How Capital Allocation Shapes Long-Term Business Outcomes

A business added Rs 24,40,000/- to its standing base inside one year. Sales grew Rs 30,00,000/-, what those sales left behind grew Rs 12,90,000/-, and operating profit dropped Rs 11,50,000/-. Take the Rs 24,40,000/- away from the Rs 12,90,000/- and the fall appears whole, with no remainder left for anything else. Both the commitment and the consequence are on the record. The person who chose is not.

Revenue rose Rs 30,00,000/- and profit fell Rs 11,50,000/-. Where did the year go?

The arithmetic here is short enough to hold in mind, and it settles the question before any vocabulary is needed. So the accounts come first and the definitions afterwards. Anjani Stationers Private Limited, a business invented for these notes and trading nowhere, published two years side by side. In the second of them it sold more than in the first and earned less.

One subtraction carries the whole argument. Across the year, contributionWhatever is left of the money coming in once every cost that swells and shrinks with output has been paid out of it. came in Rs 12,90,000/- higher. The standing baseThe spending that turns up in the same size whichever way the year's volume goes, so it neither grows with a busy month nor shrinks with a quiet one. came in Rs 24,40,000/- higher. Take the second of those away from the first and what remains is Rs 11,50,000/-, the entire drop in operating profitWhat survives after both the costs that travel with volume and the costs that stand still, and before any interest or tax is taken off.. Nothing is left over for a second cause to occupy, so the drop is not partly this and partly that: it is one movement, fully accounted for.

The single-cause finding matters because of what a reader reaches for first. Told that sales grew and profit shrank, almost everybody goes looking for trouble in the trading. Did the price slip? Did paper get dearer? Did somebody discount to win a school? The published margins close that door without an argument. The share of revenue left after the costs that move with output stood at 42.75 per cent in the first year and 42.78 per cent in the second, three hundredths of a point apart. The gross margin is the wider measure, taken before the standing costs, and it held at exactly 45.0 per cent across both years. Shareholdings are written as percentages too, so the measure has to be named whenever a percentage is: 45.0 per cent here is the gross margin, every time it is written.

The household version has the same shape. The rent went up by two thousand and the salary rose by nine hundred rupees a month, so the month feels tighter than it did. The grocery bill is not where the movement happened. Going through it line by line will find nothing, and it will find nothing however long the search runs. Nothing about what a sale leaves behind moved, and the whole of the fall sits in what the business committed itself to.

So what counts as an allocation of capital, and how many sit in this one year?

Here is the plain version, and it is worth stating as a thing somebody did rather than as a category on a list. An allocation of capital is money committed now, in a shape that is awkward to reverse, against something the business hopes will arrive later. The commitment is made before the result is known. The timing is not a weakness in the decision; that is what makes it a decision at all. If the answer were already visible there would be nothing to decide and no reason to pay anybody to do it.

Everything difficult about an allocation of capital follows from one gap: the commitment lands on a date, and whatever it was meant to produce arrives on a schedule of its own that buyers control. A cost committed in a year is charged in that year, whole. Volume that the cost was put in place to serve turns up when people decide to buy. A reader holding a single year is looking at all of the cost against a fraction of the thing the cost was for, and no amount of arithmetic on that one year repairs the mismatch.

Three commitments sit inside this one published year at this one business, and the important thing about them is not their size but their unit. Stating the amount and the unit in the same breath, every time, is what keeps the rest of the account honest.

Three commitments and three units, and none of the three is added to another, divided by another or ranked against the others. The restraint is not fussiness. Keeping the three units apart is the only thing standing between a reader and a tidy-looking number that answers no question anybody has, and the failure block further down is that number's obituary.

Each of the three is handled differently. The second has a consequence published against it, in the accounts, in the same year. The first has a published non-consequence against it, a stranger and more useful thing than it sounds. The third has a published sentence about who could reverse it, and that sentence turns out to carry more than it looks like it carries.

The standing base rose Rs 24,40,000/-. What is that in the thing the works actually sells?

