Enterprise Value vs Equity Value: Matching Value to Multiple

Enterprise value pairs with measures taken before the lenders are paid, being earnings before interest, tax, depreciation and amortisation (EBITDA), earnings before interest and tax (EBIT) and sales. Equity value pairs with measures taken after, being earnings and book value. For Sankalp Industrial Systems Limited, invented, the traded enterprise value is 7.78 times EBITDA and the share price is 13.04 times earnings. Dividing enterprise value by earnings gives 16.2, and the division measures nothing.

Small numbers make this mistake easiest to see. Start with a shop on a main road. An uncle buys the shop for Rs 60,00,000: Rs 20,00,000 of his own money and Rs 40,00,000 borrowed at eight per cent a year, so the interest bill is Rs 3,20,000. The rent comes to Rs 6,00,000 a year. How many years of income did the shop cost? There are two honest answers, and one thing that looks like a third.

Whole shop over whole rent, Rs 60,00,000 over Rs 6,00,000, is ten years, and that describes the shop rather than the uncle: it would still be ten had he paid cash. His own Rs 20,00,000 over the Rs 2,80,000 that reaches him once the bank has taken its Rs 3,20,000 is 7.14 years, and that describes his position. Then the division that is not an answer. The whole price of Rs 60,00,000 over that same Rs 2,80,000 comes to 21.4. The top of that fraction includes the bank's Rs 40,00,000 and the bottom of it is what is left after the bank has been paid. Nobody on earth has an interest in 21.4. The same division survives being done on a company at a scale where nothing about the result looks wrong.

The company is Sankalp Industrial Systems Limited, a maker of industrial valves and precision castings, and every figure below is a Year 0 figure from one locked set. Its traded enterprise valueThe value of the operating business, held by lenders and shareholders together. is Rs 22,40,00,00,000 and its equity valueThe shareholders' claim alone, being market capitalisation where the company is listed., being its market capitalisation, is Rs 18,00,00,00,000, or Rs 90.00 a share on 20,00,00,000 shares. Why a company has two values at all, and how the walk between them goes line by line, are covered separately. One rule settles which of the two goes on top of which financial measure, and the wrong choice produces a figure that still looks perfectly ordinary.

Which of the two values should actually be used?

Whichever one belongs to the same people as the number about to go underneath it. Matching the claimants is the entire rule.

Most people meet this subject as a list: enterprise value with EBITDA, equity value with earnings, enterprise value with sales, equity value with book value. The list is correct and it is the worst way to learn the subject. The moment somebody produces a ratio that is not on the list, the list has nothing to say. A list of admissible pairs is a set of consequences pretending to be a rule, and the rule underneath it is one sentence long: the numeratorThe value that sits on the top of a multiple. and the denominatorThe financial measure that sits on the bottom of a multiple. must belong to the same claimants.

An enterprise value is what the operating business is worth, and a business is funded by two groups who both have a claim on what it produces: lenders, entitled to interest and repayment, and shareholders, entitled to whatever is left. An enterprise value is those two claims added together, exactly as Rs 60,00,000 was the value of the whole shop. An equity value is one of the two taken alone, read off a screen as Rs 90.00 times 20,00,00,000 shares. Every measure placed underneath either of them represents a set of people too.

Which measures are taken before the lenders are paid?

Three of them, and the reason shows up in a walk down the profit and loss account. Revenue of Rs 12,00,00,00,000 has nothing whatever deducted from it, so every claim on the business is still to be met out of it. EBITDAEarnings before interest, tax, depreciation and amortisation, taken before the lenders are paid. of Rs 2,88,00,00,000, a flat 24.0 per cent of revenue here, has had the operating costs out but not the lenders and not the taxman. EBITDA is the most used denominator in this subject for that reason: two companies with the same operations produce the same EBITDA whether one is loaded with debt and the other has none. EBITEarnings before interest and tax, also taken before the lenders are paid. of Rs 2,40,00,00,000 is EBITDA less depreciation and amortisation of Rs 48,00,00,000, and depreciation is a charge for machinery being used up rather than a payment to anybody, so EBIT stands before the lenders too.

