What is DuPont analysis?

DuPont analysis breaks return on equity into the pieces that actually drive it, instead of leaving you with one number and no idea why it moved. Two companies can post the same ROE for opposite reasons — one on fat margins, another on heavy debt. DuPont tells them apart. Donaldson Brown built it inside DuPont Corporation in the 1920s, and it's been standard in equity research and corporate finance ever since.

How the decomposition works

ROE on its own is just net income ÷ average equity. DuPont multiplies that out into components that cancel back to the same number — so nothing about the math changes, you just see more of it:

The 3-step version

The one most people mean by "DuPont analysis" — ROE split into three ratios that multiply back to it:

A retailer and a bank can post the same ROE in almost opposite ways: one on high margins over few assets, the other on thin margins over a huge, leveraged balance sheet. The value is seeing which of the three is doing the work.

The 5-step version

Splits net margin into Tax Burden (Net Income ÷ Pretax), Interest Burden (Pretax ÷ EBIT), and Operating Margin (EBIT ÷ Revenue), keeping asset turnover and the equity multiplier. This is the version in Bloomberg and most databases; it isolates financing and tax effects from core operating performance — which matters most for companies with heavy debt, unusual tax situations, or one-off non-operating items, where a single "profit margin" would hide what's going on.

Same ROE, three different stories

Here's why the decomposition earns its keep. Take three businesses that all post the same return on equity — and get there in completely different ways:

Read only the headline number and you'd call them equally good. Decompose them and they're three different businesses with three different risk profiles: the grocer lives or dies on turnover, the software firm on margin, the bank on leverage. A leverage-driven ROE is the fragile one — it rises with debt and falls hardest in a downturn — which is exactly the distinction a single ROE hides and DuPont surfaces.

A worked example: Costco

On Costco's FY2024 Form 10-K the three-step chain runs net margin 2.90% × asset turnover 3.64× × equity multiplier 2.96× ≈ 31.2% ROE — a textbook warehouse-retail profile: razor-thin margin, exceptional asset turnover, moderate leverage. The instructive part is the trend: ROE climbed from roughly 25% to ~31% year over year, but part of that jump came from a special dividend shrinking the equity base, not from better operations. DuPont catches that — the equity multiplier ticks up while the margin barely moves — where a bare ROE would have flattered a financing decision as if it were performance.

Here's a live version, built on MacrosLM, running both decompositions on Costco's reported financials. Toggle between FY2024 and FY2023 and watch the three-step chain and the five-step table update — and notice the "Read" at the bottom, which is where the analysis earns its keep: Costco's ROE rose from ~25% to ~31%, but part of that came from a special dividend shrinking the equity base, not from operations.

Where it's useful, and where it isn't

DuPont is at its best comparing two companies with similar ROE to see why they got there, or tracking one company over several years to see which lever moved. A rising ROE driven by margin improvement is a very different signal from one driven by more debt — and the second deserves more scrutiny, not less. That's why the decomposition rarely travels alone: the same financials support a wider ratio set — profitability, efficiency, liquidity, leverage — that puts each DuPont driver in context (the equity multiplier beside debt/equity, net margin beside gross and operating margin, asset turnover beside inventory days).

Edge cases and common errors

• Use average, not period-end, balances for assets and equity, or a mid-year capital action distorts the ratios.
• A rising ROE from leverage is not a rising ROE from margin. The equity multiplier climbing means more debt, not better operations — read which lever moved before calling it improvement.
• Watch equity-base distortions. Buybacks, special dividends, or accumulated deficits shrink the equity denominator and inflate ROE and the equity multiplier without any operating change (Costco's special dividend is the textbook case).
• Negative equity breaks it. Highly levered or deficit companies can show a meaningless equity multiplier — the decomposition stops being informative.
• It inherits whatever the statements get wrong. Aggressive accounting, one-off items, and off-balance-sheet debt distort the ratios silently — DuPont tells you where ROE came from, not whether the numbers deserve trust. That's what a quality of earnings analysis is for.

Where each input comes from

MacrosLM's analysis agents pull the figures from the filings, run the three- and five-step decomposition, and build the ratio set with every number traced to its line item. Whether a swing is real operating improvement or a balance-sheet artifact stays the analyst's call.

Bottom line

DuPont analysis decomposes ROE into margin, efficiency, and leverage — and the 5-step version splits margin further into tax, interest, and operating effects. It's best for comparing similar-ROE companies or tracking one over time. It tells you where ROE came from; it doesn't tell you whether the underlying numbers deserve to be trusted.

This article is for general information and is not investment or financial advice. Ratio analysis depends on the quality of the underlying statements and should be interpreted by a qualified professional.