Rupees are a poor unit for feeling the size of a commitment. Rs 24,40,000/- is either enormous or trivial depending on a business not yet described. Registers are what this works makes and sells, and pricing the commitment in registers gives the better feel. In the published second year a single register leaves Rs 46.20/- behind once the costs that move with volume are settled, and that rate is simply the year's contribution of Rs 1,15,50,000/- shared out across the 2,50,000 registers made.

Do the division and the commitment turns into a quantity of selling. The Rs 24,40,000/- of new standing base, taken at that contribution a registerThe part of one register's price that survives the costs which climb and fall alongside output. of Rs 46.20/-, works out at 52,813.85 registers of extra selling in a year, or 21.13 per cent of the year's 2,50,000. The shortfall of Rs 11,50,000/- at the identical rate is 24,891.77 registers, or 9.96 per cent of that same 2,50,000. The convention belongs to both figures, so say it in the same breath as both: each one is priced at the rate the second year produced, and no rate for the first year was ever published. Either both figures travel with that convention attached, or neither travels.

Read the first of those two numbers slowly. The 52,813.85 is the honest size of the decision, and it does not mean selling 52,813.85 extra registers on one occasion. The requirement is to sell that many more every year, from now on, simply to stand where the works stood before the commitment was made. A commitment to a standing cost is a commitment to sell a particular quantity of things every single year, and that quantity is the real measure of what was taken on.

A shopkeeper who moves into a bigger unit knows this in the bones without doing any arithmetic. The rent is not really a rent; it is a number of extra customers a month, arriving whether they feel like it or not, forever. The lease turns a hope about footfall into an obligation about footfall, and the obligation is the part that keeps somebody awake.

Where did the Rs 24,40,000/- go, and how much of that is really disclosed?

Two published descriptions of that spending exist, written in different places for different reasons, and the difference between what each one claims is exactly what a careful reader is supposed to notice.

The first description is short. The comparison setting the two years side by side names three destinations for the growth: people, space, and a binding operation this business had bought into. Three destinations, stated plainly, and a price is attached to none of them. That is a description of where money went, not a breakdown of how much went where.

The second description is arithmetic. A separate treatment splits the same Rs 24,40,000/- three ways: employee benefits up Rs 6,00,000/-, the fixed part of other operating costs up Rs 11,40,000/- and mostly a second warehouse taken during the year, and depreciation and amortisationThe cost of an asset spread across the years it is expected to serve, rather than charged in full in the year it was bought. up Rs 7,00,000/- on the assets bought. The three amounts sum to Rs 24,40,000/- precisely. And that treatment puts its own label on its own working, in its own words: the split is an estimate, not a disclosure. The estimate label travels with those three amounts wherever they go. Without it, three estimates start reading as three disclosures within about a paragraph.

Two further things must not be done with the pair, and both are tempting. The first is to declare that nothing is known about where the money went, a claim that is plainly false: the total is disclosed and three destinations are named. The second is to lay one description over the other and match them up, joining people to employee benefits, space to the warehouse, and the binding operation to whatever is left. No document performs that join. Building it here would manufacture a disclosure out of an estimate, and it would look entirely respectable while doing it.

The total is disclosed, one account names what it went on without prices, another carries a split that its own source calls an estimate, and no document anywhere sets a disclosed price against any one of the things named. The sentence above is longer than a reader would like, and it is the shortest true version available.

Rs 21,00,000/- bought seventy per cent of a binding workshop. What did that buy, exactly?

Anjani Stationers put Rs 21,00,000/- of cash into a 70 per cent holding in Chitra Binding Works Private Limited, also invented, at the start of its second year. Chitra had until then been its own company under its own control, and it had been invoicing Anjani Stationers for binding the printed sheets. Why that particular workshop matters takes one clause: binding turns out 100 an hour where printing manages 125 and cutting 150. Binding is therefore the slowest of the three stages and the one setting the pace for the whole works, and it was the single stage of the three that got bought in.

