Institutional Ownership: What the Register Tells You

A register records what each holder is able to do, and not who the holder is. Three entries carry every one of Anjani Stationers' 4,00,000 shares, and not one entry is a name. On the one listed register in these notes, institutions appear as categories with proportions and share counts, and not a single institution is named on it either.

A register is opened and it has three lines on it. Is that a disclosure or a gap?

The document comes first, and not the word. The word is the part a reader feels is already held, and a definition offered first quietly decides what gets looked for afterwards. So two registers are read slowly below, one private and one listed, and the definitions wait their turn.

Anjani Stationers Private Limited, a company invented for teaching, has issued 4,00,000 ordinary shares of Rs 10 each, fully paid. Those shares come to the Rs 40,00,000/- of share capital sitting on the face of its balance sheet. Say the unit out loud every time that figure is written down: 4,00,000 here counts SHARES. The same business, in the same year, has a rated capacity of 4,00,000 registers a year at the binding stage of its works, which is a count of exercise books and not of claims on a company. One business, one year, one number, two entirely different quantities. The collision between a share count and an output count runs through the whole subject, and the only defence against it is the boring one of naming the unit in the same breath as the figure, without exception.

Three entries carry all 4,00,000 of those shares. The first founding household holds 1,80,000 shares, being 45.0 per cent of the shares in issue, of nominal valueThe value printed on a share when it was first issued, here Rs 10 apiece. The nominal value is a bookkeeping figure fixed at issue, and not what anybody would pay for the share today. Rs 18,00,000/-. The second founding household holds 1,40,000 shares, being 35.0 per cent, of nominal value Rs 14,00,000/-. The outside holder, whose money arrived in an injection some years back, stands against the remaining 80,000 shares, being 20.0 per cent, of nominal value Rs 8,00,000/-. Nothing is pledged and nothing is otherwise encumberedA share is encumbered when somebody other than the holder has a claim over it, most often because it has been pledged to a lender who may take it if a loan goes wrong..

Two of those figures need their division named before they are allowed to travel any further. The 45.0 per cent above is a share of a register and nothing else. This same business separately runs a gross margin held at exactly 45.0 per cent across two years, and a gross margin is a slice of revenue rather than a slice of a share count. And the Rs 8,00,000/- above is the nominal capital standing against the outside holder's 80,000 shares. The same Rs 8,00,000/- also appears in this business in this same year as the binding invoiced by a workshop it holds, as a guarantee given over that workshop's borrowing, and as the whole tax expense for the year. Four amounts, one number, and only the first of them belongs here.

Now the second register, kept by a different invented business set down elsewhere in these notes. Sarvani Coatings Limited is listed, and its published pattern reads promoter and promoter group 52.4 per cent, being 12,57,60,000 shares; foreign portfolio investorsInvestors based outside India who buy shares in Indian listed companies through a registration route made for exactly that, without taking a controlling stake in any of them. 18.2 per cent, being 4,36,80,000 shares; domestic institutionsInstitutions based in India that hold shares on behalf of other people, such as a mutual fund, an insurer or a pension arrangement. The label says where the institution sits, not how big it is. 14.6 per cent, being 3,50,40,000 shares; and retail and others 14.8 per cent, being 3,55,20,000 shares. The four blocks carry all 24,00,00,000 shares in issue, and the four proportions come to 100.0 per cent exactly.

Set the two documents beside each other and something is true of both that almost nobody notices on a first reading. Neither register carries a single name. One carries two households and an outside holder. The other carries four categories, two of which are institutional. Both lists are complete, both are correct, and a reader who came looking for a person to look up leaves either one empty handed.

The sheet of paper pinned up in the lobby of a housing society before a general body meeting works the same way. The sheet lists flat numbers and the votes each flat carries. It is accurate, it is complete, it settles who can carry a motion and who cannot, and it says nothing whatever about who lives in any of the flats. Nobody calls that sheet a failure of disclosure. The sheet was built to answer one question and it answers it.

