Private Markets: What Sits Outside the Listed World

A market is private when there is no quotation, no continuous price, no obligation to publish and no way to transfer without somebody's consent. Everything else follows from those four absences: the number on the statement is an estimate rather than a price, the exit has to be arranged, and the information arrives because a contract requires it rather than because a rule does.

Almost everybody arrives at this subject holding the wrong picture. The picture is that there is a set of things called private assets, sitting in a separate box, with their own personality and their own weather. Unlisted companies go in the box. Buildings go in the box. Loans made outside a bank go in the box. Then the box gets a label, and the label gets treated as though it described a kind of thing.

It does not. Private describes how a holding is bought, held and priced, and not what the holding is. The same company, the same office block and the same loan can sit on either side of the line, and can cross it, and nothing about the underlying object needs to change for that to happen. The machinery around it changes: who forms the price, who is allowed to buy, how and when anybody can leave, what information exists and who is entitled to see it, and how the holding is valued in the long stretches between one transaction and the next.

Every one of those five is a mechanical fact. Each can be checked by reading a document or looking at a screen, and none of them requires an opinion about whether anything is good. Every strategy that lives inside a private marketA market with no quotation, no continuous price, no publication duty and no free transfer. inherits whatever the machinery has already settled, so the five mechanical questions come first. A reader who has the mechanism can pick up any structure, in any country, in any decade, and ask the five questions. A reader who only has the box will keep meeting objects that refuse to stay in it.

One invented fund carries the worked figures throughout. Nilgiri Growth Partners Fund II, an invented private fund, is managed by Nilgiri Alternatives Advisors Private Limited, sponsored by Nilgiri Financial Holdings Private Limited, and its assets are held by a trustee, Nilgiri Trusteeship Services Private Limited. Rs 5,00,00,00,000 was committed to it by twelve investors and the manager between them. The fund bought nine companies. Every figure attached to it is as at the end of its Year 9 Quarter 2, the record date the whole worked case uses.

What makes a market private, exactly?

The four absences are easier to see against a listed market, so a listed market is the place to start. A quoted share carries a number that somebody is publicly willing to deal at, right now. The quoted share carries that number again a second later, and again all day. The share comes with a company that must publish certain things to everybody at the same time, whether it wants to or not. And it can be sold without asking anybody's permission: a button is pressed, somebody on the other side takes it, and the holder is out.

With those four removed, one at a time, what is left is a private market. Not a different kind of business. A different set of arrangements around the same kind of business.

Read the table once more and notice what is not in it. There is no mention of size, of sector, of age, of risk, of how fast anything grows or of who is running it. A hundred year old cement works and a two year old software business can both be private, and either of them can be listed. The line is drawn through the arrangements, not through the assets.

If nothing is quoted, where does a price come from at all?

An ordinary flat makes the point, the one a household actually lives in. Nobody has ever told its owner what it is worth today. There is no screen carrying its number, no closing level for it in tomorrow's paper, and no stranger standing at the gate willing to buy it at four in the afternoon. Figures get heard, of course. The broker who called last month had one. The neighbour who sold two floors down had another. Neither of those is a price. A price is what somebody actually paid, and for that flat the number will exist on exactly one day: the day it is sold.

The flat shows the first absence in its most ordinary form. In a listed market a crowd of strangers is publicly willing to deal and their willingness is displayed, so a quotationA price at which somebody is publicly willing to deal. stands there whether anybody wants it or not. In a private market nobody is standing there. A price in a private market is not hidden, not lagging and not being got wrong by anybody: it has simply not been made yet. That distinction sounds pedantic until somebody argues that private marks are wrong, at which point it becomes the whole argument. Nobody can be wrong about a number that has never been made.

So what does exist? Three quite different number-making events, and it is worth keeping them apart because readers routinely merge them into one and then wonder why the numbers disagree.

