Capital Allocation: The Recurring Decision That Compounds
Capital Allocation: The Recurring Decision That Compounds
Capital allocation is the choice of where a company's next rupee goes: back into the existing asset base, into new capacity, into buying another business, into paying down borrowings, or out to shareholders. Each route is priced the same way, as the return it earns on the capital it swallows, set against what that capital already earns inside the business. The comparison is arithmetic. The choice is not.
What decision is actually being made here?
Think about a household that finishes the month with Rs 8,000/- left over. The Rs 8,000/- is going somewhere. The money can top up the fixed deposit, it can go into the roof that needs redoing, it can pay down the loan on the scooter, it can be handed to the son starting a shop, or it can simply sit in the savings account because nobody sat down to decide. All five of those are decisions. The last one is a decision too, and it is the one people do not notice themselves making.
A company is in exactly the same position, at a much larger scale and with the same short list. Harivansh Packaging Limited, a listed maker of rigid and flexible packaging for food and personal care customers, finishes its year with revenue of Rs 3,180 crore, earnings before interest, tax, depreciation and amortisation (EBITDA) of Rs 477 crore and profit after tax of Rs 225 crore. Devyani Kulkarni, its chief financial officer, does not get to leave that money undirected. The money goes somewhere by the end of the next year whether or not a meeting is convened about it.
A company that never decides has still allocated, and the cash sitting on its balance sheet at the year end is the written record of that decision. Harivansh Packaging holds Rs 140 crore of cash. Nobody has to have chosen to hold it. The Rs 140 crore is what is left when nothing else was chosen, and it sits on the balance sheet in exactly the same ink as a deliberate reserve would be. The balance sheet cannot tell the two apart and neither can a reader.
Capital allocation is therefore a recurring decision rather than a transaction. An acquisition happens once. A rights issue happens once. Allocation happens every single year for as long as the business generates more money than it consumes, and it happens again next year on whatever is generated then. Compounding is exactly that repetition. Each year's choice is made from the position the last year's choice left behind, so a business that puts money to work at a good return year after year ends up somewhere very different from one that does not.
The record below is one year of one invented company. There is no second year in it, no third, and no history of what was chosen before. Compounding needs a run of years, and a single year cannot show one business pulling away from another. The record supports the single decision, priced properly, for one year. The repetition is real, and a compounding chart drawn from one year would have to invent the years it needed.
So the decision under the words is narrow and concrete. Not what the strategy should be, not what the market wants, not what the board would like to announce. Just this: there is an amount of money, there is a short list of places it can go, and each place has a return attached to it. Everything difficult about the subject comes from the fact that those returns are not written down in a form that allows them to be set side by side.
Where can a rupee actually go?
Here is the part that surprises people who expect this to be complicated. The list is short. Money leaving a company by any route at all leaves by one of six doors, and there is no seventh.
A rupee can go into maintaining what already exists: the machinery, the buildings and the vehicles that wear out and have to be kept standing. A rupee can go into growth inside the business, meaning new capacity or a new line that did not exist before. A rupee can go into buying another business outright. A rupee can go into repaying borrowings. A rupee can go back to shareholders, either as a dividend or by buying in the company's own shares. Or a rupee can stay as cash, the door that opens by default when none of the other five is pushed.
A short, closed list is what makes the whole comparison possible: six candidates can be set out together, priced, and ranked, in a way that six hundred could not. Compare it with the household. The rupees left at the end of the month go to the roof, the new stall, the loan, the fixed deposit, the relative who expects a share, or the savings account. Six doors, at a scale a person can hold in their head, and the same six that a Rs 3,180 crore packaging company is looking at.
Routes 1 and 2 are real and this record carries no figure for either. There is no capital expenditure line anywhere in this constructed case, so how much Harivansh Packaging Limited spends keeping its machines standing, and how much it spends adding capacity, is simply not known here. Maintenance and growth spending are therefore named but never sized. Route 5 has the same problem on one of its two legs: the record carries a repurchase and carries no dividend, so the return-of-capital route is worked through the repurchase alone.
