TL;DR: Return on equity, or ROE, divides period net income by average shareholders' equity; $2 million of annual net income divided by $10 million of average equity gives 20 percent. It measures how much profit the business generates from the capital shareholders have in it. Read alone it is ambiguous, because debt can inflate it. Read through the DuPont decomposition, into margin, asset turnover, and leverage, it becomes one of the most instructive ratios in financial analysis.
Key Takeaways
- ROE equals net income divided by average shareholders' equity. Use average equity across the period, not just the ending balance, when equity changed materially.
- The DuPont breakdown expresses ROE as profit margin times asset turnover times financial leverage, showing whether returns come from pricing, efficiency, or borrowing.
- High ROE built on leverage is fragile. High ROE built on margin and turnover with moderate debt is the durable pattern investors look for.
- Buybacks shrink equity and can raise ROE mechanically, and negative or tiny equity makes the ratio meaningless.
- ROE is a core quality input in fundamental screening, as covered in fundamental stock screeners.
How do you calculate ROE properly?
Take net income for a full year and divide by shareholders' equity. Because income accumulates across the year while equity is measured at a point, standard practice divides by average equity: beginning plus ending equity, divided by two. If a company begins the year with $9 million of equity, ends with $11 million, and earns $2 million, average equity is $10 million and ROE is $2 million divided by $10 million, which equals 20 percent. Using only the ending $11 million would give about 18.2 percent, and using only the beginning balance about 22.2 percent, so the convention exists because the choice moves the answer.
Check the equity figure's composition before trusting the output. Equity is a residual, assets minus liabilities, shaped by decades of profits retained, dividends paid, shares issued and repurchased, and accounting write downs. Two companies with identical operations can post different ROEs purely because one has repurchased many shares and carries a small equity base. That does not make ROE useless. It makes it a ratio whose denominator has a biography, and glancing at that biography, in the balance sheet and equity statement, is part of competent use.
What does the DuPont breakdown reveal?
The classic decomposition multiplies three ratios. Net profit margin, net income divided by revenue, measures pricing and cost control. For a DuPont decomposition consistent with ROE based on average equity, use asset turnover equal to revenue divided by average assets and financial leverage equal to average assets divided by average equity, using the same average-assets figure in both terms. Multiply them and the revenue and asset terms cancel, leaving net income divided by equity: ROE. A worked example: margin of 10 percent, turnover of 1.0, and leverage of 2.0 gives 0.10 times 1.0 times 2.0, which equals 0.20, or 20 percent ROE.
Now the diagnosis the single number hid. A grocer may reach a modest ROE through a 2 percent margin and very high turnover. A software company may reach a high ROE through a 25 percent margin and low leverage. A third company may reach the same ROE as the software firm through thin margins and leverage of 5 or 6, meaning a small fall in asset values wipes out a large share of equity. Same headline, three different businesses, three different risk profiles. DuPont takes thirty seconds and prevents the most common ROE error, which is admiring a levered return as if it were an operating achievement.
When is a high ROE a warning instead of a strength?
When leverage is the engine, when equity has been shrunk to a sliver, or when one year's ratio rests on one time gains. Banks have distinct assets, liabilities, regulation, and capital requirements, so their leverage requires sector-specific analysis, and ROE comparisons belong inside industries, not across the whole market. The extreme case is negative equity, arising after large losses or enormous buybacks: ROE then flips sign with earnings or explodes, and the honest treatment is to discard the ratio for that company and analyze debt and cash flow directly.
Trend matters as much as level. A company holding ROE in a band for a decade, through a downturn, is demonstrating something about its economics. A company whose ROE spikes in one boom year is demonstrating its cyclicality. A single-year ROE ranking can favor temporarily elevated earnings, so compare several periods, as discussed in screening for compounders. Durability is the quality being bought. The level is just its most visible symptom.
How should investors use ROE in practice?
As a quality screen and a conversation starter. In screening, require ROE above a threshold across several years alongside a debt ceiling, so the leverage route to the threshold is partly closed. In individual research, decompose any striking ROE with DuPont before quoting it, and compare it with return on invested capital, which measures returns on all capital, debt and equity together, and is harder to inflate with borrowing. A wide gap between ROE and return on invested capital can reflect leverage, taxes, non-operating items, or differences in capital definitions, so reconcile the numerators and denominators before drawing a conclusion.
ROE also frames management questions well. Is the equity base changing because of retained earnings, new share issuance, distributions, buybacks, losses, or other accounting changes, and what evidence shows whether incremental investment is productive? Different causes imply different investments, even at the same ROE. The ratio cannot answer for you. It can, reliably, tell you that the question is worth asking, and in a market full of single number summaries, a ratio that reliably points at the right question has earned its longevity.
FAQ
What is a good ROE for a stock? Context decides. Assess sustained ROE against documented industry peers, leverage, and the company's own history, and compare across a full cycle rather than against one universal threshold.
Why is ROE high when a company has lots of debt? Because equity is smaller when assets are funded by creditors. Net income divided by a smaller equity base gives a larger percentage. The DuPont leverage term exposes this directly: holding net profit margin and asset turnover constant, a higher assets-to-equity ratio raises DuPont ROE, but additional debt can also reduce net income through interest expense and can lower ROE.
What is the difference between ROE and return on invested capital? ROE measures profit against shareholders' equity only. Return on invested capital measures after tax operating profit against all invested capital, debt plus equity. The second is harder to inflate with leverage and is often preferred for judging the underlying business, while ROE describes the shareholder's specific claim.
Can ROE be too high? Yes. An unusually high ROE warrants checks for leverage, a small equity denominator, unusual gains, and other differences from industry peers. Figures far above industry norms should trigger a DuPont decomposition and a balance sheet review rather than enthusiasm.
Sources
- U.S. Securities and Exchange Commission, ‘Beginners' Guide to Financial Statements’, for introductory financial-statement structure and selected ratios.
Calculation note: Worked examples use hypothetical inputs and are calculated directly in the text.
Educational content, not financial advice. Any worked example uses hypothetical, illustrative figures chosen for round numbers. It is not a projection of any security's returns and not a recommendation to buy or sell any security.
