Your Return on Equity Is Hiding Something
You've built something. Revenue is coming in, the products move, the team is busy. Then you look at what's left at the bottom and it doesn't add up to the effort.
Where is all the money going?
Return on equity is supposed to answer this. Net income divided by owners' equity: for every dollar the owners put in, how much comes back each year. It's the headline number investors quote and the one most likely to be quoted at you.
And on its own, it tells you almost nothing about how a business works.
One number, three drivers
The DuPont framework — developed inside the DuPont company in the 1910s, by an explosives salesman turned finance man named Donaldson Brown — takes ROE apart. The whole thing is one line of algebra:
ROE = Profit margin × Asset turnover × Leverage
Which is to say:
ROE = (Net income ÷ Sales) × (Sales ÷ Assets) × (Assets ÷ Equity)
Cancel the sales and the assets and you're back to net income over equity. The identity is trivial. What it buys you is that each of the three factors is a different question about the business, and they can be fixed independently.
- Profit margin — how much of each dollar of sales survives to the bottom line. A question about pricing and cost.
- Asset turnover — how much sales each dollar of assets generates. A question about efficiency: inventory sitting still, equipment underused, receivables not collected.
- Leverage — how much of those assets was bought with other people's money. A question about financing, and about risk.
Four businesses, four routes to a return
Where a return on equity actually comes from
The DuPont framework splits return on equity into profitability, efficiency and leverage — so you can see which of the three is actually doing the work.
Return on equity
17.7%
margin × turnover × leverage
Return on assets
6.8%
before leverage gets involved
Leverage's contribution
10.9 pts
borrowed, not earned
Per $100 of equity
$18
of net income a year
Profit margin
2.2%
Net income ÷ Sales
How much of each dollar of sales survives to the bottom line?
Asset turnover
3.10×
Sales ÷ Average assets
How hard is every dollar of assets working?
Leverage
2.60×
Average assets ÷ Average equity
How much of those assets was bought with other people's money?
Earned or borrowed? Four businesses compared
- Earned on assets
- Added by leverage
Almost no margin on anything, but the shelves turn over three times a year. Volume, not price, is the whole model.
Two companies can report the same ROE and be nothing alike. Flipping between the presets is the point of the framework: the headline number hides which lever produced it, and leverage in particular flatters the return without improving the business one bit.
Show the numbers
| Business | Profit margin | Asset turnover | Leverage | ROE |
|---|---|---|---|---|
| Supermarket | 2.2% | 3.10× | 2.60× | 17.7% |
| Luxury brand | 21.0% | 0.70× | 1.50× | 22.0% |
| Retail bank | 24.0% | 0.05× | 11.00× | 13.2% |
| Software company | 18.0% | 0.80× | 1.90× | 27.4% |
Flip between the presets and watch the chart at the bottom. The supermarket makes about two cents on the dollar and turns its shelves three times a year — volume, not margin. The luxury brand does the exact opposite: twenty-one cents on the dollar, sold slowly. And the retail bank barely earns anything on its assets at all; almost the entire return is the bottom segment of that bar, which is leverage.
Four businesses. Returns in the same broad range. Nothing else in common.
Why the split is the point
"Our ROE fell from 30% to 20%" is a symptom. The three factors turn it into a diagnosis:
- Margin fell → costs rose or prices slipped. Look at input costs, discounting, mix.
- Turnover fell → assets grew faster than sales. Inventory piling up, a new facility not yet earning, receivables stretching out.
- Leverage fell → you paid down debt. Your ROE went down and your business got safer. This is the one that catches people out.
That last case is the reason the framework is worth the ten minutes. A falling ROE can be good news. A rising one can be a warning.
Running it on your own numbers
Everything you need is on two statements.
- Profit margin = net income ÷ sales. Both from the income statement.
- Asset turnover = sales ÷ average total assets. Sales from the income statement; average assets from the opening and closing balance sheets. Average, not closing — a company that bought a factory in December would otherwise look artificially inefficient.
- Leverage = average total assets ÷ average total equity. Both from the balance sheet, averaged the same way.
Multiply. Check it equals net income ÷ average equity. Then compute the same three for last year, and for the closest competitor who files accounts.
That comparison is where it stops being arithmetic. An ROE of 18% means nothing in isolation. An ROE of 18% built on half the margin and twice the leverage of the firm across the road is a finding.
What to do with each answer
Margin is the problem. Raise prices or cut costs, and be honest about which is available. If you're in a commodity business, a margin below peers is often structural rather than fixable, and the answer is differentiation rather than cost-cutting.
Turnover is the problem. Money is tied up. Inventory, receivables, and idle capacity are the usual three. Look at the cash conversion cycle — days of inventory plus days to collect, minus days you take to pay. Every day you shave is working capital released.
Leverage is the problem. Either you're over-borrowed and one bad quarter from trouble, or you're under-borrowed and leaving cheap capital on the table. This is the factor where "improving" the number can make the business worse, so decide it on risk appetite rather than on ratio.
The framework won't tell you what to do. It will tell you which of three conversations to have, which is most of the value — because the wrong one of those three can absorb a year.