Share Repurchase: cash goes out, shares come in, the earnings stay put
Of the six routes, one deserves its own treatment before the comparison starts. Its mechanics are the ones most often confused with a result. A repurchase is a company buying its own shares in the market and cancelling them.
Three things happen and they happen in a fixed order. Cash leaves the company. Shares are bought from whoever is willing to sell into the free floatThe portion of a company's shares actually available to trade, as opposed to the portion held by the people who control it. Harivansh Packaging's promoter and promoter group hold 58.0 per cent, so the float is 42.0 per cent.. The bought shares are extinguished, so the count of shares outstanding falls. And that is the entire mechanism.
Work it on the case. Harivansh Packaging Limited spends its Rs 140 crore of cash buying shares at the illustrative price of Rs 300/-, recorded as at 28 August 2026. Rs 140 crore divided by Rs 300/- is 0.4667 crore shares. The count falls from 18.00 crore to 17.53 crore. Nothing that produces profit has been touched, so profit after tax is still Rs 225 crore and earnings per share moves from Rs 12.50/- to Rs 12.83/-, a rise of 2.7 per cent.
Nothing inside the business changed, so the 2.7 per cent is a division effect and not a performance effect, and it carries no information at all about how well the company is run. No machine ran faster. No customer paid more. No cost fell. The same Rs 225 crore is being cut into slightly fewer pieces, and each piece is therefore slightly larger. If four people share a Rs 400/- meal and one of them leaves without eating, the remaining three each get more food and nobody cooked anything.
There is a second number in the repurchase, and it is the one that matters for the comparison. The company paid Rs 300/- for a share carrying Rs 12.50/- of earnings. Earnings over price is an earnings yieldThe earnings attached to a share divided by the price paid for it, which is the earnings multiple turned upside down. A share bought at 24.0 times earnings yields 4.17 per cent. of 4.17 per cent. Rs 12.50/- over Rs 300/- is 4.1667 per cent, and it is the same number as one divided by the 24.0 times earnings the shares trade at. Keep hold of it. The 2.7 per cent rise in earnings per share is what the repurchase looks like; the 4.17 per cent is what the repurchase bought.
Put that price in context. The Rs 300/- is the one number in this entire comparison the company does not set for itself. At the illustrative Rs 300/-, the whole of Harivansh Packaging Limited is worth Rs 5,400 crore, and adding its own net debtBorrowings less the cash held against them. Here that is Rs 740 crore of borrowings less Rs 140 crore of cash. Settled in the statements layer and used here as a ready figure. of Rs 600 crore gives an enterprise valueThe value of the whole business before deciding who is entitled to it, so it covers the lenders as well as the shareholders. How it is built belongs to the material on valuation method. of Rs 6,000 crore, or 12.58 times EBITDA of Rs 477 crore. A repurchase is the only one of the six routes whose return is fixed by a price the company is a buyer in rather than a setter of. The final drawing below places it on exactly that axis.
A company repurchases shares and earnings per share rises. What changed inside the business?
How to compare Capital-Allocation Alternatives
Now the centre of the matter. There are six routes and an amount of money. Comparing them requires two figures against each name: how many rupees the route consumes, and what those rupees return. Getting the rupees right is mostly bookkeeping. Getting the returns right is where almost everybody comes unstuck, and the reason is always the same.
A return is not a number. A return is a number plus the basis it was computed on, and two returns are comparable only when their bases match. A published comparison of these routes almost always mixes them, quietly, because each route's return is naturally quoted in whatever form its own world quotes it in. The acquisition world quotes returns before tax on capital committed. The treasury world quotes them after tax on cash paid out. The market world quotes them on price. Line those up in a column and the column looks like a ranking. The column is not a ranking.
Four things have to be named beside every percentage.
- Before tax or after taxThe same return is roughly a quarter smaller after tax at Harivansh Packaging's 25.0 per cent rate. Mixing the two puts a 25 per cent thumb on the scale of whichever route happened to be quoted before tax.
- Before depreciation or after itA return struck on EBITDA flatters any route that consumes a physical asset, because the asset wearing out is a real cost that has been left out of the top line.
- On capital committed or on cash out of the doorAn acquisition commits an enterprise value and pays out an equity value, and the two differ by the target's borrowings. A percentage means something different depending on which of the two sits in its denominator.
