Profitability & Return Ratios

A margin is a slice of the sales line. You start at revenue, subtract a defined set of costs, and express what is left as a percentage of revenue. Every margin therefore shares the same denominator — sales — and they differ only in how many layers of cost you have removed before you stop counting.

A return ratio has nothing to do with sales. Its denominator is capital: money that was put into the business and left there. It asks a landlord's question rather than a shopkeeper's — not how much of each rupee of trade you keep, but how much the whole building earns for what it cost to buy.

That distinction explains a pattern that confuses almost everyone the first time they meet it. A jeweller or a fuel retailer can run a very thin net margin and still be a sound business, because the same rupee of capital turns over several times in a year and does several rupees of trade. A speciality chemicals plant can run a fat net margin and still earn an unremarkable return, because it took years and an enormous amount of capital to build. Margin is a statement about pricing power and cost control. Return is a statement about capital intensity. You need both readings before you can describe a business at all.

Almost every margin you will see quoted is one of three, taken at three different heights on the same staircase of costs. Read them together and they tell you where a change came from; read one alone and you will attribute it to the wrong cause.

Gross margin moves with input prices, product mix and pricing power — whether a rise in raw material cost can be passed to the customer or has to be absorbed. Operating margin moves with all of that plus operating leverage: fixed costs such as salaries, rent and depreciation sit still while sales rise or fall, so operating margin swings harder in both directions than gross margin does. Net margin moves with all of the above plus two things that have nothing to do with the business operating well — how much the company borrowed, and its tax position.

This is why a falling net margin alongside a steady operating margin is a completely different story from a falling operating margin. The first is usually a financing or tax event. The second is the business itself losing ground.

Take an illustrative mid-cap, Bharat Cables Ltd (figures illustrative, not a real company). Every rupee number on this page belongs to it, so the ratios tie back to the same set of accounts rather than floating free.

Read the right-hand column downwards and you are reading the margin at each height. Nothing here is a judgement about whether 9 per cent is good — that question cannot be answered until you know what the rest of the cable industry earns and what this company earned five years ago.

This is the part most people get wrong, and it is the heart of the topic. All three ratios divide a profit by some pool of money. They differ only in which pool.

Think of a shop. The shop's total value is everything in it: shelves, stock, cash, goods owed by customers. Return on assets (ROA) divides profit by that whole value, however it was paid for. Return on equity (ROE) divides profit by only the owner's own money, meaning the capital put in plus profits kept back over the years. Return on capital employed (ROCE) divides profit by the owner's money plus the bank loans. It leaves out the short bills the shop owes suppliers, because those are not really 'invested' money.

Because ROCE counts the bank's money too, its profit figure must be the one before the bank is paid. That is EBIT, or operating profit. Using net profit here would mix a profit that has already paid the bank with a base that still includes the bank's money. For a company with debt, that quietly understates the ratio.

Run the Bharat Cables numbers through and the three answers spread across a ten-point range — from 11.3 per cent to 21.5 per cent — for one company, in one year, with one set of accounts. That spread is not noise. It is the arithmetic consequence of standing on three different rungs.

Hold Bharat Cables' operations completely still — same plant, same customers, same EBIT of Rs 140 crore — and change only how the Rs 650 crore of capital employed was raised. In Structure A the company funded it with Rs 450 crore of equity and Rs 200 crore of debt. In Structure B, the same business, funded with Rs 250 crore of equity and Rs 400 crore of debt at the same 10 per cent cost.

Nothing operational has changed. Not one extra metre of cable was sold. ROE goes from 20 per cent to 30 per cent, and ROCE does not move at all.

That is the mechanism. ROE has shareholders' funds in the denominator, so replacing equity with debt shrinks the denominator faster than the extra interest shrinks the numerator — as long as the business earns more on the borrowed money than it pays for it. ROCE has both funders in the denominator and a pre-interest numerator, so the funding mix cancels out and only the operating performance shows through.

Return on invested capital (ROIC) tidies up two loose ends in ROCE. First, it takes tax off the profit, so the return is what the business earns after the government's share. Second, it removes idle cash from the base. Cash lying unused in the bank is money waiting for a decision, not money working in the business.

Take Bharat Cables (illustrative). Its operating profit of Rs 140 crore, after 25% tax, becomes Rs 105 crore. Its capital of Rs 650 crore includes Rs 50 crore of idle cash, so the working base is Rs 600 crore. ROIC is 105 divided by 600, which is 17.5%, against a ROCE of 21.5%. The gap is mostly the tax now being counted.

Use ROIC when you want to know how well the business turns money into profit, without the effect of borrowing, tax or a cash pile. It also pairs neatly with the cost of capital in the next block, since both are after tax.

