Acquisitions From a Research Perspective: Reading the Rationale

Reading an acquisition means asking four things: what was bought, what was paid, how it was funded, and what the price implies about the assumptions the buyer must be holding. Buying earnings at a lower multiple than the buyer's own shares carry lifts per-share figures the day it closes, by arithmetic alone. The multiple gap may be exactly deserved, so that arithmetic settles nothing about whether the purchase was sensible.

Three things are settled elsewhere and none of them gets rebuilt below. The combined statements were taken apart line by line under the merger arithmetic, which worked out how the money was found and what residual sits on the buyer's balance sheet afterwards. The same Rs 480 crore transaction is read here rather than recomputed. The valuation material settled what a multiple is and what it stands for, and a multiple is applied here rather than taught again. An announcement moves a share price while the arithmetic changes the model, which is settled under announcements. The four questions below are all answerable from the announcement, and the reason the announcement was written is not answerable from anything at all. Sarvani Coatings Limited, an invented paint maker, has bought nothing and proposed nothing. Both the purchase worked through below and the target were written for teaching, and the target stays unnamed.

What are the four questions to ask about any acquisition?

Something ordinary makes the point. A cousin mentions at a wedding that she has taken over the tailoring shop two lanes away. Nobody sits there nodding at the word expansion. The questions come straight out: what the shop turns over, what she paid for it, where the money came from, and whether the price sounded like two years of the shop's earnings or ten. Four questions, asked in about eleven seconds, and at the end of them what has happened is roughly clear even to somebody who has never seen the shop.

Research asks exactly those four and in exactly that order, and the order matters because each answer sets up the next. The first, what was bought, fixes the earnings now being bought. The second, what was paid, fixes the price. The third, how it was funded, fixes what left the balance sheet to pay it. The fourth, what the price implies, turns the first two into a statement about what the buyer must believe, and is the only one that requires any thought. On the worked purchase the first three are quick: an unnamed industrial coatings maker with Rs 290 crore of sales and a 21.0 per cent margin, so Rs 60.9 crore of EBITDAEarnings before interest, tax, depreciation and amortisation. A rung of the profit ladder taken before financing costs and before assets are written down, settled in the accounting layer and used here as a given. and net assetsWhat a business is worth on its own books once its liabilities have been taken off its assets. Measured under accounting rules taught elsewhere, and used here only as a starting figure. of Rs 120 crore; Rs 480 crore paid; every rupee of cash on hand plus Rs 168 crore of fresh borrowing.

A rationale offered in words is answered in numbers, and only the numbers are checkable. The announcement will say something about fit, or capability, or a platform for the next decade. Nothing in the announcement can confirm or break any of them, so none of those is a fifth question. Claims about fit and capability are not lies and not usually even wrong. Such claims are simply not the sort of sentence that evidence attaches to, and a researcher who spends the morning arguing with them has spent the morning on the one part of the document that cannot answer back.

What does the price show once it becomes a multiple?

Rs 480 crore on its own is a large and uninformative number. Divided by the Rs 60.9 crore of earnings being bought it becomes 7.9 times, and now the figure can be set beside other figures. Comparison is the whole of what the division achieves. Turning a price into a multiple is a translation and not a verdict. The division moves the price onto the scale every other price already sits on, and a comparison becomes possible. A comparison is still not a conclusion.

The arithmetic will run on the buyer's own rating, so the buyer's own rating is the comparison worth making first. Sarvani Coatings Limited's enterprise valueWhat it would cost to take the whole business, counting the shares and the borrowings together and giving credit for the cash sitting inside. Built in the valuation layer and taken here as given. at the illustrative price of Rs 486/- is Rs 11,664 crore of quoted shares less the Rs 72 crore of cash held over its borrowings, or Rs 11,592 crore, and against year three EBITDA of Rs 446 crore that is 26.0 times. So one business changed hands at 7.9 times while the buyer's shares are rated at 26.0 times. The gap is a factor of more than three, and the gap is the single most consequential number in the whole comparison.

Why do per-share figures rise the moment a cheaper business is bought?

Here is the mechanism, and it is smaller than its reputation. No shares were issued to pay for this purchase, so the number of shares in issue is 24.00 crore before and 24.00 crore after. Set the target's Rs 60.9 crore of earnings on top of the published Rs 446 crore and combined EBITDA reads Rs 506.9 crore. EBITDA per share therefore moves from Rs 446 crore over 24.00 crore shares to Rs 506.9 crore over the same 24.00 crore shares, or from Rs 18.58/- to Rs 21.12/-. The rise is Rs 2.54/- a share, or 13.65 per cent.

The numerator grew and the divisor did not, and there is nothing else happening in that sentence. It is worth saying out loud how ordinary this is, because the word accretion makes it sound like an achievement. A landlord with two rooms let out who buys a third house with a third room in it sees the total rent go up while the headcount stays at one, so rent per person rises. Nobody would call that a triumph of household strategy. The rent per person rose because a bigger numerator sat over the same divisor, and the only reason the equivalent in a listed company sounds impressive is that the divisor happens to be called shares.