- Contractual or expectedRepaying a loan returns what the loan agreement says. Buying a business returns what that business turns out to earn. Both are quoted as percentages and only one of them is knowable on the day.
The practical form of the rule is short: one setting of all four switches is chosen, every route is restated on it, and only then is the column read downwards. Restating is duller work than it sounds and it changes answers: the worked instance below reverses which of two routes is ahead.
One term is needed for that restatement: the after-tax cost of debtThe interest rate on a borrowing reduced by the tax the interest saves, so a 9.0 per cent facility at a 25.0 per cent tax rate costs 6.75 per cent after tax. Built in the material on valuation method and used here as an input.. The after-tax cost of debt is the interest rate reduced by the tax that the interest saves. The material on valuation method settles that rate and this comparison takes it as an input.
What return does every alternative have to beat?
Every comparison needs a floor, and the temptation is to import one. Somebody will suggest fifteen per cent because that is what was used at their last job. Somebody else will suggest the cost of borrowing. Both are guesses dressed as policy.
The floor is already sitting in the accounts. Return on capital employedOperating profit measured against the money the business has tied up in it, being net worth plus borrowings. Settled in the statements layer and used here as a ready quantity. for Harivansh Packaging Limited comes out of two lines it already publishes. Divide the Rs 339 crore of earnings before interest and tax (EBIT) by the capital that produced it, being net worth of Rs 1,650 crore alongside borrowings of Rs 740 crore, so Rs 2,390 crore in all. The answer is 14.2 per cent. The 14.2 per cent is what a rupee already inside this business earns today, before tax, without anybody doing anything new.
The floor is a fact about this business, computed from its own two statements, and it is not a rule imported from anywhere. A different packaging company with a different ladder would have a different floor. A floor that shifts from company to company is not a weakness in the method, it is the whole method: the question is never whether a route clears some general standard, it is whether the money does better there than where it currently sits.
The same floor after tax, at the effective tax rateWhat tax actually cost this company as a share of what it earned before tax, read off its own accounts rather than off any statute. Here Rs 75 crore against Rs 300 crore. of 25.0 per cent that this company actually bears, is 10.6 per cent. Some routes are naturally quoted before tax and some after, so both floors are needed, and the switch has to be set the same way on the floor as on the route being tested.
Notice what a stated floor does to the conversation. The conversation stops being a question about whether an acquisition is exciting and becomes a question about whether it beats 14.2 per cent. The floor also, uncomfortably, applies to the routes nobody thinks of as investments. Repaying a loan has to beat 14.2 per cent. Buying in shares has to beat 14.2 per cent. Leaving the money in the bank has to beat 14.2 per cent. Cash contributes nothing to EBIT at all, and the Rs 140 crore held therefore does not appear anywhere in that 14.2 per cent.
Harivansh Packaging Limited earns 14.2 per cent on capital employed. An alternative offers 7.42 per cent on the same basis. What does that establish?
What do all five routes actually cost?
Devyani Kulkarni has Rs 140 crore of cash and access to borrowing, and five things she could do. Here is each one priced, with its basis written beside it, exactly as the previous section demands.
Route one, the acquisition
Sundarban Polymers Private Limited, an unlisted maker of flexible packaging films that sells to some of the same customers, is available at an enterprise value of Rs 1,320 crore, or 10.0 times its EBITDA of Rs 132 crore. Run the bridge: Rs 1,320 crore less its net debt of Rs 180 crore leaves Rs 1,140 crore, and that Rs 1,140 crore is what reaches the sellers. The enterprise value is never the amount paid.
On the hurdle's own basis, its EBIT of Rs 98 crore over the Rs 1,320 crore of capital committed is 7.42 per cent, before tax. The 7.42 per cent is directly comparable with 14.2 per cent, and it is a little over half of it. Buying Rs 1,140 crore of equity against net worth of Rs 320 crore also leaves Rs 820 crore of goodwill before any allocation to identified intangibles, and the accounting for that sits in the accounting layer rather than here.
