Series A: What Changes at the First Priced Round

Series A is a priced round led by an outside institution. In Aravalli Learning Systems Private Limited, invented, it set Rs 120.00 a share against a pre-money of Rs 1,20,00,00,000, issued 25,00,000 new shares for Rs 30,00,00,000, and took the total to 1,25,00,000 shares. The round sold 20.0 per cent, so every holder already on the register was multiplied by 0.8.

A priced roundA round in which a price per share is agreed and new shares are issued at it. fixes two numbers and derives the rest. The pre-money divided by the shares already in issue gives the price. The raise divided by the price gives the new shares. Every percentage on the register afterwards is those two divisions worked through. There is far less to argue about inside a round than the vocabulary around it suggests. Two things are left to negotiate: the pre-money, and where an enlarged employee option pool is taken from. The first has a headline, and it gets argued over for weeks. The second has no headline at all, usually occupies one sentence, and on the arithmetic below it moves more of the register than a sizeable change in the pre-money would.

What is a Series A, and what actually changes from the seed?

Picture a tuition centre run by two people out of two rented rooms. When it first needed money, the money came from people who already believed in the two of them: a small early stage fund that writes first cheques, half a dozen individuals who had watched them teach. There was very little arithmetic to do, so nobody spent four months on it. Two years on, the centre has eleven branches, a payroll, a waiting list and a landlord problem. The next cheque does not come from somebody who believes in the two of them. The money comes from an institution that has read the accounts, sent people to sit at the back of three branches, and will put a number on the whole thing in writing before it parts with anything. A change of counterparty, and not the size of the cheque, is what a Series A actually is.

The finance version runs on the same distinction. A seed round is money against a plan and a pair of people. A Series A is money against a business that has begun to work, put in by an institution whose job is to price things for a living, usually with one investor doing the pricing and the rest following. The investor doing the pricing is the lead investorThe investor that sets the price and the terms the others come in on., and its arrival is the structural change. Somebody outside the existing register now sets the number that every holder inside the register is measured against.

The company worked through here is Aravalli Learning Systems Private Limited, invented. Aravalli is one of eighteen investments held by Nilgiri Venture Fund I, a venture capital fund registered as an Alternative Investment Fund with the Securities and Exchange Board of India and holding Rs 1,50,00,00,000 of commitments. Time runs on Aravalli's own clock throughout, a different clock from any fund's. The seed was Aravalli Year 1 and the Series A is Aravalli Year 3.

Why is it called the first priced round when this seed already had a price?

The phrase first priced round is attached to Series A almost everywhere. Aravalli's round does not support it, so the phrase is worth taking apart rather than repeating. Aravalli's seed had a price. The seed price was Rs 30.00 a share, it was agreed, and 20,00,000 shares were issued at it. So the phrase and the case appear to disagree, and the reason they do not is worth setting out.

The phrase carries an assumption inside it. The assumption is that the seed was raised on an instrument that defers pricing: money goes in now, and how many shares that money converts into is settled later, at the next round, on a formula agreed in advance. Where a seed is raised that way, nobody has yet agreed what a share is worth, so there genuinely is no price until Series A arrives. The phrase is exactly right for a seed of that kind, and raising a seed that way is common. Aravalli did not do it, and where a seed was itself priced, the phrase describes a different round from the one at hand.

So what is first at Aravalli's Series A is not the price. The new thing is an outside institution setting it. At the seed, the price was agreed among people who were all about to be on the register together. At the Series A, it is agreed by an institution that was not on the register at all and will only join if the number suits it. The change is real and it deserves its own name, but the name it usually gets does not fit this case, and the arithmetic below contradicts what the phrase leads a reader to expect.

What did the register look like the morning before the round?

Everything a round does happens to a list, and the list is the share register. Aravalli's had four lines on it and 1,00,00,000 shares in total. The two founders held 70,00,000 between them. An employee option poolShares set aside for employees, including ones not yet hired. held 10,00,000. Nilgiri Venture Fund I held 15,00,000, and an angel syndicate of six individuals held 5,00,000.

Two things in that table are worth pausing on. The first is the pool. The pool's 10,00,000 shares belong to nobody yet; they are set aside for employees the company has hired and employees it has not hired. The reserved shares are still counted in the 1,00,00,000. Counting them is what people mean when they say a register is being read fully dilutedCounted as if every reserved share were already issued.. The pool sits inside the count, and every calculation in this guide divides by a count that includes it. The second is the fund's cheque. Nilgiri Venture Fund I put Rs 4,50,00,000 into that seed, and Rs 4,50,00,000 is 3.0 per cent of the fund's own Rs 1,50,00,00,000 of commitments. Rs 4,50,00,000 is not 3.0 per cent of anything else, and naming the denominator every time is the difference between a figure and a number that merely looks like one.

Where does the price of Rs 120.00 come from?

