Revenue Growth vs Earnings Growth: Why the Two Differ

Revenue growth and earnings growth differ because every cost line between them grows at its own rate, or does not grow at all. Sarvani Coatings Limited, an invented maker of industrial coatings, grew revenue 13.9 per cent and profit after tax 41.1 per cent between year two and year three. The 27.2 point gap between those two rates is arithmetic rather than achievement. A gap locates where in the ladder the movement happened, never whether it happens again.

Everything the decomposition needs is small and already to hand. The published profit ladder of Sarvani Coatings Limited for all three completed years supplies every rate used below. A margin is what one rung of that ladder leaves as a percentage of revenue, and a cost ratioA cost line expressed as a percentage of revenue for the same period. Two years of different size can then be put side by side. is any cost line written the same way. What is being judged here is a reported base rather than an outcome.

Why do the two rates differ at all?

Start with the thing nobody says out loud. Revenue and profit after tax are not two measures of the same event. Revenue sits at the top of a ladder and profit after tax at the bottom. Between them stand six or seven separate lines, each moving by its own amount for its own reason during the same twelve months.

Consider a tea stall outside an office gate. In a good month it sells a third more cups. The milk and the leaves are bought per cup, so they go up by roughly a third. The rent does not move at all. The boy who runs the counter is paid the same. So the owner keeps far more than a third more money, and anyone who knew only the cup count and the money kept would think something magical had happened. Nothing magical happened. The rent simply did not care how many cups were sold.

The gap between the top rate and the bottom rate is always the sum of what the lines in between did, and never a property of the business by itself. A business does not "have" 41.1 per cent profit growth the way it has a factory. The company reported six line movements that happened to add up that way in one particular year, and each of those six can be pointed at, sized and questioned separately.

Which cost lines open the gap, and in what order?

The order is where the answer lives. Work down the ladder one rung at a time and each rung has exactly one question attached to it: did this line grow faster or slower than the 13.9 per cent revenue grew?

Cost of materials comes first and is the biggest. Materials grew 9.9 per cent against revenue at 13.9 per cent, and the first gap opened there, turning 13.9 into 19.1. Then the costs that do not track revenue closely. Employee cost grew 10.2 per cent and other expenses 13.0 per cent, and together they turned 19.1 into 31.2. Then depreciation and amortisation at 9.5 per cent, a line set years ago by capital spending and almost indifferent to this year's sales, turning 31.2 into 38.3. Then finance cost, down 12.5 per cent as borrowings were repaid, and other income, up 22.6 per cent. The two of them together turned 38.3 into 41.1. Tax came last. Growing 40.9 per cent, almost exactly in line with profit before tax, it changed nothing at all.

A reader who computes only the top rate and the bottom rate has skipped every single place where the answer lives. Look at the table again and notice how ordinary each line is on its own. Not one of them is dramatic. Employee cost grew 10.2 per cent, the ordinary behaviour of a wage bill. Depreciation grew 9.5 per cent, the ordinary behaviour of a depreciation schedule. The drama is entirely in the addition.

One line is honestly strange, and deserves a second look. Cost of materials grew 9.9 per cent while the volume behind it grew 6.0 per cent and revenue grew 13.9 per cent. Materials there is not a cost line behaving lazily. A cost line and a revenue line are moving apart, a different animal altogether. Telling those two apart is the work of the decomposition.

What actually produces the amplification, and what does not?

One condition does all the work here, so go slowly. Amplification needs a cost that does not move with revenue. The condition is that simple. If a cost is genuinely fixed in rupees, then selling more spreads it over more units and every extra rupee of revenue drops a larger share of itself into profit. Rent on the tea stall does not move, so more cups means more money kept per cup.