The first is a transaction. Two parties negotiate, they agree, money moves, and for that one moment a real price exists for that one thing. In Nilgiri Growth Partners Fund II this happened five times across the whole nine years to its record date at Year 9 Quarter 2. Five. The second is a quarterly restatement of value made by the manager. The third is an annual valuation made by an independent party appointed for the purpose. In the worked case that party is Palani Valuation Advisors LLP, a limited liability partnership (LLP), with Kolar Fund Services Private Limited acting as administrator and striking the fund's total. How any of those valuations is actually produced, on what basis and with what judgement, is covered separately. Who makes the number, and on what timetable, is what settles what kind of number is in hand.

A reader who holds those three apart will never again be confused by a sentence like the one this invented fund could truthfully write about itself: it reported a value at the end of every one of thirty four quarters, and on five occasions in that entire stretch somebody actually paid it money for something. Both halves of that sentence are ordinary. Neither is a scandal. The reporting and the paying are simply two different machines running at two different speeds.

If there is no price, what is the number on the statement?

An investor in a private fund still gets a statement, and the statement still carries a figure. It has to. Somebody has to tell the pension pool and the insurance company what their money is currently said to be worth, and telling them nothing is not an option. So a number is manufactured, on a stated timetable, by named parties, under a policy the fund's own documents set out. The manufactured number is a carrying valueThe estimate at which an unrealised holding sits on a fund's books., and each time one is struck at a stated date it is called a markA carrying value at a stated date, produced on a timetable..

A carrying value is a different kind of number from a price, and the difference is not that one is better than the other. A price is the output of a negotiation that actually happened. A carrying value is the output of a procedure that runs whether or not anything happened. Both are honest. The two numbers answer different questions. The price answers what somebody paid. The carrying value answers what the appointed parties, following the fund's policy, currently estimate.

Here is what that produces in the worked case. Nilgiri Growth Partners Fund II reports a total valueCash already returned plus the estimate of what is still held. of Rs 7,20,00,00,000 as at the end of its Year 9 Quarter 2. The reported total is not one thing. Rs 4,38,00,00,000 of it is cash that has already gone back to investors in four distributions, and that part is settled beyond argument: it arrived in bank accounts. The remaining Rs 2,82,00,00,000 sits across five holdings that have never been sold to anybody, and every rupee of it is an estimate. The estimated part is 39.2 per cent of the reported figure. Nearly two fifths of what this invented fund is currently said to be worth rests on a mark rather than on a receipt.

Notice how the five holdings are spread. One of them, holding 4 in Bhavani Speciality Chemicals Private Limited, carries 38.3 per cent of the whole estimate by itself. One of them, holding 6 in Vaigai Edutech Private Limited, is carried at Rs 21,00,00,000 against the Rs 30,00,00,000 it cost, below what was paid for it. So the estimate is neither uniformly optimistic nor uniformly cautious; it is five separate judgements made under one policy. The five shares print as 38.3, 7.4, 13.8, 28.7 and 11.7 per cent, and that adds to 99.9 rather than 100.0. Nothing has been nudged to fix that. The unrounded shares sum to exactly 100.0 per cent and the printed column is simply what rounding does to five numbers.

What happens to the estimate when a real price finally arrives

The worked case gives one clean instance and it is worth sitting with. Holding 3, in Tungabhadra Logistics Private Limited, cost this fund Rs 60,00,00,000. At the end of Fund II's Year 7 it was carried at Rs 1,20,00,00,000, being twice its cost. In Year 8 Quarter 1 it was actually realised, and the total proceeds were Rs 1,50,00,00,000, being two and a half times its cost. The reported figure for that one holding moved by half a turn on the day a buyer finally paid, and nothing about the business itself changed on that day.

The gap between a mark and a realised price has a name in this subject, the valuation lag. A private estimate is built from information that arrives quarterly and annually. A quoted price is built from information that arrives continuously. The lag runs in both directions, too: holding 6 was carried at what it cost right through Year 5 and was written down to Rs 21,00,00,000 only at Year 6, three quarters after the deterioration had first become visible. The lag is a property of the measurement timetable and not of the business being measured.

The temptation to draw two conclusions here is strong, and both of them are wrong. The lag does not make private marks wrong. And a number that only moves four times a year does not describe something calmer, steadier or smoother than a number that moves all day. A measurement interval is not a property of the thing being measured. How often something is looked at changes the record of it and changes nothing at all about it, and a claim in either direction has no basis behind it.