The test is not whether a measure is large or small or clean or messy, but simply whether the lenders have been paid out of it yet, and on all three the answer is no. The bar below puts the interest line inside EBITDA, at Rs 48,00,00,000 of it.

Which measures are taken after the lenders have been paid?

Two of them, and both are what is left at the bottom of the same ladder. profit attributable to ownersWhat is left for ordinary shareholders after interest, tax and the minority's share. is Rs 1,38,00,00,000 in Year 0, and the bar above shows how it gets there: EBIT of Rs 2,40,00,00,000, less interest of Rs 48,00,00,000, less tax of Rs 48,00,00,000 at the company's own assumed effective rate of 25.0 per cent, less Rs 6,00,00,000 for the quarter of Sankalp Coatings Private Limited, invented, that the group does not hold. Three deductions, three sets of people paid off. An enterprise value has excluded all three, and each deduction is a reason profit attributable to owners cannot sit under one.

earnings per shareProfit attributable to owners divided by the number of shares. is that same figure over 20,00,00,000 shares, being Rs 6.90, so it carries the same claim and sits on the same side. Earnings per share adds one requirement: a per share denominator needs a per share numerator. Rs 90.00 over Rs 6.90 is 13.04 times, the identical multiple that Rs 18,00,00,00,000 over Rs 1,38,00,00,000 gives.

Where does book value sit, and what stands opposite it?

On the equity side by construction, and its opposite number on the enterprise side is a figure most readers have never been introduced to. book valueWhat the balance sheet says is left for shareholders after every liability. of equity is Rs 9,00,00,00,000 here, or Rs 45.00 a share: what the balance sheet says is left for shareholders once every liability has been taken off. The debt has already been deducted in getting to that figure. Book value is therefore a residual in exactly the way profit attributable to owners is. Rs 90.00 over Rs 45.00 is exactly 2.00 times.

The matching measure for the whole business is what the asset base itself is worth before anybody asks who funded it. The replacement cost is locked here at Rs 17,90,00,00,000, and against the Rs 22,40,00,00,000 enterprise value that is 1.25 times. James Tobin is the name on that ratio, from his 1969 paper on a general equilibrium approach to monetary theory. The replacement cost sits within Rs 10,00,00,000 of the market capitalisation of Rs 18,00,00,00,000, and that near coincidence is where a reader gets ambushed. A replacement cost is the cost of an entire asset base and a market capitalisation is what is left after the lenders. Nothing may be drawn from how close the two figures are.

What are the admissible pairs for this one company?

Five of them, and one rule produces all five and rules out everything else without a single act of memory.

The top three take the same numerator because their denominators are all measures nobody has been paid out of yet, and the bottom two take a per share numerator because their denominators are per share figures. A multiple is a question about what somebody is assuming and not an answer about what anything is worth. Calling one of the five high, low, demanding or cautious, or ranking it against another company, answers a question the arithmetic never asked.

What comes out of dividing the enterprise value by the profit?

16.2, and everything dangerous about it is contained in how unremarkable it looks. The exact quotient of Rs 22,40,00,00,000 over Rs 1,38,00,00,000 is 16.2319, printed here as 16.2. A second decimal on a figure that measures nothing would suggest it deserves precision. There is no error message and no impossible value.

The numerator is a claim held by the lenders and the shareholders together, with nothing taken out of it on anybody's behalf. The denominator is what is left after interest of Rs 48,00,00,000, tax of Rs 48,00,00,000 and the minority's Rs 6,00,00,000. So the ratio divides everybody's claim by one group's earnings. The answer is neither an enterprise multiple nor an equity multiple but a fraction belonging to nobody at all.

Compare it with the shop, where the bank sat visibly inside the numerator and was visibly taken out of the denominator in a single paragraph. Here the arithmetic is identical and the visibility is gone: the two figures came from different tabs of a model, and nobody had them side by side long enough to ask who each belonged to.

Is that figure wrong, or is it merely unusual?

Neither, and the two words send a reader in completely different directions. If a figure is wrong, the mistake is the thing to look for, and a correct version sits behind it. If a figure is unusual, the question is what makes this company different, and there is something to learn. A mismatched multipleA ratio whose numerator and denominator belong to different sets of claimants. invites both responses and deserves neither. There is no question that 16.2 is a slightly wrong answer to.