So what is a register actually a list of, if not people?

A register is a list of entries, and an entry is whatever the document was built to record. That single sentence carries the rest of this guide, so it is worth slowing down on. The register of membersThe book a company keeps of who holds its shares. Every company has one, listed or not, and it is the document that settles who may vote at a meeting and who receives a dividend. exists to settle two things and only two: who may vote, and who gets paid. The register was never built to record who anybody is. Identity is somebody else's job, and a document does the job it was designed for rather than the job a reader happens to want from it.

So one register carries a founding household, then a second founding household, then an outside holder, and stops. The other carries promoter and promoter group, then foreign portfolio investors, then domestic institutions, then retail and others, and stops. Both lists are complete. Both are correct. Both settle every vote at every meeting either company will ever hold. Neither contains a person.

Here is the move that most readers make next, and it is the wrong one. Readers treat the category as a name that has been blurred, as though somebody started to write the holder down and then thought better of it. A category is not less information than a name. A category is different information, and which of the two is needed depends entirely on the question brought to the document.

A name answers questions about a particular somebody. What else does that somebody hold, who are they connected to, and what did they do the last time a matter like this came up. A name is a thing that can be looked up somewhere else, so a name is the key that opens all three. A category answers a different set. Does this kind of holder turn up and vote, or does the paper sit in a drawer. Can it leave at all. If it wanted to leave, how long would that take and what would it have to do. The second set is about behaviour, and a category predicts behaviour rather well precisely because a category describes an arrangement rather than a person.

The consequence runs through everything that follows. The question a register answers is what a holder is able to do, and ability is a property of the holder and the occasion together. The same 6 per cent is a thin 6 per cent on an ordinary matter and something much larger on the one kind of matter where a controlling holder is kept out of the count. The proportion did not move. The occasion moved. So an account of a kind of holder has to be an account of a document, and the document turns out to be the more useful of the two.

And one more thing rests underneath all of it. A kind of holder that is absent from a register is a fact about the company's form before it is a fact about anything else. A company whose shares are not listed anywhere has no market in which an institution could have bought them. Such a company lodges nothing with any exchange. A promoter block, a public shareholding total, an encumbrance column carrying figures, a schedule of movement since last time: every one of those is missing, and not one of them went missing because somebody decided what to disclose.

What does the word institutional actually do on a register?

The word suggests a list of funds with names against them, and a good deal of writing about shareholding quietly implies the same. So the definition comes first, plainly, and then the document.

An institutional holder is a body that holds shares with money belonging to other people. A fund holds for the people who subscribed to it. An insurer holds against the policies it has written and will one day have to pay. A pension arrangement holds for the people who will draw on it in thirty years. In every case the shares stand in the institution's name and the money behind them does not.

Notice what is not in that definition. Size is not in it. The defining feature is that the money is somebody else's and that somebody else can ask for it back. A single wealthy individual holding a very large block is not an institutional holder, however large the block. A small fund holding a modest one is. The plain test needs no document opened at all, and it is a pair of questions: whose money would leave if this holder sold, and would this holder have had any choice about it.

Think of the treasurer of a school committee. The account is in the committee's name, the treasurer signs the cheques, and the treasurer cannot spend a rupee of it on what he personally would prefer, however certain he is that he is right. He answers to the committee, and the committee can call for the money. The treasurer's arrangement is the whole of it, scaled up and dressed in a filing.

And here is the consequence the rest of this guide needs. A holder of somebody else's money answers to that somebody on a clock the register does not print. Redemptions, mandates, review dates, the quarter in which somebody upstairs asks why this holding is still here: none of that appears on the document. The document gives the label, and the label is the reason such a clock can be presumed to exist at all.

What has this business's own register got no room for?

With the three entries already read, the interesting part of Anjani Stationers' register is not what is on it but what is missing: the lines a reader coming from listed company work goes looking for and does not find. There are six of them.

