The Asset Sale: Selling Part Rather Than the Whole

An asset sale hands a part of a business to a buyer for cash and keeps the rest running. The sale improves what the people owed money end up with only when the buyer pays a higher multiple than the remaining business could itself have carried. Below that price it makes them worse off, and the cash arriving disguises it.

The reason sits in one detail that is easy to skip. The earnings of the part sold do not stay behind. The earnings leave with the part sold, so the business left standing supports less borrowing than it did the day before, and the cash has to cover that loss before it covers anything else. The proceeds are not an addition, they are an exchange, and an exchange has a rate at which it breaks even.

Three things are taken as already settled. How much borrowing a given level of earnings can carry is set out under debt capacity, and Meghdoot Coated Products Limited was introduced there. Drawing a line inside a business without finding anybody to buy the part was worked through where separations were compared, and it changes no ratio by itself. And the two ranking rules that decide who receives what out of any pot were both run when insolvency was covered, so the proceeds here are handed into that work rather than derived again.

What exactly is being sold, and what is not?

Start with the difference from a forced sale of the whole business. There, every asset, every contract and every customer goes into one transaction, and whatever the buyer pays becomes the entire amount available to the people owed money. Nothing survives on the seller's side except the claims themselves. In an asset sale a business is still standing the morning after completion, and its own ability to carry debt is now part of the answer rather than something the transaction ended.

The surviving business is what makes the arithmetic harder and more interesting. The question is no longer a single question, what will a buyer pay. There are two, and they interact: what will a buyer pay for the part, and what can the rest of the business carry once the part has gone. An answer that covers only the first has answered half a question and will reach the wrong total.

Meghdoot Coated Products Limited, invented for this subject and carrying borrowings of Rs 900 crore against Rs 60 crore of annual operating earnings, is the case throughout. Fifteen times earnings is the position, and no schedule of payments reaches it. At a judged sustainable level of 3.5 times, the whole business supports Rs 210 crore. The Rs 210 crore was not looked up. Rs 60 crore multiplied by three and a half lands on Rs 210 crore, and 3.0 times and 4.0 times sit beside it at Rs 180 crore and Rs 240 crore because the sustainable level is a judgement somebody makes rather than a property anybody measures.

One thing has to be said before a single figure of the sale is written, and it is said again every time the figure reappears. The record behind this business names no product, no plant, no line and no division. There is nothing in it to sell. So the part being sold here is constructed: of the Rs 60 crore it is given a quarter, Rs 15 crore, and Rs 45 crore stays behind. The quarter is an assumption made here to let the arithmetic be shown, taken as a fraction of a figure that is locked, and no business unit has been invented underneath it. If the real division were a third, or a tenth, every rupee below would change. The test itself would not change, and the crossing point below shows why.

Why do both sides of the ratio move at once?

Here is the move almost nobody runs all the way through. Everything else in a restructuring meeting is a subtraction, so when the room hears that a part of the business can be sold for cash, the cash is welcome. So the cash gets added to what is already there. The proceeds are not added to an unchanged business; they are exchanged for a slice of it, and the slice was carrying debt of its own.

Work it in two steps and it stops being abstract. Before the sale, Rs 60 crore of earnings at three and a half turns supports Rs 210 crore. After the sale, only Rs 45 crore of earnings is left inside the business, and Rs 45 crore at the same three and a half turns supports Rs 157.5 crore. The business has lost Rs 52.5 crore of borrowing capacity, not because anybody wrote anything down, but because the earnings that justified it now belong to somebody else.

So the question is never whether the cash is welcome. The question is whether the cash is bigger than the Rs 52.5 crore of capacity that walked out of the door with the asset. A comparison has a crossing point.

Where is the crossing point, and why is it there?

Take the comparison seriously and it produces a test rather than an opinion. The business gives up the borrowing capacity that the sold earnings were supporting. The business receives, in exchange, whatever the buyer pays for those same earnings. Both quantities are the sold earnings multiplied by something: the capacity given up is the sold earnings at the sustainable multiple, and the cash received is the sold earnings at the buyer's multiple. Set the two side by side and everything cancels except the two multiples. The crossing point therefore sits exactly at the multiple the remaining business could have carried.

