Capital Raises: Dilution, Signal and Use of Proceeds

A capital raise brings money in and creates new shares, so earnings per share falls while book value per share rises. Whether a holder is worse off depends on the terms: a rights issue hands existing holders the entitlement that offsets the dilution, and a placement does not. Where the money goes decides everything after that.

Two figures move whenever shares are created, and they move in opposite directions at the same instant. Profit is divided among more shares, so the earnings figure per share falls. Net worth is enlarged by the money that arrived, and it is enlarged proportionally more than the count wherever the issue price sits above the existing book value per share, so the book value figure per share rises. Everything below is that one sentence, worked out on Sarvani Coatings Limited and then run backwards.

Sarvani Coatings Limited has not raised money, has not placed shares and has not bought any back. All three actions below are HYPOTHETICAL and none is proposed, and no company runs all three at once; each is set on the same published figures so the arithmetic can be compared across them.

What actually happens to a holder when new shares are created?

Start at the kitchen table. The finance version is the same picture. A household cooks one pot of rice for twenty people. Two more sit down. Every plate is now smaller, and nothing about the rice changed. Dilution is that smaller plate, and it is worth noticing what dilution is a statement about: a proportion, not a quantity of rice.

Now change one thing. The two who sat down carried in a sack of rice. The plates are still divided twenty two ways, so each share of the pot is a smaller fraction than it was, and the pot is bigger than it was. Both of those are true at once, and almost every bad sentence written about a capital raise comes from reporting one and forgetting the other.

Dilution is a statement about the size of a holder's proportion, never about the size of their money, and the two come apart the moment cash arrives alongside the new shares. The loose use of the word dilution is what makes a raise sound like something taken from a holder, so proportion and money are worth holding apart throughout.

Sarvani Coatings Limited published profit after tax of Rs 278 crore in its most recent year, on a share count of 24.00 crore. The count of 24.00 crore is carried forward from the record of a completed split and a completed bonus, and every per-share figure below is built forward from that count rather than solved backwards out of a printed per-share number. Rs 278 crore over 24.00 crore shares is Rs 11.58/- of earnings per share. Create 1.20 crore new shares and the same Rs 278 crore is divided 25.20 crore ways instead. The result is Rs 11.03/-, a fall of 4.76 per cent. The 4.76 per cent is not an opinion about the business. The figure is 24.00 divided by 25.20, and nothing else.

Why does one per-share figure fall while the other rises?

Because the two figures are built on different tops and the same bottom. Earnings per share is profit over the share count. The raise did nothing to profit and added to the count, so it can only fall. Book value per share is net worthWhat is left over for equity holders once every liability has been taken off the assets. It is a balance sheet leftover, not a market value. over the same share count, and the raise added to both. Which one wins is settled by one comparison rather than by a judgement.

Here is the comparison. Book value per share rises whenever the issue price is above the book value per share that already exists, and it falls whenever the issue price is below it. Written out, the new figure is net worth plus price times new shares, all over count plus new shares, and that is larger than net worth over count exactly when the price exceeds net worth over count. Sarvani Coatings carries net worth of Rs 1,486 crore, so book value per share is Rs 61.92/-. A rights price of Rs 350/- is many times the existing book value per share, so the figure rises hard.

Work it through. Rs 1,486 crore plus the Rs 420 crore raised is Rs 1,906 crore of net worth, over 25.20 crore shares, or Rs 75.63/-. So in one transaction earnings per share went from Rs 11.58/- to Rs 11.03/-, down 4.76 per cent, and book value per share went from Rs 61.92/- to Rs 75.63/-, up 22.16 per cent. An account that reports the fall and not the rise has reported one half of an arithmetic identity and called it a result.

Note what did not move. Capital employedNet worth plus borrowings, meaning the total money put to work in the business whoever supplied it. rose by exactly the money raised and not a rupee more. Return on equityProfit after tax measured against net worth, so how much a rupee of equity produced in a year. falls once the same Rs 278 crore sits on a larger net worth. The fall is arithmetic too, until the money is put to work. Nobody has yet done anything with the money at this point in the story, and the whole of the interesting question lives there.

What is a rights issue, and why does it leave a holder exactly where they started?

A rights issue offers the new shares to existing holders in proportion to what they already hold, at a price below the market price. The shopkeeper version: the shop is taking in more stock and offering the regulars first refusal at a lower price, so a regular who takes it up is not made worse off by the expansion. The entitlement is the offer itself, and it has a value of its own even before anybody takes it up.