A cost holding a constant share of revenue produces no amplification whatever. A company whose every cost ratio is flat grows profit at exactly the rate it grows revenue, however large that company is. If materials are 56 per cent of revenue this year and 56 per cent next year, they have contributed nothing at all to the gap. If the wage bill is 8.8 per cent of revenue in both years, it has contributed nothing either. Doubling the size of the business does not change this. Multiply every line in the statement by 1.139 and the profit line is also multiplied by 1.139, and the gap is zero.

A trap hides in plain sight here. Hold the flat ratio case beside what Sarvani Coatings actually reported. Employee cost and other expenses together were Rs 593 crore in year two and Rs 665 crore in year three. As a share of revenue that is 27.97 per cent falling to 27.54 per cent. The fall of 0.43 points is genuine amplification: real costs that grew more slowly than sales. But gross marginRevenue less the cost of materials, expressed as a percentage of revenue for the same period. Gross margin is what each rupee of sales leaves behind before any other cost is met. moved by nearly five times as much, and a gross margin move is not amplification at all. A margin move changes what each rupee of revenue is worth in the first place.

How much of the gap is margin, and how much is simply scale?

Two different things feel identical inside a growth rate and behave completely differently next year. A scale move changes how many rupees of revenue there are. A margin move changes what each rupee leaves behind. Both make profit go up. Only one of them carries a reason to persist.

Here is the household version. A tailor takes twenty more orders this month at the same price and the same cloth cost: that is scale, and if the orders keep coming, the money keeps coming. The same tailor takes the same number of orders but raises the price by ten per cent while cloth costs the same: that is margin, and whether it lasts depends entirely on whether customers keep paying and whether the shop next door follows. On a bank statement the two look identical. In a forecast they are opposites.

The arithmetic that separates them is short enough to write in one line and it holds exactly, not approximately.

Run it on the published figures. Revenue rose Rs 295 crore. Last year's EBITDA margin was 16.04 per cent. Even if nothing else improved, the new revenue alone would have brought Rs 47.3 crore of EBITDA with it, and that is the scale part. EBITDA margin then rose from 16.04 per cent to 18.47 per cent, a move of 2.43 points, and 2.43 points applied to the whole of Rs 2,415 crore is Rs 58.7 crore: that is the margin part. The two add to Rs 106.0 crore, exactly the published movement from Rs 340 crore to Rs 446 crore. The decomposition ties.

Two very different things sit inside the margin part. Open it. Gross margin contributed Rs 48.2 crore of it. Employee cost and other expenses falling as a share of revenue contributed Rs 10.5 crore. Of the Rs 106 crore that EBITDA rose, only about Rs 10.5 crore came from costs failing to keep up with revenue. One rupee in ten is the only part of the movement that the phrase operating leverage actually describes.

A note on why the figures above are 48.2 and 10.5 rather than the 48.3 and 10.4 the rounded headlines would give. Sarvani Coatings reports gross margin as 44.0 per cent in year two and 46.0 per cent in year three, and everybody reads that as a two point move. Computed from the published rupee figures, Rs 933 crore on Rs 2,120 crore and Rs 1,111 crore on Rs 2,415 crore, the two margins are 44.01 and 46.00 per cent and the move is 1.99 points. Build the decomposition on the rounded headline and the parts stop adding back. Catching exactly that failure is what the fourth step of the procedure below is for. When a published absolute exists, it is the figure to use; a figure should never be rebuilt out of a percentage somebody has already rounded.

Why does the amplification run in both directions?

Every mechanism above is symmetric, and symmetry is what gets quietly forgotten in a good year. Costs that fail to rise when revenue rises are the same costs that fail to fall when revenue falls, and the ladder has no memory of which way anyone was hoping it would go.

One division sizes it. Holding every cost below gross profit in the published year three ladder exactly where it is in rupees, namely Rs 665 crore of employee cost and other expenses, Rs 92 crore of depreciation, Rs 21 crore of finance cost and Rs 38 crore of other income, a net Rs 740 crore of charges, profit before tax is gross profit less Rs 740 crore. Gross profit of Rs 1,111 crore divided by profit before tax of Rs 371 crore is 2.99, and that ratio is the multiplier. Revenue moving one per cent with the margin held moves profit before tax about three per cent, up or down, no exceptions.