The one condition worth stating precisely is the divisor. The rise happens because no new shares were issued. Had the purchase been settled in shares instead, the count would have risen alongside the earnings and the whole direction of the arithmetic would depend on which rose faster. Settlement in shares is a different case and not the one worked here. How a purchase would be structured or settled is covered in the transactions material.

What does the accretion arithmetic not establish?

The arithmetic runs out at exactly this point. A gap of 7.9 times against 26.0 times has two readings, and the arithmetic here cannot separate them.

The first reading is the flattering one. The buyer found something for less than the earnings are worth and picked up the difference. The second is the deflating one, and it is at least as common: a business is worth fewer rupees per rupee of current earnings when those earnings are not going to grow, are not going to last, or need more capital held against them each year to stay where they are. An industrial coatings maker with any of those three properties deserves to change hands at a lower multiple, and the price would then be neither generous nor mean but simply correct.

Both readings produce exactly the same Rs 21.12/- a share on day one. The accretion therefore cannot be used to choose between them. The number is identical under a bargain and under a fair price for a poor asset, and a figure that comes out the same under two opposite explanations carries no information about which explanation applies. The accretion is not weak evidence. It is not evidence.

The price is not in the arithmetic at all, so a third reading can be constructed in which the buyer overpaid for a declining business and the per-share figure still rises. Overpayment is taken up under the balance sheet below. No arithmetic on this transaction chooses between the two readings, and none of it prices the purchase. Whether the purchase was a good one is not settled by any figure above, and every honest treatment of an acquisition stops in the same place.

What does the balance sheet do at exactly the same instant?

The other half of the trap, and the half nobody draws. While EBITDA per share was rising 13.65 per cent, Rs 480 crore left the balance sheet. All Rs 312 crore of cash and investments went first, and Rs 168 crore of new borrowing covered the rest. Borrowings end the day at Rs 408 crore with nothing left standing on the cash line. Net debtBorrowings less the cash and investments a business is holding. A reading below nil means the cash is larger than the borrowings, and the business is said to be in net cash. therefore lands at plus Rs 408 crore where it had read minus Rs 72 crore that morning. An issuer holding more cash than debt turns into a geared one inside a single afternoon.

Per share, that is a move from minus Rs 3.00/- to plus Rs 17.00/-, a step of exactly Rs 20.00/-. Look at where that Rs 20.00/- comes from: it is Rs 480 crore divided by 24.00 crore shares. The price is invisible on the earnings side and appears in full, to the rupee, on the balance sheet side. The two are never shown apart. At the same instant, goodwillThe part of a purchase price left over once everything identifiable has been valued. Goodwill sits on the buyer's balance sheet as an asset. How it is measured and later tested belongs to the accounting material. of Rs 360 crore arises, being the Rs 480 crore paid less the Rs 120 crore of net assets acquired, which is Rs 15.00/- a share.

Now the figure that catches people out. Book value per share does not move at all. Take Rs 1,486 crore of net worth across the same 24.00 crore shares. Nothing was issued and nothing was distributed, so the figure reads Rs 61.92/- on the morning of the purchase and Rs 61.92/- on the evening of it. But tangible book valueNet worth after removing goodwill and other assets that cannot be touched, so what remains is only what the business physically holds or is owed. per share falls to Rs 46.92/-, and the fall is exactly the Rs 15.00/- of goodwill. A reader who checks only the reported book value sees a figure standing perfectly still while the substance underneath it changed shape.

Acquisition-Led Growth: why does the revenue arrive as a step?

The step in revenue has its own mechanism and is not the same subject as the accretion arithmetic. Name the step properly. Acquisition-led growth is revenue growth produced by consolidatingAdding a purchased business into the buyer's own statements line by line, so that one set of accounts covers both. The rules for when and how this happens belong to the accounting material. a business that was bought, rather than by selling more of anything. Combined revenue reads Rs 2,705 crore, being the published Rs 2,415 crore with the target's Rs 290 crore consolidated on top of it, which stands 12.0 per cent above the published year, and not one rupee of that 12.0 per cent came from selling more paint.

Three marks identify it, and all three are visible without any inside knowledge. First, it arrives as a step rather than a slope: revenue jumps at the date the purchase closes and then sits at the new level, where growth produced by selling more arrives as a gradient across quarters. Second, it arrives with a matching balance sheet movement. Something had to be paid, so the same statements that show the revenue step show the cash gone or the borrowings up. Third, it stops repeating as soon as a full year with the purchase inside it is what the comparison runs against, unless another purchase follows. The following year compares a combined business with a combined business, and the 12.0 per cent is simply not there any more.