Route two, the project
Tapti Crossing Infrastructure Private Limited is a single-asset toll road company costing Rs 1,800 crore and generating EBITDA of Rs 248 crore. The toll road returns 13.8 per cent on the money committed. The 13.8 per cent is an upper bound and not a comparable figure: this record carries no depreciation for the crossing, and EBIT can only be lower than EBITDA. So the true like-for-like return sits somewhere below 13.8 per cent, and that puts it below the 14.2 per cent floor before the missing line is even filled in.
Route three, repaying borrowings
Harivansh Packaging Limited carries Rs 740 crore of borrowings and a finance cost of Rs 60 crore, so the implied average rate on its own book is 8.11 per cent. Repaying Rs 140 crore of it saves 8.11 per cent of interest, and because interest is deductible, the saving after tax at 25.0 per cent is 6.08 per cent. Repayment is a contracted rateA rate written into a specific agreement for a specific borrower, and therefore a fact about that agreement alone. It says nothing about what borrowing costs generally. outcome: the loan agreement says what the money costs, so repaying it returns exactly 6.08 per cent and not a paisa either side.
Route four, the repurchase
Rs 140 crore at the illustrative Rs 300/- of 28 August 2026 buys 0.4667 crore shares carrying Rs 12.50/- of earnings each, an earnings yield of 4.17 per cent. Earnings per share is already after tax, so 4.17 per cent is an after-tax figure and sits directly against the after-tax floor of 10.6 per cent.
Route five, holding the cash
Cash earns whatever it earns in the bank and contributes nothing to operating profit. The absence is not a slight against cash. The absence is a definition, and it is precisely why the Rs 140 crore held does not appear in the 14.2 per cent: the Rs 2,390 crore denominator is built from net worth and borrowings alone, and the cash sits outside the operating engine that produces the Rs 339 crore of EBIT.
Now set all four switches the same way
The ladder above is how these routes are usually presented, and it is unusable as a ranking. So restate everything after tax, on the cash that actually leaves the company. The floor becomes 10.6 per cent. Debt repayment was already on that basis, so it stays at 6.08 per cent. The repurchase stays at 4.17 per cent, for the same reason. The acquisition has to move: its Rs 61 crore of profit after tax over the Rs 1,140 crore actually paid is 5.35 per cent. And the project cannot be restated at all: this record carries no depreciation line for it.
| Route | Rupees out | As usually quoted | All after tax, on cash out |
|---|---|---|---|
| What the capital already earns | Rs 2,390 cr in place | 14.2% | 10.6% |
| Repaying borrowings | Rs 140 cr | 6.08% | 6.08% |
| The acquisition | Rs 1,140 cr | 7.42% | 5.35% |
| The repurchase | Rs 140 cr | 4.17% | 4.17% |
| The project | Rs 1,800 cr | at most 13.8% | not computable here |
| Holding the cash | Rs 140 cr | nil in EBIT | nil in EBIT |
Read the two right-hand columns against each other and the acquisition and debt repayment swap places: 7.42 per cent led 6.08 per cent, and 5.35 per cent trails it. Not one thing about either route changed. All that happened is that the tax switch and the denominator switch were set the same way on both. The swap is the entire content of the previous section, demonstrated on the case rather than asserted.
And the leverage, on both bases, every time
Route one is funded with Rs 140 crore of the company's own cash plus Rs 1,000 crore of new borrowing at a contracted 9.0 per cent, so the borrowings climb from Rs 740 crore up to Rs 1,740 crore while the cash balance empties to nil. There are two honest ways to state what that does to leverage and no third.
Standalone, the acquirer's own net debt of Rs 1,740 crore over its own EBITDA of Rs 477 crore is 3.65 times, against an opening 1.26 times. Buying the whole of Sundarban Polymers Private Limited carries its Rs 180 crore of borrowings across as well, so the consolidated figure is Rs 1,920 crore set against the two businesses together at Rs 609 crore, and that ratio is 3.15 times. Both are true. The two figures describe different perimeters, and every sentence quoting either has to say which one it is on. Both businesses carried a 15.0 per cent margin before, so the combined business, for completeness, has revenue of Rs 4,060 crore and EBITDA of Rs 609 crore at that same margin.