The readings people quietly attach to 16.2 are worth saying out loud. Some read it as a conservative price to earnings figure because it is bigger than 13.04, but a price to earnings figure carries a price on top and this carries an enterprise value. Others read it as a rich enterprise multiple because 16.2 is far above the 7.78, but an enterprise multiple carries an enterprise measure underneath and this carries a shareholders' measure. The figure is a version of neither. An honest description is that a division was performed rather than a ratio measured.

Why does one mismatch survive and the other get caught?

Because of where the resulting number happens to land. Luck decides that, and not care. There are two ways to cross the line. One is the enterprise value over a shareholders' measure, just worked. The other is the market capitalisation over an enterprise measure: Rs 18,00,00,00,000 over EBITDA of Rs 2,88,00,00,000, being 6.25. Both are meaningless by precisely the same argument, and in a working week they behave completely differently.

The six invented peers in this record carry price to earnings figures of 11.0, 12.5, 14.0, 15.0, 17.0 and 28.0 times, against the company's own genuine 13.04. A 16.2 dropped into that column sits between the 15.0 and the 17.0: not the largest, not the smallest, not round, not extreme. Set the 6.25 against the same six peers on enterprise value to EBITDA and it falls below every one of them, so somebody queries it within a minute, having spotted an outlier rather than a mismatch. The dangerous mismatch is not the one that produces the largest error but the one that produces the most believable number.

How is a multiple checked?

With two questions, asked in the same breath, and they work on a ratio nobody has ever taught. Who does the number on the top belong to? Who does the number on the bottom belong to? Same answer twice and the ratio is admissible, whether or not it appears on any list. Different answers and it is not a multiple, however reasonable the figure looks.

The two questions run on a ratio not mentioned above. Enterprise value over free cash flow to the firm: the top belongs to lenders and shareholders together, and so does cash the operating business generates before any payment to a lender, so the ratio is admissible and nobody had to teach the pair. Enterprise value over free cash flow to equity: what is left after interest has been paid belongs to the shareholders alone, so it is not a multiple. A principle survives being handed something new and a list does not. The rule is one sentence rather than five rows for exactly that reason.

Which side is used when two companies are funded differently?

The enterprise side, whenever the businesses are what is being compared, and this is where the matching rule stops being tidiness and starts changing what the work measures. Set Sankalp Industrial Systems Limited beside an unlevered twin, invented for this comparison and identical on every line but one: same operations, same cash of Rs 1,20,00,00,000, same non-operating assets of Rs 1,00,00,00,000, same minority interest of Rs 60,00,00,000, and no debt at all where Sankalp carries Rs 6,00,00,00,000 of it. The operations being identical, the twin has the same enterprise value of Rs 22,40,00,00,000.

The funding never enters the enterprise fraction, so on that side the two are indistinguishable, at 7.78 times each to as many decimals as anyone cares to take it. On the equity side they part company, at 13.79 times for the twin against Sankalp's 13.04. The gap is 0.75 of a turn, 5.75 per cent of the levered company's own figure, between two companies making the identical product at the identical margin from the identical asset base. Ranking those two on price to earnings produces a ranking of loan agreements while appearing to produce a ranking of businesses.

One caution about the direction of that gap. The direction is what a reader carries away as a rule and then gets wrong. No law says an unlevered company shows the larger price to earnings figure. Borrowing here swapped Rs 6,00,00,00,000 of equity value for Rs 36,00,00,000 of earnings, being the interest less relief at 25.0 per cent. The swap is a yield of 6.00 per cent. Because that is below the twin's unlevered earnings yield of 7.25 per cent, being Rs 8.70 over Rs 120.00, what remains yields more and the multiple falls. Reverse the two rates and the gap runs the other way.

Is the equity side ever the right one to use?

Often, and treating it as a lesser option is its own mistake. Neither side is the careful choice and neither is the lazy one; they answer two different questions and the only error is answering the one nobody asked. The enterprise side answers what the operating business is worth, the question behind comparing businesses, valuing a division or thinking about what a buyer of the whole company might pay.