A promoter block, set out and separately totalled, is absent, and these notes have already settled why: the listed sense of that classification has nothing to attach itself to when a company's shares never trade. A public shareholding total is absent too. A company whose shares never trade has no free floatThe part of a listed company's shares that is available to trade in the market, being everything outside the block a controlling holder sits on. A company whose shares do not trade has none. for anybody to total. A periodic filing is absent, no exchange being anywhere in the picture. An encumbrance column carrying figures is absent, nothing having been pledged. A schedule showing movement since last time is absent, there having been no last time. And a market in which an institution could have bought was never there to begin with, so an institutional line is absent.

Where does a reader find the ownership of a company like this, then? In three ordinary places. The company's own register of members. The annual returnA once-a-year filing made to the registrar, covering a company's basic particulars, its shares and the entries against them. Quite distinct from the annual report, and not the accounts either. it lodges with the registrar under the Companies Act 2013. And the share capital note inside the accounts themselves. All three exist, all three are consulted every day by people who do this work, and none of them is a shareholding pattern in the sense a reader trained on listed companies means by the phrase.

Six absent lines are a different reporting obligation and not a withheld disclosure, and that ruling is set down word for word in these notes where this register was first published. Most Indian companies never list at all. So a three entry register is the ordinary shape of the document rather than an odd one, and far more registers are built like this one than like the other.

Where do the actual obligations live? The Companies Act 2013, kept by the Ministry of Corporate Affairs, carries what every company must maintain and lodge. The listing requirements set by the Securities and Exchange Board of India carry what a company whose shares trade must report about how those shares are spread. No level, no proportion, no frequency and no cut-off date is stated here as fact. All of that shifts. The live text comes from the source itself on whichever day the question arises, and that date belongs beside whatever is taken.

Is there a register anywhere in these notes where institutions do turn up?

There is one, and it turns the argument from a refusal into a finding. Sarvani Coatings Limited is a listed business invented elsewhere in these notes, and its published pattern is set out there in full. Taken slowly, one line at a time, each line turns out to carry some things and not others.

Two of its four lines are institutional. Foreign portfolio investors hold 18.2 per cent of Sarvani Coatings Limited, being 4,36,80,000 shares. Domestic institutions hold 14.6 per cent of Sarvani Coatings Limited, being 3,50,40,000 shares. Every one of those figures is published. Every one belongs to Sarvani Coatings Limited and to no other business. And not one of them is attached to an institution.

The finding is plain. The register records that institutions hold, how much they hold, and which kind of institution each block is. The register does not record which institution. There is a proportion, there is a share count, and where a name would sit there is a category.

None of that is a complaint, and it changes what the next move should be rather than whether there is one to make. A category is enough to predict how a block behaves and it is never enough to predict a person. So a reader who wants to know whether this particular fund has held this company before, or whether it has ever voted against this promoter, is holding the wrong document. The register was never asked to carry that answer, and rereading it more carefully will not produce one.

One warning travels with this block and it costs nothing to carry. 52.4 per cent belongs to Sarvani Coatings Limited and 52 per cent belongs to Aravalli Agro Foods, a different invented business set down elsewhere in these notes. Two businesses, one decimal apart, and a reader who lets either figure travel without its business attached has just merged two companies by accident.

What is an institutional holding actually able to do?

There is exactly one reading of that worth carrying, and it is a single sentence with a great deal of work inside it. On the governance reading set out earlier, an institutional investor uses the one lever it holds, the vote, and a fund holding 6 per cent of Aravalli Agro Foods holds an eighth of the votes that count on a related-party resolution where the promoter is excluded.

Read the sentence twice. The interesting part is what did not happen. The fund's holding did not change. The fund held 6 per cent of Aravalli Agro Foods before the resolution was called and holds 6 per cent afterwards. Nobody bought anything and nobody sold anything. The arithmetic underneath the vote changed.