A third figure appears in drafts and is never right. Rs 1,740 crore over Rs 609 crore is 2.86 times, and it puts a standalone numerator over a combined denominator. The 2.86 times understates the standalone row by 0.79 turns and the consolidated row by 0.29 turns, and it is wrong on both.
One route buys earnings at 4.17 per cent on the illustrative Rs 300/- price and another at 5.35 per cent on the Rs 1,140 crore paid. Which one would be expected to raise earnings per share?
A draft says leverage after the acquisition moves to 2.86 times. What must be checked before quoting any leverage ratio?
Why does an earnings per share test point the wrong way?
Now the two rankings, set against each other. Rank the acquisition and the repurchase by what each rupee returns, and the acquisition is ahead: 5.35 per cent against 4.17 per cent, both after tax on the cash paid out. Now rank the same two by what each does to earnings per share.
The repurchase, funded with the Rs 140 crore of cash the company holds, takes the share count from 18.00 crore to 17.53 crore and lifts earnings per share from Rs 12.50/- to Rs 12.83/-, a rise of 2.7 per cent. The acquisition, funded with Rs 1,000 crore of borrowing at the contracted 9.0 per cent, brings in the target's Rs 61 crore of profit after tax and carries Rs 67.5 crore of after-tax interest, so profit after tax becomes Rs 218.5 crore on an unchanged 18.00 crore shares. Earnings per share is Rs 12.14/-, a fall of 2.9 per cent.
The two rankings point opposite ways, and not one thing about either route's return moved between them. The route that bought the lower yield produced the better earnings line, and the route that bought the higher yield produced the worse one.
The mechanism is not subtle once seen, and it is worth being precise about. Interest on borrowed money is a line in the profit ladder. Cash sitting in the bank contributes nothing to operating profit, so spending it removes nothing from the ladder either. A route funded with borrowing is therefore charged in the earnings test and a route funded with cash is charged nothing at all. The test is reporting the funding mix and not the return.
Prove it by holding the route fixed and moving only the cheque. Fund the same Rs 140 crore repurchase entirely with borrowing at the same contracted 9.0 per cent. The interest is Rs 12.6 crore, or Rs 9.45 crore after tax, so profit after tax becomes Rs 215.55 crore. Rs 140 crore at Rs 300/- buys the same 0.4667 crore shares whatever paid for them, so the share count is identical at 17.53 crore. Earnings per share is Rs 12.29/-, a fall of 1.7 per cent. The identical transaction on the identical shares has swapped verdicts, and the 4.17 per cent it bought never moved.
Put both sides on one base, or the arithmetic does not tie
There is a trap in reconciling the acquisition's 2.9 per cent fall, and it catches people who are being careful rather than careless. The target's earnings yield is 5.35 per cent, struck on the Rs 1,140 crore actually paid. The funding cost has to be struck on that same Rs 1,140 crore: Rs 67.5 crore over Rs 1,140 crore is 5.92 per cent. The spread is minus 0.57 points. Minus 0.57 points on Rs 1,140 crore is Rs 6.50 crore, and Rs 6.50 crore across 18.00 crore shares is Rs 0.36/-, and Rs 12.50/- less Rs 0.36/- is Rs 12.14/-. The chain reconciles exactly.
Now the tempting alternative. Rs 67.5 crore over the Rs 1,000 crore of new borrowing alone is 6.75 per cent, a perfectly correct after-tax cost of that borrowing and a completely different base. Set 5.35 per cent against 6.75 per cent and the spread of 1.40 points on Rs 1,140 crore implies a fall of Rs 15.95 crore, against an actual fall of Rs 6.50 crore. The chain now reconciles to nothing. Use 5.92 per cent wherever the arithmetic has to tie, and reserve 6.75 per cent for describing the cost of the borrowing itself.
The target's profit after tax is locked in this record at Rs 61 crore and it computes to Rs 61.35 crore: EBIT of Rs 98 crore less Rs 16.2 crore of interest on its own Rs 180 crore at 9.0 per cent is Rs 81.8 crore, and at 25.0 per cent tax that is Rs 61.35 crore. On the exact figure the chain lands at Rs 218.85 crore, Rs 12.16/- and a fall of 2.7 per cent rather than Rs 12.14/- and 2.9 per cent. The route dilutes on either figure, so the teaching point survives and only the headline was a shade overstated. No exact decomposition is built on the 2.9 per cent.