The equity side answers what the shareholders' claim is worth, or how it behaves. Somebody who will hold the shares wants to know what a share is worth to them, and a share is a claim on the residual rather than on the operating business, so the numerator has to be Rs 90.00 and never Rs 22,40,00,00,000.

Reaching for the enterprise side to answer a shareholders' question produces an accurate figure to a question nobody asked. The failure leaves no trace in the arithmetic. Where the borrowing is the business rather than a way of funding it, the distinction stops behaving as it does here, and those cases are covered separately.

Is there a check that catches a mismatch inside a column of figures?

There is one worth carrying and it costs nothing. Enterprise value over EBITDA, times EBITDA over sales, cancels the EBITDA. So multiply any company's enterprise value to EBITDA multiple by its EBITDA margin and out comes its enterprise value to sales multiple, every time. The identity holds across all seven companies in this record.

Three rows look at first glance as though the identity has failed, and Kaimur Industrial Limited is the clearest, at a product of 1.7480 against a printed 1.75. Nothing has failed: the identity holds on the unrounded values and the last column is rounded to two decimals for display. Every identity is computed on the unrounded value, and no figure is ever rebuilt out of another figure's printed form.

The check belongs here because it works only when all three multiples come from the same side of the line. The two enterprise multiples share a numerator and the EBITDA margin is a ratio of two enterprise side figures, so the cancellation is clean. Put a price to earnings figure in the middle and there is nothing to cancel. The numerators represent different people.

What is the habit that prevents all of this?

Saying whose claim each number represents, out loud or on paper, before one is divided by the other. The sentence takes about four seconds and sounds too small to matter. Almost nobody says it for that reason. But nobody sits down intending to build a nonsense ratio: a model has an enterprise value on one sheet and a profit figure on another, somebody wants a quick sanity check, and two numbers get divided because they were both to hand and both had the right number of zeros. The failure is one of proximity, not of understanding. Every person who has ever produced a 16.2 knew that enterprise value includes debt, and the reason they produced it anyway is that nobody says the sentence at the moment the division is typed. So make it a written habit: label the numerator with its claimants in the cell beside it, and keep enterprise multiples and equity multiples in separate blocks rather than one shared column.

When does the distinction stop mattering?

In two situations, and both are narrower than they look. The first is when nothing stands between the two values. The bridge on this company runs enterprise value, plus cash of Rs 1,20,00,00,000 and non-operating assets of Rs 1,00,00,00,000, less debt of Rs 6,00,00,00,000 and less the minority's Rs 60,00,00,000. Those four lines take Rs 22,40,00,00,000 down to Rs 18,00,00,00,000. Empty all four of those lines and the two values are the same number, every measure below EBITDA belongs to the shareholders alone as well, and the ten divisions in the grid above collapse to five. There is nothing left to mismatch.

Notice what that condition is not. The condition is not the absence of debt. The unlevered twin has no borrowing at all, and its two values are still Rs 22,40,00,00,000 and Rs 24,00,00,00,000, Rs 1,60,00,00,000 apart. The twin holds the same cash, the same non-operating assets and the same minority. Being debt free narrows the gap and does not close it, so treating no debt as permission to use either value is the one shortcut that cannot be afforded.

The second is when a change is being measured rather than a level, with the funding held where it is. The gap between the two values here is Rs 4,40,00,00,000, and for as long as the debt, the cash, the non-operating assets and the minority stay put, that gap is a constant. Add Rs 1,00,00,00,000 of operating value and the enterprise value goes to Rs 23,40,00,00,000 while the shareholders' value goes to Rs 19,00,00,00,000, the same rupee on each. So either value answers what an operating improvement is worth to the shareholders. The distinction governs the ratio and never the increment.

Both conditions are things that were true on a balance sheet date and get consumed afterwards, without announcing themselves. A company with an empty bridge draws its first term loan. A company whose funding was fixed buys back shares or sells a plant. Nothing in a column of multiples changes appearance on the day either happens, the 7.78 and the 13.04 still print to two decimals, and the only signal that the condition has expired sits on a balance sheet nobody re-opened.