The result is stronger than it looks, so run the algebra slowly. Call the sustainable level three and a half turns, the whole year of earnings Rs 60 crore, and the sold slice Rs 15 crore. The amount available after a sale is the earnings that stay at three and a half turns, plus the buyer's multiple on the slice. The amount available if nothing is sold is the whole Rs 60 crore at three and a half turns. Written out the long way, that is the earnings that stay at three and a half turns plus the slice at three and a half turns. The first term is identical in both. So the entire difference between selling and not selling is the slice multiplied by the gap between the buyer's multiple and three and a half.

Two consequences fall straight out of that, and both are worth more than the arithmetic they came from.

The first is the test. If the buyer's multiple is above three and a half, the difference is positive and the sale helps. If it is below, the difference is negative and the sale hurts. If it is exactly three and a half, the difference is zero and the whole exercise, with its diligence, its legal work and its months, has moved not one rupee. There is no band, no zone and no roughly. The crossing point is a single number, and it is the same number the remaining business was being valued at all along.

The second is that the crossing point does not depend on how big the sold part is. The slice sits in that difference as a multiplier, so it scales the size of the gain or the loss but cannot move the point where the gain becomes a loss. Sell a tenth of the earnings, or a quarter, or half, and the answer to the question is the price still three and a half times. The independence is genuinely useful. The quarter used here is a construction, and it is reasonable to ask how much of the conclusion rests on it. None of it does. The quarter decides how much is at stake. The quarter does not decide which way the arithmetic points.

What does the crossing look like in rupees?

A test nobody can run is not a test, so now put rupees on it. The remaining business earns Rs 45 crore and at three and a half turns carries Rs 157.5 crore. The Rs 157.5 crore is recomputed on the earnings that stay, and it is not the Rs 210 crore that the whole business carried. Recomputing on the earnings left behind is the entire discipline. A buyer paying three and a half times the Rs 15 crore the part earns pays Rs 52.5 crore. Add them and the total available is Rs 210.0 crore, exactly what the undisturbed business supported before anybody drafted a sale document.

Nothing lucky happened there. The equality is forced, and it is worth seeing in one line why. Rs 157.5 crore is the kept earnings at three and a half turns; Rs 52.5 crore is the sold earnings at that identical three and a half turns; and the two lots of earnings, Rs 45 crore beside Rs 15 crore, are the whole Rs 60 crore the business started with. Multiplying two parts by the same number and adding them gives the same answer as adding them and multiplying once. Every rupee has been moved from one pocket to another and the pockets belong to the same people.

Notice what has quietly been spent to achieve that. Months of process. A buyer found, courted and taken through diligenceThe examination a buyer runs before committing, over the contracts, the numbers and the liabilities, so that it knows what it is taking on. How that examination runs is set out in the transaction process material on this platform.. Legal work on both sides. A business disrupted while its people wondered which side of a line they would end up on. All of it, at that price, for a total that has not moved by a rupee.

What happens above and below the crossing?

Four prices, all on the same Rs 15 crore of earnings, and each one computed to a total rather than described. The four multiples below are settings chosen to bracket the crossover. Nobody has bid anything, and this record carries no offer of any kind.

The middle column never moves, and that fixed column is the shape of the whole table. Rs 157.5 crore depends on the earnings that stay and not on what anybody pays, so it is fixed the moment the perimeter is fixed. Everything that varies varies in one column, and it varies at a steady rate. A turn is one times the earnings being sold, so each additional turn of the buyer's multiple is worth exactly Rs 15 crore, and Rs 15 crore on Rs 900 crore is 1.7 points of recovery. That is the exchange rate between negotiating harder and getting paid.

Now the row that matters most, and it is the first one. At two turns the buyer pays Rs 30 crore, the total available comes to Rs 187.5 crore, and that is 20.8 per cent of the Rs 900 crore owed against the 23.3 per cent that doing nothing would have produced. Rs 22.5 crore has gone, two and a half points of recovery, and it has gone to the buyer. A sale can reduce what the people owed money receive, and it does so almost silently. Rs 30 crore of cash landing in an account looks like progress in every report that describes it.

The absence of a villain is the part worth carrying away. Nobody stole anything in the 2.0 times row. A buyer offered a price, a seller accepted it, cash moved, and the position of everybody owed money got worse. The only place the damage is visible is in a comparison nobody was obliged to run.

Two readings from the ends of that slider are worth having in words as well, so they survive with the control switched off. Drag it all the way down to one turn and the buyer pays Rs 15 crore, the total falls to Rs 172.5 crore and recovery drops to 19.2 per cent. Drag it to the top at eight turns and the buyer pays Rs 120 crore, the total reaches Rs 277.5 crore and recovery lifts to 30.8 per cent. So across the full width of the control, the gap either side of the do nothing line runs from Rs 37.5 crore short of it to Rs 67.5 crore beyond it, and the arrow underneath the bar prints that distance as a plain magnitude with the words above or below carrying the direction.