The same Rs 140 crore repurchase is funded with borrowing instead of cash. What happens to the return it bought?
One repurchase, one control, and one quantity that refuses to move
The repurchase is fixed at Rs 140 crore, the price at the illustrative Rs 300/- of 28 August 2026, the borrowing rate at this company's contracted 9.0 per cent and tax at 25.0 per cent. The control moves how much of that Rs 140 crore is borrowed rather than taken from held cash. Watch the left panel and then watch the right one.
The acquisition takes Rs 1,140 crore. How many years of this company's own cash generation is that?
What does each route take off the table?
Capital is spent once. The sentence sounds obvious and it is the part that never appears in the arithmetic of a single route. The arithmetic of a route is done on its own, and its own arithmetic has no term for what else there was.
Size it. Harivansh Packaging Limited generates about Rs 342 crore in a year: EBITDA of Rs 477 crore, less the Rs 60 crore finance cost, less the Rs 75 crore of tax. The Rs 1,140 crore paid to the sellers is three and a third years of everything this business produces, so taking that route removes the growth spending, the repayment and the return of capital for that long. The project equity of Rs 540 crore is about a year and a half. The Rs 140 crore that funds either the repayment or the repurchase is about five months.
The Rs 342 crore is a constructed proxy, not an operating cash flow. The record carries no cash flow statement, so EBITDA less finance cost less tax is the closest thing the figures support. The proxy sits before any working capital movement and before anything spent on keeping the machines standing, and the record carries no figure for that spending. So the three and a third years is a floor: the true foreclosure is longer, by an amount this record does not disclose. The proxy is also one year's rate, applied as a size conversion. Nothing in this record says the next year produces Rs 342 crore, and the rate is not projected forward.
Is a certain 6.08 per cent the same as an expected 7.42 per cent?
Two percentages that look alike can be entirely different kinds of statement, and this is the fourth switch from the comparison section doing its real work.
Repaying Rs 140 crore of borrowing at an implied 8.11 per cent returns 6.08 per cent after tax, and it returns exactly that. The interest stops. There is no version of the world in which the loan quietly charges more or less than the agreement says. Buying Sundarban Polymers at 7.42 per cent returns 7.42 per cent if the acquired business earns what it earned last year, and nobody knows on the day of signing whether it will. Customers move. Input costs move. The people who ran it may or may not stay.
Setting a certain 6.08 per cent against an expected 7.42 per cent, with nothing said about which is which, has hidden the only difference that actually separates them. Think of the household again: a fixed deposit paying six per cent and a nephew's shop that might return seven are not two versions of the same choice, and no reasonable person treats them as one.
Debt repayment returns 6.08 per cent and the acquisition 7.42 per cent. Is the acquisition the better use?
Who reads this, and what does each of them look for?
The same five routes are read differently depending on who is reading them, and all three readings are worth seeing. The differences are not opinions, they are different questions.
A lender reads the routes backwards, starting from what each one does to the ratio it has written into its agreements. A credit officer looking at Harivansh Packaging Limited cares that the acquisition takes standalone leverage from 1.26 times to 3.65 times, and that the repurchase and the repayment leave it at 1.26 times or improve it. The 5.35 per cent and the 4.17 per cent barely register. The figure that registers is that one route consumes three and a third years of generation and lands the borrowings at Rs 1,740 crore, and the lender's own question is whether the schedule it has agreed still gets serviced afterwards. Where a covenant ceiling is written on the standalone base, reading it against the 3.15 times consolidated row is the same basis mistake in different clothes.
An analyst covering the company reads the routes for what they say about the people running it. Not because the return figures are uninteresting, but because they are computable by anybody and the pattern of choices is not. A company that repeatedly buys in shares at a 4.17 per cent yield while its own capital earns 14.2 per cent is saying something about how it thinks, and so is one that keeps finding projects. The analyst also has the one weapon the basis rule provides: every route restated on one basis before any published ranking is believed. On this case the restatement changed the order, and it usually will.
