Return on equity is net income divided by shareholders' equity. If a company earns $20 on $100 of equity, that's a 20% ROE — for every dollar shareholders have tied up in the business, it produced twenty cents of profit last year.
It's one of the cleanest one-number reads on quality: persistently high ROE usually marks a business with a real edge. "Usually" is doing work in that sentence, though.
What actually drives it
The DuPont breakdown splits ROE into three levers, and it's worth knowing because it tells you why a number is what it is:
- Net margin — profit per dollar of sales. Pricing power and cost control.
- Asset turnover — sales per dollar of assets. How hard the asset base works.
- Equity multiplier — assets per dollar of equity. This is leverage.
Multiply the three and you get ROE. The catch is the third lever: a company can lift ROE simply by borrowing more, not by running the business better. Two firms can both post 20% ROE — one because it's genuinely efficient, the other because it's loaded with debt. Same headline, very different quality.
A quick example
| Driver | Value |
|---|---|
| Net margin | 10% |
| Asset turnover | 1.0x |
| Equity multiplier | 2.0x |
| ROE | 20% |
Swap the equity multiplier to 1.0x (no leverage) and the same operating performance yields a 10% ROE. The business didn't change — only the financing did.
How to read it
- Look at ROE over time and against peers, not in isolation.
- Always check the equity multiplier or debt level alongside it. High ROE on low debt is the good kind.
- Cross-check with ROIC (return on invested capital), which includes debt in the base and so isn't flattered by leverage.
Bottom line
ROE tells you how much profit a company wrings from shareholders' capital. High is good — but decompose it before you cheer, because leverage can manufacture a great-looking number from a mediocre business.