The pre-moneyThe value agreed for the company before the new money arrives. agreed for Aravalli at its Year 3 round was Rs 1,20,00,00,000. The Rs 1,20,00,00,000 is the whole of what was negotiated. The figure is not a fact about the company, it is not the output of a model that both sides ran and agreed on, and nothing anywhere confirms it: it is the number at which one institution was willing to put money in and the existing holders were willing to let it. Everything else in the round is arithmetic performed on that one negotiated figure.

The price per share is the pre-money divided by the shares already in issue. Rs 1,20,00,00,000 over 1,00,00,000 shares is Rs 120.00 a share. The denominator repays attention. The right count is the one before the round, not the one afterwards, and using the wrong one is the single commonest slip on this subject. Dividing the same pre-money by the 1,25,00,000 shares that exist afterwards gives Rs 96.00, a number that reproduces nothing. Rs 96.00 will not give the right number of new shares, it will not rebuild the post-money, and it does not correspond to a price anybody paid.

How many shares does Rs 30,00,00,000 buy at that price?

The second division answers it. Rs 30,00,00,000 divided by Rs 120.00 a share is 25,00,000 shares. The 25,00,000 shares did not exist before and are created by the round, so the total in issue rises from 1,00,00,000 to 1,25,00,000. Nothing was taken from anybody to make them. A round issues new shares; it does not move existing ones.

A build that only reconciles one way is not a build, so check the round in the other direction. The post-moneyThe pre-money plus the raise, which is also the share count afterwards times the price. should be the pre-money plus the raise: Rs 1,20,00,00,000 plus Rs 30,00,00,000 is Rs 1,50,00,00,000. The post-money should also be the share count afterwards multiplied by the price: 1,25,00,000 times Rs 120.00 is Rs 1,50,00,00,000. The two agree at Rs 1,50,00,00,000, and that agreement is the only evidence available that the price and the share count are consistent with each other. And the new investor holds Rs 30,00,00,000 of a Rs 1,50,00,00,000 post-money, or 20.0 per cent, the same 20.0 per cent given by 25,00,000 shares over 1,25,00,000.

What does the register look like once the money lands?

Five lines now, where there were four. The four original holders hold precisely the share counts they held yesterday. The number underneath every one of them is what changed.

56.0 plus 8.0 plus 12.0 plus 4.0 plus 20.0 is 100.0, and the share counts add to 1,25,00,000. Both columns reconcile. A register that does not reconcile in both directions has an error in it somewhere, and no amount of reading will find it. Doing the check by hand once is what turns the next sentence from an assertion into something verified.

Why is every existing holder multiplied by exactly 0.8?

Look down the last two columns of that table. 70.0 becomes 56.0. 10.0 becomes 8.0. 15.0 becomes 12.0. 5.0 becomes 4.0. Every one of them has been multiplied by 0.8, and none of them was singled out, argued over or treated differently from any other. The multiplier is the old total divided by the new total: 1,00,00,000 over 1,25,00,000 is 0.8. The same division is done for everybody on the register, so the same number comes out for everybody on the register.

There is a shorter way to see it. The round sold 20.0 per cent of the company, so 80.0 per cent of it was left for the people who already held it, and 80.0 per cent is 0.8. One division, applied to everybody at once, is the whole of dilutionThe fall in a holder's percentage caused by the total share count rising. at a round like this. Dilution is not a transfer, it is not a penalty, and nobody has lost a share. The denominator got bigger. Both of those figures follow from the numbers chosen in this invented case rather than from any general rule about rounds, and a round that sold a different share of the company would produce a different multiplier.

What happened to the founders, in both halves of the sentence?

Half of dilution is the percentage, and an account that stops there covers half the subject. The founders' holding fell from 70.0 per cent to 56.0 per cent, and their share count did not move at all: 70,00,000 shares before, 70,00,000 shares after. The other half is what those same shares are counted at. At the seed price of Rs 30.00 they came to Rs 21,00,00,000. At the Series A price of Rs 120.00 the identical 70,00,000 shares come to Rs 84,00,00,000. The percentage fell and the count at the latest round price rose, and both of those sentences are true of the same unchanged 70,00,000 shares.

The nature of those rupee figures matters more than the figures do. Neither one is a valuation of anything. Each is a number of shares multiplied by the price of one round, and the price of a round is the number one institution was willing to transact at on one day against a company whose shares nobody can sell on an exchange. Nobody has received Rs 84,00,00,000. The eventual proceeds depend on a sale that has not happened, at a price nobody has yet agreed.

What is the option pool shuffle, and where is the pool taken from?

The finance version is easy to nod along to and hard to feel, so the everyday version comes first. A buyer agrees to purchase a flat for a stated price. Weeks in, the seller mentions that the flat needs repainting before handover, that everybody agrees it needs repainting, and that the repainting will be done out of the agreed price rather than out of the seller's pocket. The price on the agreement has not changed by one rupee. The buyer now receives less for it. Nothing about the size of the repainting job is in dispute; the only question is which side of the agreed number it sits on.