Where does the information come from, if nobody has to publish it?

Private information surprises people in the opposite direction from the one they expect. The instinct is that private means dark: that because there is no obligation to publish, there is less to see. Sometimes that is true for an outsider. For the person actually holding the position it is very often the reverse, and the mechanism that makes it so is worth understanding exactly.

Public disclosure is a rule of general application. The rule reaches everybody at once, it is the same for everybody, and nobody negotiates their own version of it. General application is the great virtue and also the ceiling: the rule delivers what it says and no more, at the moment it says, and delivers exactly that to every other person in the country as well.

Private information does not work like that at all. The information arrives because a contract says it must. In the worked case, each of the nine companies Nilgiri Growth Partners Fund II bought into sits under a shareholders agreementThe contract among a company's shareholders, which is where private information rights live., and that agreement fixes what the company sends the fund and how often. The contracted arrangement includes a monthly pack and a board meeting every quarter. The rest of that arrangement, item by item, is covered separately. The information exists because it was negotiated before the money went in, not because a rule of general application requires it.

Then the same thing happens one level up. The fund itself sits under its own constitutional documents and a contribution agreement with each investor, and those fix what the fund sends its investors: a report each quarter, an audited one each year, a notice for every capital call and every distribution. Again, negotiated rather than imposed. Again, reaching a defined list of people and nobody else. The full list of what an investor of such a fund receives is covered separately.

The consequence of that mechanism is the one readers underestimate. If information arrives by contract, then whoever did not negotiate for it does not get it, however reasonable the request. An investor who signed without securing a particular right cannot later insist on it because it seems fair. A person outside the list gets nothing at all, not because anything is being concealed but because no duty ever ran in their direction. The documents therefore matter enormously in private markets and hardly at all to somebody holding a quoted share: in one arrangement the terms are the product, and in the other the terms are the same for everybody and the price is the product.

How does anything change hands at all?

On an exchange, selling is one action. A decision, a press, and it is done, and the person on the other side is a stranger the seller will never identify. The whole apparatus exists to make that single action possible, and its success is measured by how little thought it demands.

In a private market that single action becomes four separate things, each of which has to succeed, and each of which can fail. First somebody has to be found who wants the thing at all. Second a number has to be agreed with no quoted price sitting there to anchor the argument. The number is genuinely argued rather than accepted. Third the people whose consentThe agreement of another party that a transfer requires. the contracts require have to give it. Only then does the position move.

The consent step is the one that has no equivalent anywhere in a listed market, and it changes the character of the whole arrangement. Under this invented fund's own documents an investor cannot simply pass its interest to somebody else: the manager's agreement is required. Under the shareholders agreement at each of the nine companies, the fund's own stake is similarly fenced. None of this is unusual and none of it is a defect. Consent is the direct consequence of a market where the counterparty is identified rather than anonymous: if it matters who the other shareholders are, then it has to matter who is allowed to become one.

A fifth property follows straight from consent, and it is rarely named. In a private market it also matters who may buy in the first place. Nilgiri Growth Partners Fund II has twelve investors, and they are institutions and pools rather than individuals: a domestic life insurance company, a development finance institution, an overseas sovereign-linked investor, a bank's treasury, a fund of funds, a founder's own investment office, a corporate treasury, a domestic pension pool, a university endowment, a charitable trust, a co-investment vehicle for the manager's own senior staff, and a feeder vehicle standing in front of twenty-two individuals who reach the fund through that structure rather than directly. Twelve names, and the twelfth is a door rather than a person.

Who is permitted to be admitted to a registered vehicle of this kind in India, and on what conditions, is set by the Securities and Exchange Board of India at sebi.gov.in. The conditions change, and the current text at that source is the one to read. The restriction itself, whatever its current terms, is one of the things that keeps a market private. An arrangement anybody may join and anybody may leave has already stopped being one.

What does a private investor get that a public one does not?

Put the four absences together and they look like a list of things taken away. Something is handed back in exchange, and it is not a better return. A private investor gets a different relationship with the terms.

