Mergers From a Research Perspective: What Changes in the Model

When two businesses combine, the model is rebuilt rather than adjusted. Revenue, every cost line, earnings before interest, tax, depreciation and amortisation (EBITDA) and the margin all become blended figures. The price has to be funded, so the balance sheet changes hardest: cash leaves, borrowings arrive, and goodwill appears for whatever was paid above what was actually acquired. How the combination is executed is a different subject entirely.

Four things stand behind that answer. Sarvani Coatings Limited, an invented maker of decorative paints and industrial coatings, has a published year three ladder and balance sheet, and every rupee below comes from there. Consolidation and goodwill are accounting outcomes, settled under consolidation accounting and applied here rather than taught again. The event categories and the announcement day order of work are settled under corporate events. An announcement is the catalyst and the arithmetic is the event, and the arithmetic is what follows.

What does a researcher actually rebuild?

The shape is the same at every size, so a household picture makes it plain. Two brothers each run a tiffin service. One does eighty boxes a day at a good price, the other does thirty at a slightly better one. The brothers decide to run one kitchen. The question is what the combined business earns per box. That question cannot be answered by taking the first brother's figure and nudging it. The boxes have to be added, the rupees added, and the division done again. The per box figure that comes out is a new number, and it belongs to neither of them.

A combination in a listed company is that, with more lines. Revenue is added. Each cost line is added. EBITDA is added. A margin is a ratio, and ratios do not add, so the margin is not added. Then the balance sheet: the price has to come from somewhere, so cash falls or borrowings rise or new shares appear, and whatever was paid above the value of what was acquired sits on the asset side as goodwill.

A stale cell looks exactly like a fresh one, so adjusting a live model line by line in place is how errors survive a combination. A combination is one of the very few events where the honest response is to start the sheet again. There is no partial version of this. Revenue cannot be rebuilt while the cash line is left where it was. The two halves of a purchase are joined at the price.

1. Add the two revenue linesAnd say out loud what share of the new total is acquired rather than earned, because that share is not growth and will never repeat.
   Checking: how much of the increase came from a purchase and how much from trading?
2. Add each cost line separatelyMaterials with materials, employees with employees. Adding at the EBITDA level only hides which side of the business each rupee of cost belongs to.
   Checking: does the sum of the rebuilt cost lines still reconcile to the rebuilt EBITDA?
3. Recompute the margin, never carry it forwardThe margin is the last thing touched on the ladder and the first thing a careless reader misreads.
   Checking: is the new margin between the two old ones, and nearer the larger business?
4. Fund the price in the same passCash used, borrowing raised, shares issued. Every rupee of the price lands somewhere on the sheet.
   Checking: do the funding sources add to exactly the price paid, with nothing left over?
5. Redraw the balance sheet, not just the ladderBorrowings, cash, net worth, goodwill and capital employed. A ladder on its own shows only the small half of the transaction.
   Checking: has net debt been recomputed, or quietly carried over?
6. Write down the lines that could not be computedThen leave them blank. A blank cell is information. A filled one that came from nowhere is not.
   Checking: could a reader tell which figures are arithmetic and which are assumption?

What happens to the revenue line, and what happens to the margin?

Sarvani Coatings Limited published revenue of Rs 2,415 crore and EBITDA of Rs 446 crore in year three, a margin of 18.47 per cent. In the hypothetical purchase it acquires an unnamed industrial coatings maker with revenue of Rs 290 crore at a 21.0 per cent EBITDA margin, worth Rs 60.9 crore. The additions are the easy part. Revenue becomes Rs 2,705 crore, an uplift the shared record carries as 12.0 per cent. EBITDA becomes Rs 506.9 crore.

Now the part that trips people. Rs 506.9 crore over Rs 2,705 crore is 18.74 per cent, against the published 18.47 per cent. The margin went up by 0.27 points. And here is the question worth sitting with: which part of Sarvani Coatings got better? None of it. The two revenue lines grew the denominator by about 12.01 per cent while the two EBITDA lines grew the numerator by 13.65 per cent, and a ratio whose top grows faster than its bottom rises. Nothing else happened. The target was simply a higher margin business, and a weighted average pulled towards it.

A blended margin rising is a weighted average moving, not a business improving, and a note that reports the new margin without saying which part of it is arithmetic has misled its reader whether or not it meant to. The tell is easy to check. A blended margin always lands between the two it came from. A rebuilt margin sitting outside the range set by the two businesses signals an error somewhere, and it is usually a cost line added at the wrong level.

There is a second movement here, and it is far larger than the one everybody quotes. Sarvani Coatings published its segment split as decorative Rs 1,811 crore and industrial Rs 604 crore, so industrial was 25.01 per cent of revenue. Add Rs 290 crore of industrial revenue and it becomes Rs 894 crore out of Rs 2,705 crore, or 33.05 per cent. The industrial mix jumped 8.04 points, against 0.27 points of margin, and that is close to thirty times as far. The shape of the business changed much more than the headline number did, and only one of those two moves will be in the headline.