Sarvani Coatings Limited, hypothetically, offers one new share for every twenty held at Rs 350/-, against an illustrative price of Rs 486/-. The discount is 27.98 per cent, and the first instinct on seeing it is that something is being given away. Work the theoretical ex-rights price and the instinct dissolves. Add twenty holdings priced at Rs 486/- to the single new one bought at Rs 350/-. The total is Rs 10,070/- spread across twenty one shares, or Rs 479.52/- each. Rs 479.52/- is what a share is worth on this arithmetic once the new one exists.

Now the identity, and it is the single most useful thing here. A holder of twenty shares loses Rs 486/- less Rs 479.52/- on each one. The loss is Rs 6.48/- a share, or Rs 129.52/- across the twenty on the unrounded Rs 6.4762/-. The same holder gains Rs 479.52/- less the Rs 350/- paid on the one new share, a gain of Rs 129.52/-. The loss on the old shares and the gain on the new one cancel exactly, and a holder who takes up the entitlement ends the transaction precisely where they started.

Multiplying the rounded Rs 6.48/- by twenty instead gives Rs 129.60/-, eight paise adrift. The eight paise gap is rounding and nothing else. Anybody who recomputes will find it, and an account that hides an eight paise gap teaches the reader to distrust the eighty rupee ones.

The proportion test says the same thing from the other end. Twenty shares had a claim on Rs 231.67/- of the year's profit before. Take up the entitlement and twenty one shares out of 25.20 crore have a claim on Rs 231.67/-, unchanged to the paisa. Let the entitlement lapse and twenty shares out of 25.20 crore claim Rs 220.63/-. The dilution was never automatic. The outcome was conditional on what the holder did.

One caution, said plainly. The theoretical ex-rights price is an arithmetic construct and not a forecast of where the share will actually trade. The identity holds on that construct, and the arithmetic guarantees only that the terms themselves take nothing from a holder who takes them up. The price afterwards is a different question, settled elsewhere.

What does a follow-on offering change, and what does it not?

A follow-on offering places the new shares with buyers who need not be existing holders, usually institutions. The shop is still taking in more stock, and the discounted first refusal now goes to a wholesaler down the road instead of to the regulars. Everything about the expansion is the same. The regulars simply do not get handed anything.

Sarvani Coatings Limited, hypothetically, places the same 1.20 crore shares at Rs 460/-, a 5.35 per cent discount to the illustrative price, raising Rs 552 crore. The share count goes to the same 25.20 crore. Earnings per share falls to the same Rs 11.03/-. More money arrived for the same number of shares, so book value per share rises further, to Rs 2,038 crore over 25.20 crore, or Rs 80.87/-, up 30.62 per cent.

The whole comparison sits in one line. The dilution is identical, the money raised is Rs 132 crore larger, the book value figure is better for everybody, and the existing holder receives no entitlement to take up or sell. The share count effect is the same in both, and the only thing that differs is whether an offset is placed in the existing holder's hands. A rights issue and a placement are always compared for exactly that reason.

Two other things move that the per-share figures do not show. The shareholding patternThe disclosed split of a company's shares between the promoter group, institutions and the rest of the public. changes. The new shares sit with the buyers rather than spreading across the existing register. And the promoterThe holder or group of holders who control a listed Indian company, as the disclosed shareholding pattern reports it. share falls unless the promoter subscribes. In a rights issue the promoter can subscribe on the same terms as everyone else. Neither of those is a per-share number and both are read separately.

Does the issue price reveal what management thinks the shares are worth?

No, and this is where a lot of confident writing goes wrong. A rights issue is priced below the market price because an issue priced at or above it would not be taken up: a holder would simply buy in the market instead. The discount is a mechanical feature of getting the issue subscribed. The wedding hall booked eight months ahead is cheaper than the hall booked on the day, and the discount says something about filling the calendar rather than about the value of the hall.

Look at what the two hypothetical prices actually do. Rs 350/- is 27.98 per cent below the illustrative Rs 486/-, and once the theoretical ex-rights price is worked, the reference price is Rs 479.52/- and the holder identity nets to nil. Rs 460/- in the placement is 5.35 per cent below the same reference and raises Rs 132 crore more. Two very different looking discounts, and neither carries a message about worth that the arithmetic can extract.