In a quoted share the terms are already fixed, identical for every holder, and not open to discussion. The price is the only variable a buyer controls. In a private vehicle the sequence runs the other way round: the documents come first, they are read before any money moves, and some of the terms are genuinely negotiable before signature. In the worked case six of the twelve investors in Nilgiri Growth Partners Fund II hold a side letter recording something specific to them, and what those letters contain is covered separately. One investor sits on the fund's investor advisory committee. The private investor's real privilege is not information about the future but access to the contract before becoming bound by it.

Two vocabularies run side by side in this subject, and somebody meeting both will otherwise think something has gone wrong. All six of the invented Nilgiri vehicles are settled as trusts under an indenture of trust, with Nilgiri Trusteeship Services Private Limited as trustee and Nilgiri Alternatives Advisors Private Limited as investment manager. There is no limited partnership here and no general partner as a matter of Indian law. And yet the economics were designed in the imported global form and carried across, so the documents and the investors use the vocabulary of limited partner, general partner, capital account and carried interest throughout. So both are true at once: the general partner's role is discharged by the manager and the trustee between them, and the binding contract is a trust deed and a contribution agreement rather than a partnership agreement.

How often did a price actually exist in one invented fund?

The result of the arithmetic is more startling than any argument about it could be, so the arithmetic is worth doing slowly.

Nilgiri Growth Partners Fund II ran from its final close to its record date at the end of Year 9 Quarter 2. The stretch is 8.50 years, or thirty four quarters. The fund reported a value at the end of every one of those quarters, and held nine companies across them. Exactly five events in the whole period saw a position leave the portfolio in whole or in part, and those five are the only moments at which a price for any part of it existed. Each of those events is a realisationThe event in which a holding actually becomes cash., and outside those five moments every figure the fund published about its holdings was produced rather than paid.

Look at where the five sit. Not one of them falls in the first five years. Twenty two consecutive quarters passed in which this invented fund reported a figure every quarter and nobody paid it anything for anything. Then four of the five arrive inside eight quarters, and the fifth of them, at Year 6 Quarter 4, is a write-off producing nil. So the price events are not merely rare, they are clustered, and their clustering is a feature of how positions are actually sold rather than of how the businesses performed.

This fund can be counted three ways, and only one of the three gives five. Four holdings are gone entirely: holdings 1, 2 and 3 sold and holding 5 written off. Five holdings are still held: 4, 6, 7, 8 and 9. Four plus five is nine, the whole portfolio. Holding 9 was 40 per cent realised and the other 60 per cent is still held, so there are five exit events across five holdings. Holding 9 is the only position sitting on both sides of that count, and any sentence using the number five has to say which five it means.

Can the same business sit on both sides of the line?

One of this fund's holdings did exactly that, and the crossing is the cleanest illustration of why private is a description of mechanism rather than a description of assets.

Holding 3 is Tungabhadra Logistics Private Limited. Nilgiri Growth Partners Fund II entered it at Fund II's Year 2 Quarter 2 for Rs 60,00,00,000. For five and three quarter years it was an unlisted company: no quotation, no continuous price, no publication duty of general application, and a stake nobody could move without the agreement of the people the shareholders agreement named. Then at Year 8 Quarter 1 it made an initial public offering. Its shares began to be dealt on an exchange, and the fund realised its position in two stages, Rs 60,00,00,000 at the offering itself and Rs 90,00,00,000 from selling what remained once the lock-in on those shares had ended, being Rs 1,50,00,00,000 in total.

Now ask what actually changed about the company on the day of the offering. The lorries were the same lorries. The warehouses were the same warehouses. The contracts with customers, the people, the debts and the prospects were all exactly what they had been the previous week. All four of the absences were filled at once, so the company moved from one side of the line to the other without one fact about the business changing. That is the proof, in a single transaction, that private was never a property of the business.

The line is not even a clean line for everybody at the same moment, and the fourth row of that comparison shows why. On the day of the offering most holders of that company could deal freely. The fund could not. Its remaining shares stayed locked until the lock-in attaching to them ended, and that is why its Rs 1,50,00,00,000 arrived in two stages rather than one. So for a stretch the company sat on the listed side while this particular holder's own position still behaved in the private way: it could not be moved at will. The conditions attaching to such sales in India are set by the Securities and Exchange Board of India at sebi.gov.in, and they change.

