Three margins, peeled off one income statement
When you did vertical analysis in the earlier guides, you divided every line of the income statement by revenue, turning dollars into percentages of sales. The [[profit-margin-ratios|profit margins]] are simply the three most-watched of those percentages, given names. Each one stops the income statement at a different floor — after costs, after operations, after everything — and asks: of every dollar of sales, how much was still profit by the time we reached *this* line?
The first floor is gross profit margin — gross profit divided by revenue. It captures the raw spread between what customers pay and what the goods themselves cost: pure pricing power and production efficiency, before a single salary or rent cheque enters the picture. A grocer might live on a 25% gross margin and a software firm on 80%, because their cost structures are utterly different. The second floor is operating profit margin — operating income divided by revenue — which keeps going down past wages, marketing, and rent to show how much profit the *core business* throws off before interest and tax muddy the water.
The bottom floor is net profit margin — net income divided by revenue — the share of each sales dollar that survives absolutely everything, including interest, tax, and one-off items, to reach the owners. Reading the three together tells a story no single number can. If gross margin is healthy but net margin is thin, the problem is overhead or interest, not the product. If gross margin itself is falling, costs are creeping up or prices are eroding at the source. The margins do not just measure profitability; they locate *where* it is being won or lost.
From margins to returns: profit against the money tied up
Margins judge profit against *sales*. But a margin can be high while the business is still a poor investment, because it ignores how much money had to be parked in the firm to generate those sales. A jeweller earning a 30% net margin on a handful of sales a year, using a building and vault worth millions, may be a worse use of capital than a supermarket grinding out a 2% margin on enormous volume. To see this, we stop dividing by revenue and start dividing by the capital itself — and we cross from the income statement onto the balance sheet.
[[return-on-assets|Return on assets]] (ROA) divides net income by total assets. It asks the cleanest possible question: for every dollar of *stuff* the company controls — cash, inventory, factories, receivables — how many cents of profit did it squeeze out? It rewards a firm that does a lot with little and exposes one that hoards expensive assets that barely earn their keep. Because it measures profit against everything the business owns, regardless of who funded it, ROA is the purest gauge of how well management actually *operates* the assets in its care.
[[return-on-equity|Return on equity]] (ROE) narrows the question to the owners. It divides net income by owners' equity — the slice of the assets the shareholders actually funded, after the lenders' claims are stripped out. ROE answers the question an investor cares about above all: of every dollar *I* put in (and left in, as retained earnings), how much profit is the company generating for me? It is the single most-cited measure of how good a business is at multiplying its owners' money — and, as we are about to see, the most easily flattered.
DuPont: pulling ROE apart to see what's really driving it
Here is the trap. A high ROE feels like unambiguous good news, but it can be produced in three very different ways — and two of them are not the same as being a genuinely better business. The famous [[dupont-analysis|DuPont decomposition]], devised at the DuPont chemical company a century ago, cracks ROE open into three factors that multiply together. It is just algebra: ROE is net income over equity, and you can wedge revenue and assets into that fraction without changing its value, splitting one ratio into three.
ROE = Net income = Net income x Revenue x Total assets
---------- ---------- -------- ----------
Equity Revenue Total assets Equity
= Net profit x Asset x Equity
margin turnover multiplier
(profitability) (efficiency) (leverage)
Example: 18% = 6% x 1.5 x 2.0
(a 6% margin, $1.50 of sales per $1 of assets,
and assets twice the size of equity -> 18% ROE)Read left to right, the three factors are profitability, efficiency, and leverage. Net profit margin says how much profit each sale yields — the pricing-and-cost story from the first section. Asset turnover (revenue divided by assets) says how hard the assets are worked — how many dollars of sales each dollar of assets generates. The equity multiplier (assets divided by equity) is the one to watch: it measures leverage, how much of the asset base was funded by debt rather than owners. A firm with twice as many assets as equity has an equity multiplier of 2, and that factor *alone* doubles ROE.
From the company's books to the share price
The ratios so far all live inside the financial statements. The last group reaches out and touches the *market* — what investors are willing to pay for a share. The bridge is [[earnings-per-share|earnings per share]] (EPS), which you met at the close of the income-statement rung: net income divided by the number of shares, turning a company-sized profit into one share's slice. EPS is the hinge on which every market ratio swings, because it is the per-share earnings that a per-share price gets compared against.
The [[price-earnings-ratio|price-to-earnings ratio]] (P/E) is share price divided by EPS. If a share trades at 40 dollars and earned 2 dollars of EPS, its P/E is 20 — investors are paying 20 dollars for each 1 dollar of annual earnings. The cleanest way to read P/E is as the number of years of *today's* earnings it would take to pay back the price, if earnings never grew. A high P/E is not automatically "expensive": it usually means the market expects earnings to *grow*, so it will pay up front for profits it believes are coming. A low P/E can mean a bargain — or a business the market quietly expects to shrink. P/E measures expectation as much as value.
[[dividend-yield|Dividend yield]] comes at the share from the income angle: it is the annual dividend per share divided by the share price, the cash return an owner pockets each year just for holding. A 2-dollar dividend on a 50-dollar share is a 4% yield. But a fat yield is not free money, and beginners are often caught here: a yield can spike simply because the *price* collapsed, and a dividend that looks generous may be one the company cannot actually afford and will soon cut. Yield must always be read against whether the earnings and cash flow behind it are real and durable.
The honest limits of the profitability numbers
Every ratio in this guide rests on net income, and you already know its deepest flaw from the last rung: profit is not cash. A company can post a rising net margin and a glowing ROE while quietly bleeding money, because revenue is booked when earned, not when collected. The principle that profit is not cash means a profitability ratio is only as trustworthy as the earnings underneath it — and earnings can be propped up by aggressive accounting, one-time gains, or sales to customers who will never pay. A profitability story and a cash-flow story should always be read side by side.
There is a quieter trap in the *denominators*. ROA and ROE divide by assets and equity that sit on the balance sheet at historical cost, often far below what they are worth today. A company that bought its land decades ago carries it at a tiny number, which shrinks its asset and equity base and makes ROA and ROE look spectacular — not because it earns more, but because the yardstick is out of date. Book value is not market value, and a return computed against a stale book figure flatters the firm. The same effect lets a company with lots of valuable but unrecorded intangibles — brand, know-how — post returns that look almost too good to be true.
None of this is a reason to distrust profitability ratios — it is a reason to read them like an adult. Used well, they form a chain you can climb: margins locate where profit is won on each sale, ROA and ROE judge profit against the capital employed, DuPont reveals whether a strong return rests on real performance or borrowed leverage, and the market ratios show what investors will pay for it all. Used badly — one number, no comparison, no glance at cash or at the staleness of the denominators — they become a story the figures tell about themselves. The limitations of ratio analysis are not a footnote to this skill; they are half of it.