Return on equity tells you how much profit a company earns per dollar of shareholder investment, but the same ROE can come from very different businesses — a high-margin software company and a highly leveraged bank can post identical returns for completely different reasons. DuPont analysis breaks ROE apart so you can see which lever is actually being pulled.
Why decompose ROE at all
A single ROE number answers "how much did shareholders earn," but not "why." DuPont analysis, developed at the DuPont corporation in the 1920s, splits ROE into net profit margin (profitability), asset turnover (efficiency), and the equity multiplier (leverage). Two companies with the same 20% ROE might get there completely differently — one from fat margins and modest leverage, another from thin margins offset by heavy borrowing. Decomposing the number reveals which is which, and that distinction matters enormously for how sustainable and risky that return is.
The leverage trap
Leverage is the factor investors should watch most closely. Margin and turnover reflect how the business actually operates; the equity multiplier reflects how it's financed. A company can boost ROE simply by replacing equity with debt — the operating business doesn't improve at all, but the same profit is now divided among fewer equity dollars. That works beautifully until earnings dip: fixed interest payments don't shrink with revenue, so the same leverage that inflated ROE on the way up accelerates losses on the way down. A high ROE built mostly on a large equity multiplier is a different risk profile than one built on margin or efficiency, even if the headline number looks identical.
When the 5-step view earns its keep
The 3-step breakdown is usually enough for a quick read. The 5-step version — which further splits net margin into tax burden, interest burden, and operating margin — is worth the extra inputs when you're comparing companies with very different capital structures or tax situations, or when you're trying to isolate how much of a margin change came from operations versus financing versus taxes. If two competitors report similar operating margins but different ROE, the 5-step view usually shows the gap living in interest burden (debt cost) or tax burden (effective tax rate), not in how well either company actually runs its core business.
Limits of the framework
DuPont analysis explains an existing ROE — it doesn't validate the accounting behind it. Aggressive revenue recognition, one-time gains, or a temporarily depressed equity base (from a large buyback or a prior write-down) can all inflate the factors without reflecting a durable improvement in the business. Always sanity-check inputs against multiple periods and, where possible, against industry peers before drawing conclusions from a single snapshot.