And a household faces the identical short list with the identical trap. The Rs 8,000/- left at the end of the month can pay down the scooter loan at a known rate, or go into the nephew's shop at a hoped-for one, or into the fixed deposit. The trap that catches a finance team catches a household in exactly the same form. Paying down the loan makes the monthly outgo look better immediately, and that feels like progress. The shop shows nothing for a year and might be the better use of the money. The visible monthly number is measuring how the money was found, not what it bought.
The error that gets made, and what it costs
A finance team screens the acquisition and the repurchase on earnings per share alone. The screen shows minus 2.9 per cent against plus 2.7 per cent, and the note concludes that the repurchase creates value and the acquisition destroys it. The ranking is upside down. The repurchase bought earnings at a yield of 4.17 per cent on the illustrative Rs 300/- it paid; the acquisition bought them at 5.35 per cent on the Rs 1,140 crore it paid. Both sit far below the 14.2 per cent the existing capital earns before tax and the 10.6 per cent it earns after.
The entire difference in the earnings test came from one route being funded with cash that carries no charge in the profit ladder and the other with borrowing that does. Funding the identical repurchase with borrowing at the same contracted 9.0 per cent slides its earnings verdict from plus 2.7 per cent to minus 1.7 per cent, crossing over at about 61.7 per cent borrowed. The 4.17 per cent it bought sits perfectly still throughout.
Who makes it: very nearly everybody. The earnings test is one line and the return test is four, and a paper showing accretion is easier to present than one showing a 4.17 per cent yield against a 14.2 per cent hurdle. The cost: Rs 140 crore committed on a test that would have reported the opposite verdict on the same transaction under a different funding mix.
The fix is one sentence long. Compute what each rupee returns and name the basis; use the earnings effect to describe the funding, never to rank the uses.
India, and what is left to the source
The Securities and Exchange Board of India, at sebi.gov.in, sets out what a listed company must do and disclose in order to buy in its own shares. The Ministry of Corporate Affairs, at mca.gov.in, holds what the Companies Act asks of the act of a repurchase itself. How any of these routes is taxed, in the hands of the company or of the shareholder, sits with the income tax authority at incometaxindia.gov.in.
Every requirement, threshold, period and permission belongs to the three authorities named above and should be read from them. The 25.0 per cent used throughout is this company's own effective rate taken from its own accounts rather than any statutory figure. The comparison above is division and subtraction, so it holds in any market: a rupee spent once is unavailable everywhere.
Where does the arithmetic stop?
Everything above is computable, and none of it settles the choice. A costed comparison that stops short of the answer is not a disappointing ending, it is the honest description of what this kind of work produces.
Four things can be known: what each route costs in rupees, what each returns as a percentage with its basis named, how certain that return is, and how many years of the company's own production each one consumes. Three things cannot be known from any of it: what the business needs next, what the alternative would have grown into, and whether the people involved can actually run what they are buying. The output of capital allocation work is a costed comparison sheet with a blank last column, and the blank is not a gap in the workings.
So the costs, the returns, the certainties and the foreclosures are stated for Harivansh Packaging Limited, and nothing further. Which route Devyani Kulkarni should take is not settled by any of it: that answer depends on information no published figure carries.
What does this comparison produce once the five routes are costed?
References
The mechanism above is division. Three sites are named because the subject brushes against acts a company must be permitted to perform.
| Named for routing | What sits with it | Site |
|---|---|---|
| Securities and Exchange Board of India | What a listed company must do and disclose in order to buy in its own shares | sebi.gov.in |
| Ministry of Corporate Affairs | What the Companies Act asks of the act of a repurchase itself | mca.gov.in |
| Income tax authority | How any of these routes is taxed, for the company and for the shareholder | incometaxindia.gov.in |
| This constructed case record | Every rupee and percentage above, each recomputed here rather than transcribed | built for teaching |
Harivansh Packaging Limited, Sundarban Polymers Private Limited, Tapti Crossing Infrastructure Private Limited, Devyani Kulkarni and Ashwin Rege are invented.
Educational material. Not advice on any investment, tax, budget or market position.