On an ordinary resolutionThe routine kind of motion put to a company's shareholders, carried on a simple count of the votes cast. The level needed to carry one is set in law and is not stated here., every share is counted, and 6 per cent is 6 per cent. On a matter where a controlling holder has an interest of its own, that holder is kept out of the count altogether. The 52 per cent of Aravalli Agro Foods that would otherwise settle the outcome before anybody arrived is not entitled to be counted, so the base becomes the 48 per cent that remains, and 6 of 48 is an eighth. Same shares. Same register. Eight times the weight.

An ability is a property of the holder and the occasion together, and a register prints only the first half of that. This is the sentence to carry away if only one is carried. The proportion is the input. The occasion has to be supplied by the reader, out of knowledge of what kind of matter is coming up. Which is why a register is read against a question rather than read straight through from top to bottom.

The same thing happens on a much smaller scale. A resident holds one vote at a residents' meeting. On the annual accounts that vote decides nothing whatever. The two largest households between them can carry anything they like. On the one motion the chair is not allowed to vote on, the contract in question being with his own brother, that one vote is suddenly worth arguing about. The vote never changed size. The room did.

What does the register permit as the next question, and does the answer change with its shape?

The shape of a register changes which question is worth the time, and answers none of the questions by itself. Ask what each entry is able to do, one entry at a time, rather than what the shape of the whole thing suggests.

Take the three entry register first. Not one of its entries sits above half of the shares in issue, so no entry settles an ordinary matter by itself, and any contested one turns on which two of the three find each other. Which two of the three find each other is the next question, and its shape is worth marking: the question asks about a conversation nobody outside the room can see and was never going to see. With nothing pledged, no lender is in a position to shift a block out from under anybody, so the entire line of enquiry about lenders closes before it opens. And there is no market in these shares at all, so an entry that wanted out would have to find its buyer among the other two entries or bring somebody new in. A change on that register is a negotiation with a document at the end of it, rather than an accumulation or a block trade.

Now the listed register. A single block sits above half, so any question that starts by supposing the remaining holders object has already been answered before the notice of meeting goes out: the other entries cannot carry a motion between them on an ordinary matter. The next question is therefore the narrow one, being the one kind of matter where that block may not vote. Two blocks are institutional, so a large part of what remains gets voted as a matter of standing policy rather than left in a drawer, and the question after that is what those blocks did the last time something like this came up.

The register gives the proportion, the reader supplies the occasion, and the next question is whatever the two together leave open. That is the general form, and it works on any register at all. On a three line register the useful next questions are about arrangements and conversations. On a four block register they are about occasions and behaviour. Both sets are real, both are answerable somewhere, and neither set is the better one to be holding.

What can a register never tell, however carefully it is read?

Three of the limits are already worked at length in these notes where the first register was set down, and they are carried here rather than rebuilt.

Who actually decides stays hidden. The register hands over the arithmetic of a vote, and that arithmetic is a different object altogether from the conversation held before anybody walked into the room. A holder's intentions stay hidden too. Five years without movement looks the same on paper whether the holder is patient or has simply found nobody to sell to, and the wording of the disclosure does not shift by a letter between those two cases. And whether the entry standing on the register is the party who ends up better off stays hidden as well. Separate beneficial ownership requirements exist for exactly that reason.

Now the fourth, and it is precisely the one the opening question walked into. A register cannot say which institution. On the one register in these notes where institutions appear at all, they appear as categories. There is a kind, a proportion and a share count. There is no name, no date of arrival and no history.

So a reader who wants to know whether a particular fund has held this company before, whether it left the last time the price fell, or whether it has ever voted against this promoter, is asking questions the document was not built to answer. The answer was never on the document to begin with, so more care with it produces nothing. The honest output is the name of the document that would answer the question, and not a guess dressed up as a reading. An estimate built from the size of an institutional block looks like work and behaves like a fact, and the next person to pick up the note has no way of telling the two apart.