An option pool is the same shape. A Series A investor will often want the pool enlarged, and the reason is ordinary and not in dispute either: a company about to hire twenty people needs shares to offer them, and a pool that is empty is not a pool. The question is not how big it should be. The size is a matter for whoever is doing the hiring. The question is whether the enlargement happens before the price is struck or after it. Enlarging the pool out of the agreed pre-money rather than after the round is what is meant by the option pool shuffleEnlarging the pool out of the agreed pre-money rather than after the round., and everything that follows is one counterfactual worked to the rupee to show what that ordering does.

What does that ordering do to the price per existing share?

Work it through. The pool goes from 10,00,000 shares to 15,00,000, so the pre-round count goes from 1,00,00,000 to 1,05,00,000 shares. The pre-money does not move: it is still Rs 1,20,00,00,000, exactly the number that was negotiated. But the price per share is the pre-money divided by the pre-round count, and the pre-round count just got bigger. Rs 1,20,00,00,000 over 1,05,00,000 shares is Rs 114.29 a share, against Rs 120.00 before. The headline did not move and the price per existing share fell by Rs 5.71, and that fall is the whole of what happens here.

The rest of the round then follows the same two divisions as before. Rs 30,00,00,000 at Rs 114.29 a share buys 26,25,000 shares rather than 25,00,000, and the total in issue afterwards is 1,31,25,000 rather than 1,25,00,000. There is a neat check available here: because the raise is exactly one quarter of the pre-money in this case, the total afterwards is always 1.25 times the pre-round count, and 1.25 times 1,05,00,000 is 1,31,25,000 on the nose.

So who actually pays for an enlarged option pool?

Put the two registers side by side and the answer falls out. In the counterfactual the Series A investor holds 26,25,000 shares of 1,31,25,000, or exactly 20.0 per cent, the same 20.0 per cent it held in the round as it happened. Its number did not move by a decimal place. The pool ends at 15,00,000 shares of 1,31,25,000, being 11.4 per cent against 8.0 before. The founders end at 70,00,000 of 1,31,25,000, being 53.3 per cent against 56.0. The two seed holders together end at 15.2 per cent against 16.0. The buyer is identical in both worlds and the sellers are not, so the pool was carried entirely by the people who were already on the register.

The rounded figures hide a little of the cost, so state it precisely rather than loosely. The founders would need 73,50,000 shares to hold 56.0 per cent of 1,31,25,000, and they hold 70,00,000, so they are 3,50,000 shares short. Unrounded, that shortfall is 2.6667 percentage points. The one decimal figures show it as 56.0 against 53.3, a gap of 2.7. Counted at each round's own price, their same 70,00,000 shares are Rs 84,00,00,000 at Rs 120.00 and Rs 80,00,00,000 at Rs 114.29, a difference of Rs 4,00,00,000. Neither of those rupee figures is a valuation and neither is money anybody has received: each is a share count multiplied by the price of one round.

What tightens at a Series A, read at a structural level?

A term sheetA short document setting out the shape of a proposed investment. is the document a round is agreed on the shape of, and one is read structurally rather than drafted. Feld and Mendelson, in Venture Deals, 2011, put a structural reading ahead of a clause by clause one for the reason the arithmetic above has already shown: how a document is arranged decides which numbers a reader looks at, and the sentence that moved the register at Aravalli carried no figure in the headline block at all.

Read structurally, a reader is looking for four things. The first is what is being priced: the pre-money and the count it is divided by. The second is who is buying: the institution and the amount. The third is what is created before the price is struck, and an enlarged pool lives exactly there. The fourth is what changes hands: the share counts and the rights that go with them. Three of those four sit in the largest type on the term sheet, and the third one does not.

The rest of what tightens at a Series A is real but is not arithmetic. A round of this kind normally creates a class of shares with rights of its own rather than issuing more of what is already there, adds a seat at the board table for the investor putting the money in, fixes a list of decisions the company cannot take without that investor agreeing, and sets out what information the investor gets and how often. None of those is a number, none of them appears on the register, and all of them together are why the arrival of an outside institution is a structural change rather than simply a larger cheque.

How does somebody on a private capital desk actually use this?

Three uses, and none of them is exotic. The first is the most common question an analyst gets asked about a round the fund is not putting money into: where does the fund's percentage land? If the round sells 20.0 per cent, every existing holding is multiplied by 0.8, so Nilgiri Venture Fund I goes from 15.0 per cent to 12.0 without anybody needing a model. The multiplier answers that question in one line, and the only way to get it wrong is to have used the wrong share count.

The second is the reporting and operations side. Somebody there has to hold the register afterwards and reconcile it against what the documents say. Reconciling needs the pre-round count the price was computed on, the price, the new shares, and the fully diluted total including the pool. Every one of those four is checkable against the others, and the check is the one done above: the count afterwards times the price has to rebuild the post-money. A register that fails that test disagrees with the round that created it.

The third is diligence on a company that has already done a round. Given a term sheet or the record of a past round, the single most useful thing to find is not the pre-money. The share count the pre-money was divided by is the more useful figure, and that number silently carries the pool decision. Two rounds with the same headline pre-money and different pre-round counts are different rounds, and the register is the only place the difference shows.