Why would a buyer pay above the crossing?

The test says the sale helps above three and a half turns. The test does not say anybody will pay that. The crossover is a threshold the arithmetic produces; a price is something a negotiation produces, and the two have no obligation to agree. Precision matters twice over, then: about the places a higher price could come from, and about the fact that this record contains no bid at all.

The honest answer is that the same asset is worth different amounts in different hands, and there are only a few reasons why. Each is a reason a higher number is possible. None is a reason to expect one.

The last row cuts both ways, so note it carefully. The row explains why a part can fetch a full price when the whole cannot, and it is also the argument a seller will over-argue. The reason a buyer might pay more is a description of the buyer's position, never a forecast of the buyer's behaviour, and never a statement of what price this business would achieve.

What else leaves with the asset?

Everything so far has treated the sold part as a quantity of earnings. It is not. The sold part is a set of customers, people, contracts and equipment, and the earnings are only the arithmetic left behind once that list has been drawn up.

Customers go. The buyer is not paying Rs 52.5 crore for machinery, it is paying for a stream of orders, and those orders are placed by people who now deal with somebody else. People go. A production line without the team that runs it is scrap, and both sides know it. Contracts go where their terms permit, and which terms permit it is settled in the transaction process material on this platform. And the perimeterWhich assets, contracts and people a deal takes, settled by drawing a boundary and putting everything else outside it. Negotiating where that boundary falls is covered in the transaction process material on this platform. around all of it is negotiated, so a great deal of what follows depends on where somebody chose to draw it.

Then there is the item that decides whether every figure above held. A business runs a cost baseThe whole of what a business spends in a year to keep operating, taken together rather than line by line. How it is set out and classified belongs to the accounting material on this platform. in which a good deal is paid once and used by more than one activity. A finance team. A quality function. A single warehouse. A software contract sized for the business as it stands. When the part leaves, the cost of serving it very often does not leave with it, and every rupee of that stranded cost comes out of the Rs 45 crore the arithmetic assumed the remainder would keep.

What does a stranded cost do to the test?

Put a number on it and the point stops being a caveat. Suppose Rs 3 crore of cost that the sold part was using stays in the business. The remainder no longer earns Rs 45 crore, it earns Rs 42 crore, and at three and a half turns Rs 42 crore supports Rs 147 crore rather than Rs 157.5 crore. Ten and a half crore of borrowing capacity has evaporated without anybody selling anything for it.

The bite lands on the price at which the sale breaks even. Follow the stranded cost through to it. The buyer must now cover both the capacity the sold earnings were carrying and the capacity the stranded cost has destroyed. The crossing point moves from 3.5 times to 4.2 times, a shift of 0.7 of a turn, and it moves for a reason the reader can check: three and a half turns applied to Rs 15 crore of sold earnings plus Rs 3 crore of stranded cost, all divided by the Rs 15 crore actually being paid for.

The consequence in practice is uncomfortable. A sale agreed at three and a half times was exactly neutral in the table above. It now delivers Rs 147 crore of supported borrowing plus Rs 52.5 crore of cash, so Rs 199.5 crore, or 22.2 per cent of the Rs 900 crore owed against the 23.3 per cent available from doing nothing. The deal that looked like a draw was a loss, and Rs 3 crore of overhead is a very small number to hide a result like that behind.

Where do the proceeds actually go?

A sale produces an amount. A sale does not produce a recipient. A room can agree an asset sale with total sincerity and be no closer at all to agreeing who gets the money, so the distinction between an amount and a recipient does more work in a real restructuring than almost anything else here.

Take the Rs 210 crore the crossing row produces and hand it into the two ranking illustrations already worked when insolvency was covered. Strict ranking sends every rupee of that pot to the securedMoney owed with named property of the borrower pledged against it. A pledge of that kind, and how the papers create it, sits in the banking material on this platform and is taken as known here. side and leaves the unsecuredMoney owed with no particular property pledged against it. Whoever is owed it looks only to whatever survives once the pledged assets have done their work. side with nothing, so the recovery on that side reads 33.9 per cent once it is measured against the Rs 620 crore that side is separately owed. Pro rataSharing in proportion. Each amount owed draws the same fraction of itself, instead of one group being satisfied in full before the next group starts. instead sends Rs 144.7 crore one way and Rs 65.3 crore the other, and both sides then read 23.3 per cent when measured that same way. Naming the base is not pedantry. A 33.9 and a 23.3 are not two readings of one outcome: the first is struck on Rs 620 crore and the second on Rs 900 crore, and somebody holding unsecured paper who sees only the larger figure will take it for theirs.