The same structure that turned 13.9 per cent of revenue growth into 41.1 per cent of profit growth will turn a 5 per cent revenue fall into a fall of about 15 per cent in profit, and a reader admiring the first has already been shown the second. Nothing new has to happen for the bad version. The fall is the identical arithmetic read from right to left.

Push the control down to 44.01 per cent, where gross margin sat in year two, and watch what survives. EBITDA growth falls from 31.2 per cent to 17.0 per cent and profit after tax growth from 41.1 per cent to 22.8 per cent. Push it further, to 42.03 per cent, and profit after tax growth falls to 4.6 per cent while revenue still grows 13.9 per cent: the fan has turned over completely and the bottom of the ladder is now growing slower than the top. The business did not change. One line did.

What if the margin had merely held?

The single most useful number in the whole decomposition comes from a counterfactualA deliberately rebuilt version of a published result in which one input is changed and everything else is left exactly as reported, used to size what that one input was worth.: rebuild year three with gross margin exactly where year two left it and every other published line untouched.

Year two's gross margin, taken from Rs 933 crore on Rs 2,120 crore, is 44.01 per cent. Apply it to year three revenue of Rs 2,415 crore and gross profit becomes Rs 1,062.8 crore instead of Rs 1,111 crore. Take off the same Rs 665 crore of employee cost and other expenses and EBITDA is Rs 397.8 crore, growing 17.0 per cent instead of 31.2. Take off the same Rs 92 crore of depreciation and EBIT is Rs 305.8 crore, growing 19.5 per cent instead of 38.3. Same finance cost, same other income, and profit before tax is Rs 322.8 crore. At the published effective tax rate, profit after tax is Rs 241.9 crore, growing 22.8 per cent instead of 41.1.

Check it ties. Published EBITDA of Rs 446 crore less counterfactual EBITDA of Rs 397.8 crore is Rs 48.2 crore, exactly the gross margin contribution the waterfall showed. Two completely different routes, the same rupee figure. A decomposition that is right looks precisely like that.

Of the 27.2 percentage points by which profit growth exceeded revenue growth, about 18.3 points came from a gross margin move of under two points. Two thirds of the whole gap rests on one line. The remaining 8.9 points is the ordinary part: more revenue meeting costs that grew a little more slowly. The 8.9 points is the part that repeats if revenue repeats. The 18.3 points is the part that needs a reason.

What does a fast earnings line say about next year?

A fast earnings line gives the shape of the ladder, and the shape is a real thing to know. The ladder multiplies by about three, Rs 740 crore of net charges sit below gross profit and barely move, and a small movement in materials cost is worth several times a similar movement in the wage bill.

A fast earnings line says precisely nothing about repeatability until every contributing line has been asked separately whether it can move again in the same direction. Ask that question of each of them. Can revenue grow 13.9 per cent again? Possibly, that is a question about demand and price. Can employee cost grow only 10.2 per cent again? Possibly, that is a question about wage settlements and headcount. Can gross margin rise another two points? The margin question is a different kind of question entirely. A margin cannot rise by two points every year for long without arriving somewhere absurd.

Watch how fast the absurdity arrives, faster than most people expect. Carry 41.1 per cent profit growth forward for three years against 13.9 per cent revenue growth. Profit after tax reaches about Rs 781 crore and revenue about Rs 3,570 crore, so the net margin has gone from 11.5 per cent to 21.9 per cent. Hold every other cost ratio and let depreciation grow as it has been growing, and the gross margin those figures require is about 59.7 per cent, against 46.0 per cent today. No forecaster would defend that margin. A growth rate carried forward quietly assumes it anyway, when nobody decomposes the rate first.