The traffic runs the other way too, and just as ordinarily. A listed company whose shares are bought up and taken off the exchange becomes, on that day, exactly the kind of object described here: no quotation, no continuous price, disclosure by whatever the new arrangements require, and no transfer without agreement. Nothing about the building, the machines or the staff moves an inch.

What does four years of an unchanged number actually show?

There is one mistake worth preventing above all the others, and it is not an exotic one. The mistake is the ordinary, comfortable misreading that a flat reported figure is evidence of a calm business.

The third row of the table is the one that closes the argument. In each of those four years the gap between what investors had paid in and what the fund reported was exactly the management fee and the fund expenses drawn to that date and nothing else: Rs 13,10,00,000, then Rs 23,70,00,000, then Rs 34,30,00,000, then Rs 44,90,00,000. The gap is not a performance figure of any kind. The gap is the arithmetic of a vehicle that has drawn money for its own running costs and has not yet revalued anything it bought. The mechanism of the fee itself, its rate and the way its basis changes over a fund's life, is covered separately.

What does this cost, and who pays it?

Everything described so far has to be built by somebody. A quoted market hands over a price, a disclosure regime and an exit for free, in the sense that the machinery producing them is not billed to anybody line by line. A private arrangement has no such machinery lying around, so it has to be constructed, staffed and paid for, and the bill lands somewhere.

In the worked case the arithmetic is plain. Nilgiri Growth Partners Fund II has drawn Rs 4,80,00,00,000 from its investors to the record date at Year 9 Quarter 2. Of that, Rs 4,00,00,00,000 bought the nine companies. The other Rs 80,00,00,000 bought no company at all: Rs 70,20,00,000 of management fee and Rs 9,80,00,000 of fund expenses. The fee and the expenses are what the machinery costs. The bill covers the administrator that strikes the fund's total, the independent valuer that values each unsold holding annually, the auditor, legal, custody and the investor advisory committee's own costs, and it includes Rs 2,50,00,000 of organisational expenses drawn at formation. A private market has to manufacture, and pay for, the price discovery and the disclosure that a listed market produces as a by-product. How the fee itself is calculated, at what rate, on what basis and how that basis changes over a fund's life is covered separately.

Who pays is the more interesting half of the question, and the answer has a sting in it. The investors pay, and they pay through exactly the same capital calls that buy the companies. The money arrives as capital drawn, not as a deduction from a price anybody watches, and no statement ever shows a subtraction happening. The cost is therefore borne whether or not it is noticed. The same Rs 7,20,00,00,000 of total value also reads two different ways depending on the denominator used: 1.80 times against the Rs 4,00,00,00,000 the holdings cost, and 1.50 times against the Rs 4,80,00,00,000 the investors actually paid in. Both are true of the same fund on the same day. The full arithmetic of multiples and which denominator each of them uses is covered separately.

One investor makes it concrete. Investor 1 is a domestic life insurance company that committed Rs 1,00,00,00,000, being 20.0 per cent of the fund's total commitments of Rs 5,00,00,00,000. Because every investor is drawn strictly pro rata, it has paid in 96.0 per cent of its commitment.

Two cautions on that table, both of which catch careful readers. First, investor 1's unfunded commitment of Rs 4,00,00,000 is its own figure, derived as Rs 1,00,00,00,000 less Rs 96,00,00,000. The fund's unfunded commitment is a different number entirely, Rs 20,00,00,000, being Rs 5,00,00,00,000 less Rs 4,80,00,00,000, and investor 1's 20.0 per cent share of that fund-level figure is exactly the Rs 4,00,00,000 in the table. The two figures are five times apart and both are correct at their own level, and that is precisely why they get confused. Always name the denominator. Second, the line that is not on any statement and probably should be: what the Rs 56,40,00,000 is an estimate of. Five holdings, one of them carried below its cost, and not one of them sold to anybody.