How is the price paid for, and why does that decide the balance sheet?

Here is the part most treatments skip, and it is the part a lender reads first. Rs 480 crore does not appear from nowhere. The considerationWhatever the buyer actually hands over, which may be cash, borrowed money, newly issued shares or a mixture of all three. comes from cash the buyer already holds, from money it borrows, from shares it creates and issues, or from some mixture of the three. There are no other doors. Each door lands on a different part of the balance sheet. The funding choice is therefore not a footnote.

Consider a household buying a scooter for Rs 90,000/-. Paid from the savings account, the household holds one less bank balance and one more scooter. Paid on a loan, the savings account is untouched and there is a new repayment every month. Paid by taking in a cousin as a part holder of the scooter, nobody owes anything and the household simply keeps a smaller share of it. Same scooter, same price, three completely different household sheets afterwards. Nothing about the scooter itself reveals which one happened.

The profit ladder cannot show which door the money came through, and so the balance sheet has to be drawn beside the ladder every single time rather than after it. Revenue of Rs 2,705 crore and EBITDA of Rs 506.9 crore are identical under all three funding routes. A note publishing only those has shown its reader the smaller half of the transaction.

In this hypothetical, the record fixes the route: Rs 480 crore is met by all Rs 312 crore sitting in cash and investments on the year three sheet, plus Rs 168 crore of new borrowing. No shares are issued, so the share count stays at 24.00 crore throughout. The unchanged share count does a great deal of work later on.

What does the funding choice do to the balance sheet?

Sarvani Coatings ended year three with borrowings of Rs 240 crore and cash and investments of Rs 312 crore. Set one against the other and net debtBorrowings after cash and investments are set against them, so it turns negative when a business holds more cash than debt. is MINUS Rs 72 crore. The business is in net cash. Sarvani Coatings holds more money than it owes, and a lender would have called that a comfortable position the day before the announcement.

Now pay Rs 480 crore. Cash goes to nil, so the Rs 312 crore that was setting off the borrowings is gone. Borrowings go from Rs 240 crore to Rs 408 crore, a rise of 70.0 per cent. Net debt is therefore PLUS Rs 408 crore. Net debt read MINUS Rs 72 crore the day before and PLUS Rs 408 crore the day after, a swing of Rs 480 crore. The swing is no coincidence. Every rupee of the price either left the cash line or arrived on the borrowings line, so the swing has to equal the price exactly.

A business sitting in net cash can be a geared one on the day a purchase completes, with not one operating line having moved. That sentence is the reason this block exists. Nothing about how the paint is made, sold or priced changed between the two states. The whole of the movement is funding.

One more line moves with it. Capital employedNet worth plus borrowings, the base a return on capital is measured against. was net worth of Rs 1,486 crore plus borrowings of Rs 240 crore, or Rs 1,726 crore. Afterwards it is Rs 1,486 crore plus Rs 408 crore, or Rs 1,894 crore. The cash that left was never in the capital employed figure to begin with, so capital employed rose by Rs 168 crore, precisely the new borrowing. Remember that Rs 1,894 crore. The figure comes back to bite in the failure below.

Where does goodwill come from, and what is it doing there?

Treat goodwill as arithmetic and it stops being mysterious. The price paid was Rs 480 crore. The net assetsWhat is left of a business once everything it owes is taken off everything it holds. acquired were Rs 120 crore. Take one from the other and Rs 360 crore is left over. The leftover has to sit somewhere on the buyer sheet, and it sits on the asset side under the name goodwill. The Rs 360 crore is 75.0 per cent of everything that was paid.

Read that number the right way round. Goodwill was produced by the price, not by the target. Had Sarvani Coatings paid Rs 300 crore for exactly the same business, goodwill would be Rs 180 crore, and nothing whatever about the target would have been different: same plant, same customers, same order book. Goodwill is a residual and not a discovery, so a large goodwill figure is a statement about what was paid rather than a statement about what was bought.

How goodwill is measured, and how it is tested afterwards for impairmentThe later writing down of an asset whose carrying figure the accounts no longer support., is an accounting question with a standard behind it. The standard is held by the body that issues it, named below. For the model, what matters is simply that Rs 360 crore of the asset side is now a figure produced by a price.

One comparison is worth keeping. No shares were issued and no reserves were touched by the purchase itself, so net worth is unchanged at Rs 1,486 crore. Book value per share therefore holds at Rs 61.92/- on the 24.00 crore shares. But tangible book valueNet worth once goodwill and the other intangibles have been stripped out of it. per share does not hold. Strip the Rs 360 crore of goodwill out and net worth falls to Rs 1,126 crore, or Rs 46.92/- per share. The gap is Rs 15.00/- exactly, being Rs 360 crore over 24.00 crore shares. Two figures that were the same number the day before are Rs 15.00/- apart the day after, and the reported one is the one that did not move.