Reading an issue price as management's opinion of value is an inference the disclosure does not support. There are things a raise genuinely does say: that money was wanted, roughly how much, and from whom. The use the money is applied to is disclosed, and that disclosure is read alongside the price. The worth of the shares is not in the price of the issue.

What does the use of proceeds change, when the per-share arithmetic does not move?

Consider a household borrowing Rs 5,00,000/-. The instalment is the same whether the money buys a shop counter, clears a costlier loan, or sits in the bank. The household afterwards is what differs, and no amount of staring at the instalment reveals which of the three happened. A capital raise works the same way. The use of proceeds is therefore read as its own question rather than as a footnote.

Take the same hypothetical Rs 420 crore three ways. Repay borrowing, and the Rs 240 crore of borrowings goes to nil with Rs 180 crore left over, so cash stands at Rs 492 crore. Spend it on capacity, and cash returns to the published Rs 312 crore while assets are Rs 420 crore larger. The unfinished coatings line, parked since before the raise in capital work in progressAn asset still being built or installed, carried at what it has cost so far and producing nothing until the day it is commissioned. at Rs 118 crore, would be completed with that spending. Or hold it, and cash stands at Rs 732 crore with borrowings untouched.

Earnings per share is Rs 11.03/- in all three. Book value per share is Rs 75.63/- in all three. A raise is not one event with one answer. The per-share arithmetic is identical across every use of the money, so the use has to be read as a separate question.

The balance sheets are not identical, and the differences are worth naming. Net debtTotal borrowings less the cash and investments held against them. A negative figure means the cash is larger than the borrowing. lands at MINUS Rs 492 crore whether the borrowings are repaid or the cash is simply held. The match is not a coincidence: net debt is borrowings less cash, and a repayment moves both by the same rupee. Repayment changes the gross borrowing and the interest bill, not the net figure. The cash went out again on the third path, so spending the money on assets is the one route that leaves net debt back at the published MINUS Rs 72 crore.

The record does not state what remains to be spent on that coatings line, so the whole Rs 420 crore is applied to capacity below rather than assigned to that one asset. A number nobody has is better left uninvented than made tidy.

How does a buyback run the same arithmetic backwards?

A tea stall with five partners buys one partner out, paying him from the till. Four partners are left, the stall earns what it earned, and each remaining share of the earnings is bigger. The till is lighter by exactly what was paid. Nothing about the tea changed. The tea stall has run a buyback, and every term in the identity above simply flips its sign.

Sarvani Coatings Limited, hypothetically, returns Rs 240 crore by buying 0.40 crore shares at Rs 600/-, a premium the record states as 23.5 per cent over the illustrative Rs 486/- and which recomputes to 23.46 per cent. The purchase takes the count down to 23.60 crore. Profit after tax has not moved and stays at Rs 278 crore, so the division on its own lifts earnings per share to Rs 11.78/-, up 1.69 per cent. Net worth drops to Rs 1,246 crore against a smaller count, and since Rs 600/- sits far above the Rs 61.92/- of book value already there, book value per share drops to Rs 52.80/-, down 14.73 per cent.

The balance sheet does the loudest thing. Rs 240 crore of the Rs 312 crore held in cash and investments walks out of the door, leaving Rs 72 crore behind, and net debt therefore swings by that same Rs 240 crore: MINUS Rs 72 crore beforehand, PLUS Rs 168 crore afterwards. A business that was in net cash is geared after one action, and not one line of the profit ladder shows it.

A rising earnings per share after a buyback and a falling one after a raise are the same identity seen from two ends, and neither is evidence of anything at all about how the business is trading. The 1.69 per cent lift here is a pure divisor effect: fewer shares, same profit. Anybody who reports it as an improvement has reported a division.

What does a researcher write down about a raise?

Five entries, and they fit on one card. The shares created or cancelled. The money in or out. Both per-share figures, before and after, sitting on the same line. The use of proceeds, in the words the disclosure used. And the balance sheet that results, the one entry where the same raise stops looking like one event.

Ravindra Setlur, as chief financial officer, would be the source for the third and fourth of those in any real setting, and what he says about the use of proceeds is a disclosure to be recorded rather than a view to be adopted. The researcher records what was said and what it does to the model, and keeps the two separate in the write-up.

Recording both per-share figures together, on one line, is the single habit that prevents almost every error described here. Separate them and the writer will reach for whichever one supports the sentence they had already decided to write. The failure below is exactly that.