Skip to main content

DuPont Analysis

Updated Jun 24, 2026 at 2:35pm

Research Draft High 1,247 words

DuPont analysis is a diagnostic framework that decomposes return on equity (ROE) into its multiplicative drivers, so an analyst can see why a company earns the ROE it does rather than just what the number is. Its core insight — and core tension — is that two firms can post identical ROEs for completely different reasons: one through fat margins, one through high asset turnover, and one merely through heavy borrowing. By splitting ROE into profitability, efficiency, and leverage, DuPont separates genuine operating quality from financial engineering. The framework is named after the DuPont Corporation, where engineer-turned-treasurer F. Donaldson Brown formalized the return-on-investment chain around 1912–1914; Brown later carried the technique to General Motors in the 1920s, where it became a cornerstone of Alfred Sloan's management system.

How it's calculated / formed

The identity is algebraic — each version is the same ROE with terms inserted that cancel out.

Three-step (the classic form):

> ROE = Net Profit Margin × Asset Turnover × Equity Multiplier > ROE = (Net Income / Revenue) × (Revenue / Avg. Total Assets) × (Avg. Total Assets / Avg. Equity)

Revenue and total assets cancel, collapsing back to Net Income / Equity. The three terms read as: how profitable each sale is, how much revenue the asset base generates, and how much the asset base is amplified by debt.

Five-step (the expanded form) splits net profit margin into three pieces to isolate taxes and interest from operating performance:

> ROE = Tax Burden × Interest Burden × Operating (EBIT) Margin × Asset Turnover × Equity Multiplier > = (Net Income / EBT) × (EBT / EBIT) × (EBIT / Revenue) × (Revenue / Avg. Assets) × (Avg. Assets / Avg. Equity)

Tax burden is the fraction of pre-tax profit kept after tax (higher = lower tax drag); interest burden is the fraction of EBIT surviving interest expense (lower ratio = heavier interest cost); EBIT margin is the pure operating profitability. Conventions vary on whether to use average or period-end balances — averaging balance-sheet items against flow items (revenue, income) is the more defensible practice (Wall Street Prep, CFI).

How it's used in practice

The framework's value is comparative and diagnostic, not as a single score. Analysts use it to:

  • Attribute an ROE change to its source. If ROE rose, was it operating improvement (margin or turnover) or just more leverage? The first is durable; the second adds risk.
  • Benchmark within an industry. A retailer and a luxury-goods maker can share an ROE — the retailer earns it on thin margins and rapid turnover, the luxury house on rich margins and slow turnover. DuPont makes the business model legible.
  • Spot leverage-flattered returns. A rising equity multiplier lifting ROE while margins and turnover stagnate is a warning, not a strength.
  • Trace the five-step chain. A falling interest-burden ratio flags growing debt-service strain; a swing in tax burden may be one-off (a tax credit) rather than structural.

The discipline is to compare each component across time and against peers, then ask which driver is doing the work and whether it is sustainable.

Adoption, debate & evidence

DuPont is one of the most widely taught and applied frameworks in fundamental analysis — standard in CFA curricula, corporate finance courses, and equity research. It is descriptive accounting algebra, so it isn't "contested" in the sense a trading indicator might be; the debate is about which decomposition is most informative and which components carry predictive signal.

The most cited academic evidence comes from Mark Soliman's "The Use of DuPont Analysis by Market Participants" (The Accounting Review, 2008), which decomposes return on net operating assets into profit margin and asset turnover. Soliman found that changes in asset turnover carry information for future earnings that is incremental to other known signals, and that both analysts' forecast revisions and stock returns respond to the components — i.e., the decomposition is useful, but the market under-reacts to some of it. This builds on Fairfield and Yohn (2001), whose result is more nuanced than often quoted: disaggregating the level of ROA into turnover and margin does not improve one-year-ahead forecasts of the change in ROA, but disaggregating the change in ROA into a change in turnover and a change in margin does carry incremental forecasting information. A practical corollary from this strand of accounting research: asset turnover tends to be more persistent, while profit margins mean-revert more strongly (competition erodes outsized margins faster than it erodes structural efficiency) — so a turnover-driven ROE improvement is often a better quality signal than a margin-driven one. Treat the persistence claim as well-supported but data-dependent across samples, not a law.

A separate, important refinement is the Nissim and Penman (2001) "advanced DuPont," which strips leverage out of the operating analysis by building ROE from return on net operating assets plus a financial leverage spread. The argument: mixing operating and financing effects in one margin/turnover/multiplier chain muddies the measure of management's operating skill. This modified form is now standard in serious financial-statement-analysis texts.

Strengths & limitations

Strengths. It turns one opaque ratio into a transparent causal chain; it requires only income-statement and balance-sheet data; and it makes cross-business-model comparison intelligible.

Limitations. (1) It is built on ROE, which is itself leverage-sensitive — the equity multiplier can inflate ROE even as the business deteriorates, and a firm with negative equity (from buybacks or losses) produces a meaningless multiplier. (2) It inherits all accounting distortions in the inputs: one-time items, off-balance-sheet financing, goodwill, and aggressive revenue recognition flow straight through. (3) It is near-useless for financials (banks, insurers, investment banks) where "assets" and "leverage" mean something entirely different — Wikipedia and most texts flag this explicitly. (4) It is historical and ratio-only: it explains the past, says nothing about valuation, and ignores cash flow.

The #1 misuse is treating a high ROE as good without decomposing it — celebrating a number that is really a leverage artifact masking weak operating returns. The whole point of DuPont is to prevent exactly that error.

Sources

Flags: The "asset turnover persists / margins mean-revert" claim is a statistical tendency consistent with this accounting-research strand (Fairfield-Yohn 2001; Soliman 2008; the central-tendency/mean-reversion evidence in Nissim-Penman 2001), not a universal law across all samples/sectors.