What does the whole rebuild look like, set down in two columns?

Here is the hypothetical purchase worked end to end, ladder and sheet together, exactly as the first block said it has to be done. Read the two tables side by side rather than one after the other. Every figure below comes from the published record for Sarvani Coatings Limited, and can be re-derived from the rupee amounts in the first column of each table.

Two notes on the tables, and both matter. The published sheet carries no goodwill, so book value and tangible book value started life as the same number, Rs 61.92/-; that is an assumption stated here rather than a figure lifted from the record. And the share count of 24.00 crore is checked forward against the two published per share figures, earnings per share of Rs 11.58/- on profit after tax of Rs 278 crore and book value per share of Rs 61.92/- on net worth of Rs 1,486 crore. The count is never solved backwards out of either of them.

Research Catalyst: why is an announcement the cleanest one there is?

A catalyst is a dated occurrence that makes people revise. Most of them are messy: a slow shift in raw material prices, a competitor gradually changing behaviour, a regulator consulting on something. Their date cannot be pinned down, so neither can the moment at which anybody should have revised. A combination announcement is the opposite of that on every count. An announcement has a date and a time. It carries a price, a target and usually a funding route. Nobody reads one and wonders whether it occurred. And everybody who follows the company reads it in the same minute.

The last property cuts both ways, and is worth pausing on. Everybody revising at the same moment means the reaction is common property. Watching a share price move on the morning of an announcement shows what other people concluded, and shows it quickly. A reaction is also the least informative thing available, because it can only ever be observed after it has happened. The arithmetic is available to anybody willing to sit down and do it, and the reaction is available to nobody. So the useful response to an announcement is to separate the two.

Meghna Iyer, working through the announcement above, has the whole of the determined column inside twenty minutes: Rs 2,705 crore, Rs 506.9 crore, 18.74 per cent, Rs 408 crore of net debt, Rs 360 crore of goodwill and a mix that moved 8.04 points. Nobody handed her those. Each figure was sitting inside the announcement and the last published accounts, waiting for somebody to do the addition. The analyst who spent the same twenty minutes watching a screen has a description of other people and nothing about the combined business at all.

Which lines cannot be rebuilt at all?

The lines that cannot be rebuilt are the ones most combined models quietly skip. The disclosure contains this much. The target has revenue, an EBITDA margin and net assets. The disclosure carries no depreciation charge for the target, no rate on the Rs 168 crore of new borrowing, and nothing on what the Rs 312 crore of cash was earning before it was spent.

The consequence runs down the ladder. Without the target depreciation, combined depreciation cannot be computed, so neither can earnings before interest and tax (EBIT). Without a rate, the combined finance cost cannot be computed. Without EBIT and finance cost there is no profit before tax, so no tax, so no profit after tax, so no earnings per share. And two ratios go with them. The first, interest coverEBIT divided by the finance cost, a ratio that needs both of those lines to be known before it exists., needs both of the missing lines. Return on capital employed needs an EBIT the record does not permit.

The asymmetry in that last one is instructive. The denominator of return on capital is fully determined: capital employed is Rs 1,894 crore and can be computed to the rupee. There is no numerator. Half a ratio is not a ratio. A note quoting an unchanged return on capital after a purchase has assumed, without saying so, that the acquired business holds nothing that depreciates.

There is exactly one honest thing that can be said about that ratio, and it is a bound rather than a figure. Even on the impossible assumption that EBIT does not move at all, Rs 354 crore over the new Rs 1,894 crore is 18.69 per cent against the published 20.51 per cent, so the ratio falls by about 1.82 points. The bound is a direction rather than a forecast: the arithmetic must go that way before the missing depreciation charge is even added, and adding it can only push further the same way.

What should be published before the combined accounts arrive?

Four things, in this order, and they fit on a single sheet. First, the determined lines, with the arithmetic shown: revenue Rs 2,705 crore, EBITDA Rs 506.9 crore, margin 18.74 per cent, industrial mix 33.05 per cent, net debt Rs 408 crore, goodwill Rs 360 crore, capital employed Rs 1,894 crore, tangible book value per share Rs 46.92/-. Second, the undetermined lines, listed by name, with the reason each one is undetermined. Third, the assumptions that would close each gap, written as assumptions: what depreciation charge on what asset base, what rate on what borrowing. Fourth, how much the answer moves when each assumption moves, so a reader can see which gap actually matters and which is noise.

The fourth item is the one people leave out, and it is what turns a list of caveats into something usable. If a plausible range of depreciation charges moves earnings per share by a few paise, the gap is real but small and can be described as such. If it moves it by rupees, the gap is the whole story and no combined earnings figure should be published without it.

Publishing three determined lines and a named list of unknowns is more useful, and far more honest, than publishing a complete looking combined model resting on four invented inputs. A reader cannot tell by looking which cells were computed and which were imagined, so the complete looking version is worse than useless. The imagined cells are indistinguishable from the rest in every downstream ratio.

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