Two smaller figures move with it, and both look like operating news. Neither is. Blended EBITDA margin reads 18.74 per cent where the published year read 18.47 per cent, and the whole of that difference is a higher margin business having been bolted on, with nothing in the existing paint business moving at all. The industrial share of revenue jumps from 25.0 per cent to 33.0 per cent. In any other setting that would be a mix shift with a story behind it. Here it is a purchase.

How does the whole hypothetical purchase look in one build?

Everything above, set out as a single build so that a reader with a calculator can walk it rather than take it on trust.

Two figures in that table deserve a second look side by side. EBITDA per share is Rs 21.12/-. Net debt per share is Rs 17.00/-. The first of those would be identical if Rs 300 crore had been paid, or Rs 900 crore. The second would be Rs 9.50/- and Rs 34.50/- respectively. One figure ignores the price entirely and the other is nothing but the price, and a reader shown only the first has been shown the half of the transaction that cannot go wrong.

What would have to be true for Rs 480 crore to have been the right price?

The move that gets a researcher past the wall is a change of question rather than a cleverer calculation. The arithmetic cannot answer whether the price was right. Stop asking that, and start asking what the price assumes. A price contains assumptions the way a sentence contains grammar: they are in there whether or not anybody wrote them down, and they can be pulled out and named.

Three of them matter here. Growth: paying 7.9 times only makes sense against a view about whether Rs 60.9 crore of earnings stays at Rs 60.9 crore, and a second year of the target's revenue and margin, once consolidated, would speak to that. Durability: industrial coatings sells to a smaller number of larger buyers than decorative paint does, so the loss of one relationship matters more, and the evidence is whether the combined industrial line holds its level through a full cycle. The capital held against it: a business with high capital intensityHow many rupees of assets and working capital a business must keep tied up simply to hold its earnings where they are. High intensity means more of each rupee earned goes back in before anything is left over. converts less of its earnings into anything the buyer can use, and the evidence is what happens to the combined working capital cycle and capital spend in the year after the purchase.

A researcher can publish that test honestly and cannot publish the verdict. Written as three named conditions with the evidence that would confirm or break each, it is work a reader can use and can later score. Written as a conclusion, it is a claim the arithmetic underneath it does not support, however carefully the arithmetic was done.

What may honestly be published about a rationale?

Five things, and the list is short enough to hold in mind. The four answers, so a reader knows what was bought, what was paid, how it was funded and what the price implies. The multiple, with the buyer's own rating beside it so the comparison is on one scale. The arithmetic accretion, with its cause named in the same sentence. Naming the cause means saying that the numerator grew and the divisor did not, rather than leaving the figure to speak. The balance sheet consequence in the same breath. The price landed there and nowhere else. And the conditions under which the price would be justified, written as conditions.

Stopping there is the finished job rather than an incomplete one, and this is worth saying because it does not feel finished. A note that ends with three named conditions rather than a conclusion reads as though the writer ran out of nerve. The opposite is true. The conclusion is the part that could not be supported, and leaving it out is the only part of the work that required any discipline at all. A reader who wanted a verdict has been given something more useful: the specific future observations that would produce one.

Which figures here were observed, and which were assumed?

Inside the invented record, four figures are observations and everything else was worked from them: the published EBITDA of Rs 446 crore, the published revenue of Rs 2,415 crore, the net worth of Rs 1,486 crore and the borrowings and cash of Rs 240 crore and Rs 312 crore. The target's Rs 290 crore, its 21.0 per cent margin, its Rs 120 crore of net assets and the Rs 480 crore paid are all invented for this exercise. The share count of 24.00 crore is an assumption carried forward from an invented history of a split and a bonus, and it is used forward on every line above and never solved backwards from anything. A per-share table with the assumptions stripped off would read as a set of facts about a company, when it is a set of outputs from one arithmetic exercise on figures somebody wrote. The distinction between an observation and an assumption matters more here than in most treatments.

Every rupee in the worked purchase was written so that it ties. The target's Rs 60.9 crore is 21.0 per cent of Rs 290 crore; the enterprise value of Rs 11,592 crore is 24.00 crore shares at the illustrative Rs 486/- less the Rs 72 crore of net cash; the goodwill of Rs 360 crore is Rs 480 crore less Rs 120 crore of net assets. Four figures above are the record's own rounded ones and are printed as the record has them: Rs 480 crore over Rs 60.9 crore recomputes to 7.8818 times against a printed 7.9; Rs 11,592 crore over Rs 446 crore recomputes to 25.9910 times against a printed 26.0; Rs 290 crore on Rs 2,415 crore recomputes to 12.0083 per cent against a printed 12.0; and the industrial share recomputes to 25.0104 per cent before and 33.0499 per cent after, against a printed 25.0 and 33.0. Every figure here belongs to one instant, the day the purchase closes, set against the published year three, and none of it is a two year figure.