DuPont analysis splits return on equity into three parts — net profit margin, asset turnover, and the equity multiplier — that multiply back to ROE exactly. The split tells you whether a company's returns come from profitability, from efficient use of assets, or from leverage. Two firms can post the same ROE for opposite reasons, and DuPont analysis is how you find out which reasons.
You need it whenever ROE alone is not enough: ratio-analysis chapters, case questions that ask what is driving a company's returns, and any comparison between two firms whose headline ROE looks similar. The decomposition was developed inside the DuPont Corporation in the 1910s as an internal efficiency tool, which is why it is also called the DuPont identity or the DuPont model.
ROE = Net Profit Margin × Asset Turnover × Equity Multiplier
ROE — Return on equity — net income ÷ shareholders' equity
Net Profit Margin — Net income ÷ revenue — the profit kept from each dollar of sales
Asset Turnover — Revenue ÷ total assets — the revenue generated by each dollar of assets
Equity Multiplier — Total assets ÷ shareholders' equity — how far the asset base extends beyond what owners funded
The identity works because the middle terms cancel: revenue appears in the denominator of margin and the numerator of turnover, and total assets appears in the denominator of turnover and the numerator of the multiplier, so the chain collapses back to net income over equity. A five-step version splits net profit margin further into operating margin, interest burden, and tax burden, but most intro courses stop at three.
Net profit margin is the profitability lever. It rises when a company charges more, controls operating costs, or pays less in interest and tax — everything above and below the operating line ends up in this one ratio. Both inputs come straight off the income statement, so a margin problem is diagnosed there: gross margin, operating expenses, interest, tax.
Asset turnover is the efficiency lever. It measures how hard the balance sheet works: a firm that collects receivables quickly, keeps inventory lean, and runs its capacity full generates more revenue per dollar of assets. Turnover is heavily shaped by industry structure — a grocer might turn its assets 2.4 times a year while a utility turns them 0.3 times — so this lever moves slowly and mostly through operational discipline.
The equity multiplier is the leverage lever. When debt finances a larger share of assets, equity shrinks relative to the asset base and the multiplier climbs; a multiplier of 2.50 means each dollar of equity supports $2.50 of assets. It is the same information the debt-to-equity ratio carries, restated: more debt relative to equity always means a higher multiplier. Borrowing raises it, paying debt down or retaining earnings lowers it — and unlike the first two levers, raising it creates no new profit, it only concentrates existing profit on a thinner equity base.
| Line | NorthPeak Software | Hartley Grocery |
|---|---|---|
| Net income | $6,120 | $3,708 |
| Revenue | $40,800 | $123,600 |
| Total assets | $51,000 | $51,500 |
| Shareholders' equity | $34,000 | $20,600 |
| Net profit margin (net income ÷ revenue) | $6,120 ÷ $40,800 = 15.0% | $3,708 ÷ $123,600 = 3.0% |
| Asset turnover (revenue ÷ assets) | $40,800 ÷ $51,000 = 0.80 | $123,600 ÷ $51,500 = 2.40 |
| Equity multiplier (assets ÷ equity) | $51,000 ÷ $34,000 = 1.50 | $51,500 ÷ $20,600 = 2.50 |
| ROE (multiply the three) | 0.15 × 0.80 × 1.50 = 18.0% | 0.03 × 2.40 × 2.50 = 18.0% |
| Check: net income ÷ equity | $6,120 ÷ $34,000 = 18.0% | $3,708 ÷ $20,600 = 18.0% |
Both companies earn an 18.0% return on equity, and the decomposition shows they earn it in almost opposite ways. NorthPeak keeps 15 cents of every sales dollar but turns its assets only 0.80 times a year, with modest leverage at 1.50. Hartley keeps 3 cents on the dollar, turns its assets 2.40 times, and stretches each equity dollar across $2.50 of assets.
NorthPeak's ROE is the higher-quality 18%. It is earned by the business itself — pricing power and cost control — and it barely depends on debt. If revenue drops 10%, the margin absorbs the hit and equity holders take a proportional loss, nothing worse.
Hartley's 18% leans on the leverage lever. Leverage amplifies returns in both directions: the same multiplier that turns a 7.2% return on assets into an 18% return on equity would turn a small loss into a much larger equity loss. And Hartley has less room for error to begin with, because a 3.0% margin means a cost shock of three cents per sales dollar erases profit entirely. Debt also brings fixed interest payments that do not shrink when revenue does.
The general reading rule: margin-driven and turnover-driven ROE is earned and tends to persist; multiplier-driven ROE is borrowed and carries the risk of the borrowing. When you compare two firms with similar ROE, the one that gets there with less leverage is usually the safer business — and when a company's ROE rises year over year, always check whether the improvement came from the first two levers or just from a bigger multiplier.
The components don't multiply back. If you round each ratio and then multiply, the product can miss net income ÷ equity by a visible amount, and graders check this. Compute each component from the raw figures, multiply, and confirm the product equals the direct ROE calculation before you interpret anything. If it doesn't, one component used the wrong input.
Mixing average and ending balance-sheet figures. Many textbooks compute turnover and the multiplier with average assets and average equity (beginning plus ending, divided by two), while others use ending balances. Either convention works — but only if every component uses the same one. Averaging assets in turnover while using ending equity in the multiplier breaks the identity, and the components will no longer multiply back.
Comparing decompositions across industries. Hartley's 3.0% margin is not evidence of a worse business than NorthPeak's 15.0% — grocery is a thin-margin, high-turnover industry by structure. DuPont components are only meaningful against peers in the same industry or against the same company over time; across industries, the mix is expected to differ.
Two companies can report identical ROE for opposite reasons. DuPont analysis makes the source visible: margin and turnover are earned, while the equity multiplier is borrowed. Verify that your three components multiply back to net income ÷ equity before you interpret anything.
It is a way of taking ROE apart. Instead of one number, you get three: net profit margin (how much profit each sales dollar keeps), asset turnover (how much revenue each asset dollar generates), and the equity multiplier (how much debt stretches the equity base). Multiplied together, they equal ROE exactly, so you can see which lever produces the return.
Take net income and revenue from the income statement, and total assets and shareholders' equity from the balance sheet. Divide net income by revenue for margin, revenue by assets for turnover, and assets by equity for the multiplier. Multiply the three; the product should equal net income divided by equity. If it doesn't, one component used the wrong input.
The 5-step version splits net profit margin into three finer pieces — operating margin, interest burden, and tax burden — while keeping asset turnover and the equity multiplier. It shows whether a margin change came from operations, financing costs, or taxes. Intro courses usually test the 3-step version.
Look at which component does the work. ROE built on margin or turnover reflects the underlying business and tends to persist. ROE built on a high equity multiplier is amplified by debt, which magnifies losses just as it magnifies gains. Compare each component against industry peers, not across industries, since normal margins and turnover differ by sector.
Follow your course's convention — many textbooks average the beginning and ending balance-sheet figures because income accrues over the whole year. What matters more is consistency: use the same convention for assets in turnover and for assets and equity in the multiplier, or the components will not multiply back to ROE.
The framework was developed inside the DuPont Corporation, starting from a 1914 internal efficiency report, and the company used it through the 1920s to evaluate its divisions. The name stuck, which is why it is also called the DuPont identity, the DuPont model, or the DuPont system.
By the FinanceBrain Team · Last verified July 11, 2026 · How we produce and verify articles