The aggregate never moves. The pot is the same pot and a ranking rule only decides which way the rupees go sideways, so Rs 210 crore over Rs 900 crore is 23.3 per cent under either rule. So agreeing an asset sale settles the size of the pot and settles precisely nothing about the ranking, and a plan that presents the sale as though it had settled both has skipped the harder half of the negotiation.

What has to be true before the sale is worth running?

Everything so far assumes an asset sale is available. Usually it is not, and the three conditions below have to hold together rather than one at a time. Two of them are commercial. The third is the scarcest, as the forced sale of a whole business under a deadline shows.

The first condition is a buyer who values the part above what the rest of the business could have carried against it. The condition is the crossover restated as a requirement, and it is not automatic. The second is a perimeter that can actually be separated. Some things come away cleanly; some are so entangled with what stays that cutting them out damages both sides, and a sale that leaves the remainder unable to operate has solved nothing while destroying the thing it was protecting.

The third is time, and it is the condition that fails most often. A sale process needs a buyer to be found, given access, allowed to examine what it is buying, and taken through documents. A business that has stopped meeting its obligations does not control its own calendar, and every one of the first two conditions is worth nothing if the third has already run out.

What does the same idea look like at the size of a street?

Strip the rupees out and the mechanism is a familiar one. A bakery runs two ovens and both are busy. The bakery has borrowed against what the pair of them bring in, the borrowing has become too heavy, and somebody suggests selling one oven.

Ask the only question that matters. How much is the second oven worth to somebody else, compared with what its own output was worth to the bakery? If a buyer pays roughly what that oven was contributing, nothing has been achieved. The bakery now has cash and half the baking, and the cash is worth about what the lost baking was worth. The sale helps only where the buyer can do more with the oven than the bakery could: run it through the night, put it into a kitchen that already has the staff standing there, feed it orders the bakery never had.

And the shared cost is the part everyone forgets. The bakery still pays the same rent on the same premises, still pays the same person to do the books, and now spreads all of it over one oven's output instead of two. The remaining half of the bakery is less profitable than half the bakery was, and that is exactly the stranded cost effect that moved the break even price from three and a half turns to four and a fifth.

Who actually reaches for this, and what do they do with it?

An adviser running a restructuring is not building this arithmetic to admire it. The adviser is building it because two or three parties in the room each want to hear a single number, and the number they want to hear is different in each case. Here is what each of them does with the crossover.

A lender uses it to decide whether to consent to a sale it is being asked to allow. A secured lender holding a hold over the assets of the business is usually being asked to release part of that security so the sale can complete. The lender has one question: is the cash coming in worth more than the security going out. The crossover is that question in a form it can test. Below the crossover, releasing security to permit the sale actively worsens its own position, and the arithmetic says so before any negotiation about who receives the proceeds even begins.

An analyst looking at a plan uses it as a fast check on whether anyone in the room did the work. The tell is simple and it takes ten seconds. Find the supported borrowing figure in the plan, find the earnings it was struck on, and see whether those earnings still include the part being sold. If they do, the plan has double counted, whatever else it says. Double counting is the single most common defect in a plan that carries an asset sale, and it survives review precisely because everything else in the document is arithmetically correct.

The business itself uses it to decide what to put in the room in the first place. If the only part that can be separated cleanly is also the part a buyer values least, then the sale that is easiest to run is the sale most likely to sit below the crossover. The shape is common and painful: the part that can be sold and the part worth selling are not the same part, and running the easy process because it is easy is how a business arrives at the 2.0 times row believing it has done something useful.

The pattern is not a corporate one, and it has a household version. A house has a room let out and the rent covers a third of the loan. Selling the right to that room raises cash today and removes the rent that was servicing the loan. Whether that helps depends entirely on whether the sum received is more than the borrowing the rent was supporting. Letting the room go is the identical test at a scale anybody can hold in their head, and it is just as easy to get wrong when the cash is in hand and the lost rent is a subtraction nobody wrote down.