One piece of arithmetic runs on the price, and it runs as a check rather than as a construction. The outside share of the workshop's net assets on the day of the purchase is published at Rs 7,50,000/-. The outside share is 30 per cent, so the whole of the net assets stood at Rs 25,00,000/-, and 70 per cent of that is Rs 17,50,000/- acquired. Rs 21,00,000/- paid less Rs 17,50,000/- acquired leaves Rs 3,50,000/-, and Rs 3,50,000/- is exactly the goodwillThe difference between what was paid for a share of a business and the value of the share of its net assets that came with it. published for this purchase. The circle closes, and closing it is the reason for running the working at all.

Carry the ruling with the figure and not only the figure. The Rs 3,50,000/- of goodwill is a difference between two amounts. The goodwill is not a valuation of the workshop's name, its customer list or its founder's judgement, and it is measured against the share acquired and never against the whole Rs 25,00,000/-. One warning about the amount itself: Rs 3,50,000/- is also this business's finance cost for the year, an unrelated figure that happens to be the same size, so say goodwill or say finance cost every time the amount is written and never leave it bare.

What did the stake earn in its first year? The workshop's own earning for the year came to Rs 10,00,000/-. Out of that, Rs 7,00,000/- is the group's share and Rs 3,00,000/- belongs to holders sitting outside the group. Rs 7,00,000/- set beside the Rs 21,00,000/- paid gives 33.33 per cent, and the convention has to be said in the same sentence or the figure should not be printed at all: that is one published year's share of profit measured against one published price. The 33.33 per cent is a ratio, not a return. Nothing publishes what the group's share was in any other year, so it is never annualised, never compounded, never called a payback, and no later period is named here.

Now the part worth the whole block. On the day those shares changed hands, binding was still turning out 100 an hour, and rated capacity still stood at 4,00,000 registers in a year. An earlier treatment in these notes puts it in a sentence worth quoting rather than improving: holding the stage bought the right to make the change, and it did not buy the change. The unchanged rates are the published non-consequence, and they are the most instructive point in the whole account. A commitment can be large, correct, and entirely invisible in the operating figures on the day it completes. How far a stake pulls one interest towards another, and where the pull runs out, is worked separately under Insider Ownership: Alignment and Its Limits.

The whole year's profit stayed in the business. Is a decision not to move money an allocation?

No dividendA payment out of profit to the people holding the shares, arrived at by a decision rather than arriving on its own. was paid by Anjani Stationers during the year, and its full Rs 30,00,000/- of profit after tax went into retained earningsThe profits of every past year that were kept inside the business rather than paid out, accumulated in one line., and those now total Rs 1,02,00,000/-. The same amount in the same year is also the rise in revenue the opening block already used, so which Rs 30,00,000/- is meant has to be named every time it is written. Here it is the profit after tax.

A decision not to move money is still a decision about money, and this one is the largest of the three commitments the year carries. Nothing left the bank, no asset changed hands, and no entry announced itself in bold on the face of anything. A claim that could have travelled to the people holding the shares was directed back into the business instead, and it will now be spent on whatever the business spends money on.

The source publishing that retention adds a sentence which is the only place in these notes where a decision about capital is joined to the people able to reverse it: three holders agreed to that, and three holders can change it at the next meeting, and on a register of thousands the same decision would take a great deal more explaining. The sentence reads in both directions, and both readings are worth stating. It establishes that the retention is reversible and that reversing it is easy to arrange, and that is a real fact about how this business is governed. The sentence does not establish who the three are. The register carries two founding households plus one holder from outside, and not a single one of those three entries carries anybody's name.

The household version is exact. Four people at one dinner table agree that this year's savings stay in the shop rather than being divided. The agreement is real, the money is genuinely committed, nobody signed anything, and the whole arrangement could be undone over another dinner in April. Keeping a year's profit inside a business set beside handing it over is covered separately under Growth Investment vs Capital Return: One Pot, Two Uses.

Why can the purchase price not be set beside the rise in the standing base?

State the rule in the positive form first. A purchase price is paid once, on a date, and then sits on the balance sheet as something the business holds. A standing base is a cost that arrives again next year and the year after and the year after that. A price and a base are two different kinds of quantity, and no arithmetic operation joins them into a third quantity that means anything.

