The Multiple: One Number Holding a Whole Assumption Set

A multiple is a compressed assumption set, not a valuation method. Sarvani Coatings Limited at Rs 486/- against earnings per share of Rs 11.58/- is 42.0 times, and that single number still carries a return somebody demanded, a length of time to wait, a path earnings were assumed to take, and a rating applied on the day they sell. Decompressing it is the work. Comparing it with another number without decompressing it is not.

Two things sit underneath that answer. The first is how a ratio of a market figure to a business figure is constructed at all, a construction that belongs to the valuation method material and is applied here rather than taught. The second is the backward arithmetic that runs from an observed price to the assumptions inside it, set out earlier in this sequence and used here as a tool. New in this guide is the compression itself: what goes into a ratio, what is lost the moment it is written down, and how far a ratio can move while the business it describes stands perfectly still.

Sarvani Coatings Limited, a constructed paint manufacturer, carries the figures below as at 28 August 2026. Its share price of Rs 486/-, its share count, its balance sheet and the three ratios built from them are constructed for teaching. The required return and the exit rating used in the decompression block are the reader's own inputs, chosen and named by the reader rather than estimated by anybody.

What is a multiple actually holding?

Something closer to home makes the point. A landlord says a shop rents for twenty times its monthly takings. The twenty was not handed down by anybody. Somebody looked at the street, the footfall, the length of the lease, what the last three tenants managed and what a buyer would settle for, and out of all of that came a price. Then the price was divided by the takings and the ratio appeared. Every one of those judgements is still sitting inside the twenty. Not one of them is legible in it.

A multiple on a listed company works exactly the same way and fails in exactly the same way. Somebody, or rather a very large number of somebodies trading against each other, arrived at Rs 486/- for a share of Sarvani Coatings Limited. The price already contains a view about how fast earnings grow, for how long, how risky the path is, and what the shares will be worth to somebody else at the end of it. Rs 486/- divided by earnings per share of Rs 11.58/- gives 42.0 times. The ratio was not the input to that process, it was the output. A multiple is therefore never the beginning of an analysis, and it is very often mistaken for one.

The reversal matters because of what people do with ratios. A number that comes out of a long argument looks identical, in print, to a number that started one. Nothing about 42.0 announces that it was solved for. The figure sits in a table next to a company name looking like a measurement, the way a temperature is a measurement, and it is not one. A ratio is the residue of an argument nobody can see.

Which multiples does this record carry, and what is each one built on?

Three ratios appear in the market data for Sarvani Coatings Limited, and the useful exercise is to rebuild each of them rather than accept them. Rebuilding is not a formality. Rebuilding is the only way to see that the three do not share a single component between them.

The first is price to earnings. The illustrative price of Rs 486/- divided by year three earnings per share of Rs 11.58/- gives 41.9689 times, printed as 42.0. There is a subtlety worth catching here. Checking a rounded figure against its unrounded source is the habit worth building. The published Rs 11.58/- is itself a rounded number: profit after tax of Rs 278 crore over 24.00 crore shares is Rs 11.5833/- exactly, and on that unrounded figure the ratio is 41.9568 times. Both round to 42.0, so the printed ratio is unaffected, and the check is what makes that certain rather than assumed.

The second is enterprise valueThe value of the whole business rather than of the shares alone: the market value of the equity plus borrowings, less cash and investments. to EBITDAEarnings before interest, tax, depreciation and amortisation. A profit figure taken before the cost of borrowing, before tax and before the writing down of assets.. Market value is 24.00 crore shares at Rs 486/-, or Rs 11,664 crore. Add borrowings of Rs 240 crore and take away cash and investments of Rs 312 crore, and enterprise value is Rs 11,592 crore. Over year three EBITDA of Rs 446 crore that is 25.99 times, printed as 26.0.

The third is price to book value per shareNet worth spread across every share on the register. What the accounts say each share stands for, as against what the market says it is worth.. Rs 486/- over Rs 61.92/- is 7.85 times. And now the tally stands as follows. Three ratios, three different numerators, three different denominators, and not one component shared across any pair of them. Nothing is shared, so there is nothing to average, and a single blended view is not available.

The last row is not a joke at anybody's expense. The average is arithmetic that runs perfectly well and produces a figure referring to nothing in the world. Nobody would write it deliberately. People do the equivalent constantly, taking a mean of ratios across a set of companies whose denominators were built on different rules. A mean of a peer set is the same error wearing a better suit.

Why does enterprise value sit below market value here?

Most companies anybody has looked at carry debt, so most readers carry a quiet assumption that enterprise value is the bigger of the two. Sarvani Coatings does not, and the ordering flips. Borrowings are Rs 240 crore, cash and investments are Rs 312 crore, so net cashThe position where cash and investments exceed borrowings, so the figure conventionally called net debt is a negative number. of Rs 72 crore sits on the balance sheet at the end of year three. Net debt is therefore MINUS Rs 72 crore, and enterprise value of Rs 11,592 crore falls below market value of Rs 11,664 crore by exactly that amount.