There is no magic number above which a return is good. The fair test is whether the business earns more than its money costs. The cost of capital is what the company pays to raise money: the after-tax interest on its loans, and the return shareholders expect, mixed in proportion.

The idea is the same as your own household. If you borrow at 9% and earn 14% on the money, you gain by borrowing more. If you earn 6%, every extra rupee borrowed makes you poorer. A company works the same way. Earn above the cost of capital, and growth adds value. Earn below it, and growth destroys value even while sales rise.

So a company earning 14% where money costs 11% is creating value. A small, cyclical company earning 18% where money costs 20% is not, despite the bigger headline. The gap matters, not the level. This is also why capital-heavy Indian sectors such as power and infrastructure can show respectable returns for years and still fall short of what their money cost.

The other side of the coin: the cost of capital is an estimate, not a fact. Two careful analysts will pick different numbers, so treat the spread as a rough guide.

The numerator of every return ratio is a flow — profit earned across the whole twelve months. The denominator is a stock — a balance sheet photograph taken on 31 March. Divide a full year of earning by a single day's capital and you have mismatched the two.

It matters most in exactly the situations you care about. A company that raised a large amount of fresh equity in February will show a swollen closing net worth that earned profits for barely six weeks. Divide the full year's profit by that inflated closing figure and ROE collapses for no operating reason at all. The reverse happens after a buyback: the closing base shrinks, and ROE looks like it leapt.

Averaging the opening and closing balances is the standard repair. It is crude — a mid-year fundraise is not smoothed properly by a two-point average — but it removes most of the distortion. The practical rule is simply to know which convention your data source used before comparing its number with one you calculated yourself, because screeners differ and the two answers will not match.

A return ratio is a statement about how much capital a business model requires. Different business models require wildly different amounts, and no amount of arithmetic makes those comparable.

An asset-light software services or consumer brand business rents its offices, owns few machines and carries most of its value in people and distribution, so its capital base is small and returns on it are structurally high. A cement plant, a steel mill or a power project has to build the asset before it earns a rupee, so the same operating skill produces a far lower ratio. Neither is better managed. They are different physical propositions, and comparing their ROCE is like comparing the fuel efficiency of a scooter and a truck.

Lenders break the frame entirely. For a bank or an NBFC, borrowing is not a financing decision sitting outside the business — it is the raw material. Deposits and borrowings are what the business buys in order to sell loans, so capital employed, debt-to-equity and ROCE as defined here simply do not carry their usual meaning. Banks are read through return on assets and return on equity against a regulated capital adequacy floor instead, which is why they get their own topic in this module.

That leaves two comparisons that do work, and only two. Compare a company against direct competitors doing substantially the same thing, and compare it against its own record over five to ten years. Everything else is noise dressed as analysis.

One year of a high return ratio says very little. A commodity producer at the top of its cycle, a company that sold a division, a business enjoying a one-off duty protection — all of them post a spectacular single year and then hand it back.

A return on capital that stays well above the cost of capital for eight or ten years, across at least one full downturn, is a different kind of evidence. In a competitive market, high returns attract entrants; entrants add capacity; capacity compresses margins; returns fall back towards the cost of capital. That is the default gravity of business. When a company defies it for a decade, something is stopping the entrants — a brand customers ask for by name, a distribution network that would take years to replicate, switching costs that make leaving painful, a regulatory licence, or a cost base structurally below everyone else's.

That is precisely what an economic moat means, and it is why the durability of a return ratio matters more than its level in any single year. A steady 18 per cent for a decade is far more informative than a 35 per cent that appeared last year. The moats topic in this module deals with where that durability comes from and how to test whether it is still intact.

Everything above is calculable from two pages of any annual report filed with the exchanges — the standalone or consolidated profit and loss statement, and the balance sheet with its comparative column. Use consolidated figures where the company has subsidiaries, because the standalone accounts of a holding company can show a return that the group as a whole is nowhere near earning.

A return ratio is a tool with a clear use and clear limits. Both halves matter.

When it works: comparing a company with its own past and with direct competitors doing the same work, over five to ten years, using consistent accounts. In that setting a steady, high return on capital is real evidence of a strong business.

When it fails: it says nothing about price. A wonderful business can still be a poor purchase if the market price already assumes everything goes right, and an average business can be a fair one at a low price. It also looks backwards, so it cannot say whether the return will last. And it is easy to distort: a buyback, an asset sale, idle cash or a fresh share issue can move the number without any change in the business. The bull reading is that a high, stable return shows a moat. The bear reading is that a high return invites rivals, and it may be a cycle peak. Test both before you believe either.