns are real. If ROE is high but ROA is thin, you're looking at a balance-sheet story — fine if the leverage is stable and well-funded, dangerous if it isn't. Then run the DuPont split across a few years: a rising ROE driven by margin is a strengthening business; a rising ROE driven only by the equity multiplier is a company taking on risk to keep the number up.
Reading them in AnalystBook
AnalystBook computes ROE and ROA for every company we cover, each year, straight from the reported financials — and handles the awkward cases honestly (a company with negative equity gets “not meaningful,” not a nonsense number). Because the ratios sit next to the income statement and balance sheet they came from, you can trace any figure back to the filing line in a click, and run the margin-vs-leverage question across a company's history without assembling a spreadsheet. For sector-specific reads — banks, insurers, REITs — see the companion guide on bank analysis in five ratios. For research purposes only; not investment advice.