Even the generous version does not get there. Add two more points of gross margin next year, already hard to write down on purpose, hold every cost ratio and let depreciation grow as before, and profit after tax grows 29.2 per cent rather than 41.1. The base is bigger now, so repeating last year in points is no longer enough to repeat it in per cent. A rate that needs an accelerating margin to stand still is not a rate that carries anywhere.

How is the decomposition run?

Four steps, and they run on any two published years with nothing but the statements and a calculator. No data has to be bought and no model has to be built.

A decomposition that does not add back has an arithmetic fault inside it and must not be published, whatever story it happens to support. This is not fussiness. The single most common source of a failed reconciliation is applying a margin move to the wrong base, and the direction of that error is not random: it usually flatters whichever explanation the writer had already settled on. The check is cheap and it is the only thing standing between a decomposition and a rationalisation.

How does an analyst actually use this on a Tuesday morning?

Meghna Iyer, an analyst covering coatings, gets the year three release. She does not start with the growth rates in the headline. She opens a blank column beside the two years, computes the growth rate of every line, and looks for the ones that are not roughly 13.9 per cent. The scan takes about four minutes and it decides what the rest of her week is about.

Her finding sends her to one place. Materials at 9.9 per cent against volume at 6.0 per cent is the only movement large enough to change the answer, so she goes to the segment split and to whatever the company said about pricing, and she leaves the wage bill alone because sizing it would not change anything she concludes. The practical payoff of a decomposition is exactly that: it identifies which question is worth a day of work and which four are not.

The same discipline shows up outside research. A lender is paid the same whether profit grows 41 per cent or 22, so a lender looking at a term loan proposal is not interested in the 41.1 per cent at all. A lender wants the downside arm of that multiplier instead: this ladder falls three per cent for every one per cent revenue falls, so a modest bad year moves interest cover a lot. A person running a household budget already knows the shape of this. When the salary rises ten per cent and the rent does not, what is left over rises far more than ten per cent, and everybody knows without being told that this is a good year rather than a new permanent condition.

Where the quality question enters

Everything above is arithmetic and none of it is contestable. The question underneath the arithmetic is not settled at all, and the published record cannot settle it.

Two thirds of the gap rests on a gross margin move of under two points, and the published statements do not say where that move came from. The record does support a per unit reading. The loose version of that reading is wrong, so state it precisely. Volume between year two and year three was up 6.0 per cent against revenue up 13.9 per cent, leaving realisationWhat one unit of output actually brought in, being revenue for a period divided by the units sold in it, so both discounts given and the spread of products sold land inside the figure. higher by roughly 7.5 per cent. Materials spending was up 9.9 per cent on that same 6.0 per cent of extra output, so each unit made cost about 3.6 per cent more in materials than it did the year before. Nothing on the input side got cheaper, and what widened the gross margin was the price side pulling away from a cost side that was itself still climbing.

Three explanations fit that gap and the published statements separate none of them. A pricing environment across the whole field that let every maker price ahead of its input costs. Sarvani Coatings pricing ahead of the field on its own account. Or a mix shiftA change in the proportions of what was sold. The average price and the average cost per unit move even when no individual product changed price at all. towards higher realisation work. The segment split narrows it: industrial moved from 24.06 per cent of revenue to 25.01 per cent, a shift of about 0.95 points. For a shift that small to produce a two point blended margin move, the two segments would have to differ in gross margin by something like two hundred points, and no coatings business carries a gap of that size. So mix is ruled out as the main explanation and two candidates remain, and the statements cannot choose between them.

A growth rate resting on an unexplained margin move is not an achievement and not a warning: it is a question, and the honest thing to do with it is carry it to the next release rather than resolve it here. The evidence that would separate the two remaining explanations is not in these statements. The evidence would be the peer margins for the same period, putting the question to Nandivarman Paints Limited and Kesaria Surface Solutions Limited about whether they gained too, and whatever the input side of the chain shows at Thottam Chemicals Limited. None of it is settled by the arithmetic above.