A rule left as a rule is only somebody's convention. Published support for this one exists, and hardly anybody knows it is sitting there. An earlier treatment in these notes states that the two years of standing base are standalone figuresThe parent company taken by itself, with nothing of the companies it controls folded in., the parent company by itself. So the wages and running costs of the binding operation have never formed part of the Rs 74,00,000/- at all. The same treatment then names the only two narrow routes by which the purchase reaches that base. The first is the one-time cost of doing the transaction, the professional fees and the stamp duty of buying a business. Such a cost is charged as incurred and therefore sits inside the second year and inside no part of the first year's comparative. The second is the relationship: work that may now be bought from a connected operation rather than from somebody unconnected.

So the two commitments touch each other by two narrow routes and meet in no single figure, and these notes established that separately beforehand. The arithmetic a reader is by now itching to try is still sitting there, and the honest handling is to name it without performing it. The two amounts are not divided by each other anywhere in these notes. The division is arithmetically valid and produces a clean-looking quotient, and that quotient has no unit at all, so it answers no question anybody actually has. The source carrying the price puts its own bar in its own words: that figure is never divided by anything there.

The everyday version is the deposit on a shop and its monthly rent. Both are real money, both go to the same landlord, both are written in the same agreement, and dividing one by the other gives a tidy figure that describes nothing. Nobody would print it. Everybody can feel that a deposit and a rent are different sorts of thing. Written as rupees in a spreadsheet, the same two quantities lose that feeling entirely.

The outcome is published. So was the decision a good one?

The decision cannot be graded, and the reason is not delicacy or a wish to seem even-handed. The material a grade would need is not published anywhere, and the useful move is to write down exactly what that material would be.

Five things would have to exist. One, a second year and a third. A base that moved 49 per cent in a single year is a base still in motion, and any reading built on it carries the same instability. Two, the volume the capacity was put in place for, and the date by which it was expected. A commitment is judged against what it was for and nothing else. Three, what the business would have earned had it committed nothing. Every judgement of an allocation quietly assumes that comparison, and almost nobody writes it down. Four, a disclosed price against each of the things the money went on. No document carries one: one account names three destinations without prices and another carries a split its own source calls an estimate. Five, who decided, on what information, and what they expected to happen.

Not one of the five is published. So the arithmetic stands, what would be needed to judge it is written out, and nobody publishes that material, and that is a finding rather than a gap.

One more thing needs saying plainly. A reader who has watched profit fall will otherwise supply it themselves. A cost that lands in one year against volume that has not arrived yet is the ordinary shape of putting capacity in place. A business in that position has not done anything foolish. Such a business has done the ordinary thing, in the ordinary order, and a single year's accounts cannot yet report on how it turned out.

Two decisions, two consequences. So who decided?

The question is worth asking directly and then following in the open. The search itself should be visible, not only its result. A name against the standing base commitment: there is none. A name against the purchase of the binding workshop: there is none either. The accounts say what was committed and what followed, and they say it in rupees, and they say nothing whatever about the person who chose.

There is exactly one place in these notes where a decision about money carries a name and a body against it, and it is a small one. A board of directors chaired by Anjani Kulkarni approves the purchase of a Rs 15,00,000/- machine for the following year, and the money is set aside. The same amount appears elsewhere as an internal figure removed when two sets of accounts are folded together, so which Rs 15,00,000/- is meant has to be named. The second meaning is not used here at all. The reach of a board of directors, and where it ends, is covered separately under The Board: Composition, Committees and What It Controls.

Now put the three side by side. The shape they make is the finding. The one decision with a name against it carries no consequence yet, and the accounts record nothing until the machine arrives. The two decisions with consequences attached carry no name at all.

The silence is not a defect in these notes and not a defect in the accounts either. It describes where a reader outside a business permanently stands. From out there, what was committed and what followed are routinely visible. Who chose, on what information, and what they were expecting are routinely not. A reader who quietly fills that gap has written the most confident sentence in their file and the least supported one, and nobody downstream will be able to tell the difference by looking at it. A shop on an ordinary street is the same case: it changed hands last spring and it is busier now. Both facts are plain from the pavement, and who decided what is not written on the window.