A corner shop bought for Rs 20 lakh, with Rs 3 lakh sitting in the till on handover, shows what this means. Rs 3 lakh of what was handed over came straight back, so the buyer paid Rs 20 lakh for the shares and Rs 17 lakh for the shop itself. Enterprise value is exactly that: what the operating business cost, once the cash inherited and the borrowings taken on are settled up.

The consequence is the single most useful thing in this guide: because one ratio is built on the price of the shares and the other on the price of the business, a change in the capital structureThe mix of borrowings and shareholders' funds a company runs on. Changing that mix changes the enterprise figure without touching the operating result. pushes the two in opposite directions, so a reader watching only one of them will see a movement they cannot explain. Hold that thought, because in a few blocks it is going to do some work.

How is a multiple decompressed into the assumptions inside it?

The assumptions that produced a price were never written down, and there were thousands of them, so not one can be recovered. The one move available is to choose two assumptions of one's own, hold them still, and solve for the third. The move is called decompression, and the procedure was built earlier in this sequence and is applied here rather than derived again.

Four steps. First, a required return: what the analyst wants for taking this risk. Second, a horizon: how long the analyst is prepared to wait. Third, a rating at the end, the ratio a buyer is expected to apply on the day of sale. Fourth, solve for the earnings path that makes those three hold together at today's price. The output is an earnings path, not a judgement, and the two assumptions that drive it belong to the analyst by name and are stated as the analyst's own.

Run it on Sarvani Coatings. At a required return of 12 per cent a year, Rs 486/- has to become Rs 856.50/- in five years. At a rating of 25 times, a price of Rs 856.50/- needs earnings of Rs 34.26/- a share. Growing Rs 11.58/- into Rs 34.26/- over five years takes about 24.2 per cent compound growthA growth rate applied to a figure repeatedly, so each year builds on the year before. Roughly 24 per cent a year triples a number in five years. a year. So 42.0 times, under the two stated assumptions, is holding an expectation that earnings compound at roughly 24 per cent a year and keep doing it for five.

Against what the earlier work established, that is something to think about rather than something to conclude. The sector these figures describe grew 11.0 per cent in the year. Sarvani Coatings grew revenue 13.9 per cent and picked up 0.13 of a percentage point of share, a real gain and a small one. And the margin gain that carried profit after tax up 41.1 per cent has not been shown to be durable. Naming that tension is as far as the arithmetic reaches; resolving it is a judgement about the business.

Can a multiple move when nothing about the business has?

The buyback that follows is HYPOTHETICAL from beginning to end. Sarvani Coatings Limited has not done it, has not proposed it, and nothing in the record suggests it intends to. Working it through has one purpose, which is to make the arithmetic visible.

Suppose Rs 240 crore is returned to shareholders through a buybackA company purchasing its own shares out of its cash and cancelling them, which reduces the number of shares in issue. at Rs 600/- a share, a 23.5 per cent premium to the illustrative price of Rs 486/-. Rs 240 crore at Rs 600/- buys 0.40 crore shares, so the count falls from 24.00 crore to 23.60 crore. Nothing about how the paint is made or sold has altered, so profit after tax is unchanged at Rs 278 crore and earnings per share rises from Rs 11.58/- to Rs 11.78/-. At the same unchanged Rs 486/-, price to earnings falls from 42.0 times to about 41.3 times.

Now follow the cash. Cash and investments fall from Rs 312 crore to Rs 72 crore, so net debt moves from MINUS Rs 72 crore to PLUS Rs 168 crore. Market value becomes 23.60 crore shares at Rs 486/-, or Rs 11,470 crore, and enterprise value becomes Rs 11,638 crore. Over the same unchanged EBITDA of Rs 446 crore, enterprise value to EBITDA rises from 26.0 times to about 26.1 times.

One ratio went down by about seven tenths of a turn, the other went up by about a tenth, the price never moved by a paisa, and revenue, margin and cash generation did not change at all. A third ratio moves hardest of the three and in the same direction as the enterprise one: net worth falls from Rs 1,486 crore to Rs 1,246 crore, so book value per share falls from Rs 61.92/- to Rs 52.80/-, and price to book rises from 7.85 times to 9.21 times. Earnings per share rising while book value per share falls, on one action, is the cleanest statement of what these ratios are actually sensitive to.

Worth naming plainly: the Rs 240 crore of this hypothetical buyback happens to be the same figure as total borrowings of Rs 240 crore. The match is a coincidence in the record, not a mechanism. The buyback is funded from cash, the borrowings are untouched throughout, and nothing in the arithmetic connects the two numbers.

One setting is worth pausing on: Rs 72 crore. At that setting the cash spent exactly equals the net cash the company started with, so net debt lands on nil and enterprise value equals market value for the only time on the whole slider. The crossing point is not a target and means nothing about the business. Nil net debt is simply where the arithmetic passes through zero, and it is a useful place to see that the enterprise ratio has no floor holding it below the market one.

The reader test that follows carries the whole weight of this block. One fact is handed over and nothing else: price to earnings fell from 42.0 to 41.3 times. Three things could have caused it. A fall in the price is one. A rise in earnings from the business actually earning more is a second. And a fall in the share count from a capital action, with earnings flat and the price flat, is a third. Only the middle one is about the company. The ratio does not disclose which happened.

What can a multiple not be compared across?

Four tests follow, to be run before a comparison rather than nodded at as warnings. Each one, if it fails, means the two ratios are not measuring the same thing, and each one is completely invisible in the ratios themselves.

First, a different earnings base. One denominator may be a reported figure and the other a management adjusted one. The earnings quality work on this very company showed reported EBITDA of Rs 446 crore against a management adjusted Rs 452 crore and a symmetric Rs 448 crore, a spread of Rs 6 crore on one company in one year. Two ratios built on the top and bottom of that range differ by roughly a third of a turn before anything real has been compared.

Second, a different share count. If one company has had a bonus, a split, a rights issue or a buyback and the other has not, the per share denominators are on different footings unless both have been restated. Third, a different capital structure. One company in net cash and another in net debt will have their enterprise ratios and their earnings ratios pointing in opposite directions relative to each other, exactly as the block above showed. Fourth, a different accounting choice inside the denominator: two issuers can capitalise differently, provide differently, or treat a lease differently, and both can be entirely correct.

All four are common, and not one of them can be detected by looking at the ratio. A comparison between two multiples therefore feels safest at precisely the moment it stops being safe.

When is a multiple genuinely the right shorthand?

The honest positive case is this. A ratio is genuinely good at one job. Given thirty companies, all reporting on the same basis, none of which has had a corporate action, all with broadly similar capital structures, and a need to know within thirty seconds which three are worth an afternoon of work, a multiple does that job better than anything else available. A ratio is fast, a ratio is arithmetic, and speed is exactly what was needed.

A multiple compares to price per kilogram at a vegetable market. Price per kilogram sorts twenty sellers in a glance. Price per kilogram is completely useless for settling whose tomatoes will still be good on Thursday, and nobody would ever argue about that using the price per kilogram. A buyer would pick one up. A multiple is a good index and a poor argument, and knowing which of the two is in hand at any moment is what decides whether it misleads.

Who actually does this, and what they do with it

A lender sizing a facility is looking at the whole business rather than at the shares, so a lender rarely uses an earnings ratio at all and reaches straight for the enterprise figure and the borrowings inside it. On this record that lender would notice within a minute that net debt is negative, and would treat the Rs 72 crore of net cash as a fact to check the durability of rather than as comfort. If it walks out through a buyback, the lender's covenant headroom changes even though the profit and loss account did not.

A research analyst uses a ratio in exactly two places. First, to sort a screen of names down to a handful. Second, at the very end, to sanity check a conclusion reached some other way: if the discounted work implies a rating far outside where the whole set trades, the analyst goes back and finds out which assumption is doing that. The ratio never generates the view.

A long term investor uses decompression rather than comparison. Asking whether 42.0 times is high gets no answer from the ratio. So they run the four steps, write down that the price contains about 24 per cent growth for five years under their own assumptions, and then spend their time on the only question that is actually answerable: is there evidence for or against that path. And a household choosing between two of anything does the same thing without the vocabulary, when it stops asking which is cheaper and starts asking what each price is assuming will happen.

What does a multiple never tell?

A multiple never tells whether a price is high or low. The silence is not modesty and it is not a dodge. High and low are statements about assumptions: a price is high if the assumptions inside it will not be met, and low if they will be exceeded. The ratio is the one object in the whole exercise that has had those assumptions removed from it, so asking a ratio whether a price is high is asking the only part of the record that has been stripped of the answer.

The ratio is where the assumptions went to be forgotten, so the work is always to put them back, and the number itself is never the destination. That is why three multiples can be rebuilt in full, two of them moved in opposite directions by a hypothetical action, and the whole exercise still end without a single word about whether Sarvani Coatings Limited is worth Rs 486/-. No conclusion is being withheld. There is no conclusion in a ratio to withhold.

Where would every figure above be found for a real issuer?

Every rupee above comes from the invented case record. The routes below are where the equivalent items would be found for a real issuer.

Sarvani Coatings Limited, Nandivarman Paints Limited, Kesaria Surface Solutions Limited, Thottam Chemicals Limited and the analyst Meghna Iyer are invented.
Educational material. Not advice on any investment, tax, budget or market position.