GP vs LP: Who Decides, Who Funds, and Who Holds

One side puts up almost all the money and decides almost nothing. The other side puts up a small fraction of it, makes every investment decision, and is paid a fee for doing that and a conditional share of profit. In Nilgiri Growth Partners Fund II, invented, twelve investors committed Rs 4,90,00,00,000 and the manager Rs 10,00,00,000, which is 98.0 per cent against 2.0 per cent.

Three questions separate these two sides and nothing else needs to. Who decides. Who funds. Who bears the loss if it goes wrong. All three are written into the fund's own documents before a rupee moves, so all three have mechanical answers on any particular fund. Every other difference anybody draws between the two words follows from those three. Whether the split between the two sides is fair is a judgement of a different kind. A reader who has those three answers straight is far better placed to make that judgement than any account of the mechanism could be.

The arrangement is much older than the vocabulary, and it appears at household scale first. Thirteen people in a neighbourhood decide to back a wedding catering business. Twelve of them put in money and go back to their own jobs. The thirteenth puts in money too, a much smaller amount, and then does all of the work: choosing which weddings to bid for, hiring the cooks, buying the rice, deciding when to buy a second van and when to sell the first one. The twelve are told what is happening. The twelve are not asked. The twelve agreed to that when they handed over the money, and they agreed to it in writing, and the whole reason they handed the money to the thirteenth person rather than doing it themselves is that they did not want to choose the weddings. A private fund is that, with a long written contract instead of a conversation.

Which of these two words names a job, and which names a position?

The two words are not the same kind of word, and a reader who treats them as a matched pair gets confused later. Settle the difference before anything else. General partnerThe imported term for the side that makes the investment decisions. names a job somebody does. Somebody raises the fund, decides what it buys, sits on the boards of what it bought, decides when to sell, and answers for all of that. Limited partnerThe imported term for the side that provides the capital. names a position somebody holds. A pension pool, a charitable trust and a person putting money in through a feeder are all in that position, and being in it involves doing nothing at all on most days.

The difference between a job and a position is why one word describes a small number of people who work full time at it, and the other describes a very mixed collection of institutions with almost nothing in common except the position they hold. In Nilgiri Growth Partners Fund II, invented, that position is held by twelve investors, and the list runs from a domestic life insurance company at Rs 1,00,00,00,000 down to a co-investment vehicle for the manager's own senior staff at Rs 5,00,00,000. An insurer and a staff vehicle have nothing in common. Neither of them chooses a single company the fund buys.

The worked instance throughout is that fund. Nilgiri Growth Partners Fund II, invented, is a closed-end growth and buyout fund with total commitments of Rs 5,00,00,00,000, managed by Nilgiri Alternatives Advisors Private Limited, invented. The record date for every figure is the end of its Year 9 Quarter 2. Twelve investors have committed Rs 4,90,00,00,000 between them and the manager side has committed Rs 10,00,00,000. Those two sums make Rs 5,00,00,00,000 exactly, and that addition is the first arithmetic check a reader can run on any fund at all.

Who decides what this fund buys, and what does the other side get instead?

The manager decides. Not the largest investor, not the seven investors who sit on a committee, not all twelve of them together. In Nilgiri Growth Partners Fund II, invented, every investment and every realisation of one is approved by an investment committee of five members, four of them from the manager and one external to it. Nine companies went into that portfolio through that committee and four positions have come back out of it the same way. No investor voted on any of the nine.

The investor side gets a consent rightA right to agree or refuse one named thing, short of deciding it. instead, a different power and a much narrower one. A consent right is a right to agree or refuse one named thing. A consent right is not a right to choose, and it does not reach anything that is not on the list. In this fund an investor advisory committee of seven members, drawn from investors 1, 2, 3, 4, 5, 6 and 8 and chaired by Meera Sathe for investor 1, consents on four matters: conflicts, valuation policy, the first extension of the term, and any change to the investment policy. Not one of those four is an investment, and that committee can neither approve an investment nor reject one.

The seven investors on that committee committed Rs 4,25,00,00,000 between them, 85.0 per cent of the Rs 5,00,00,00,000 of total commitments and 86.73 per cent of the Rs 4,90,00,00,000 committed by investors. A group holding that much of a fund still cannot stop the manager buying a company, and that is not an oversight in the drafting. The twelve signed that arrangement, and it is the reason the manager can move in a competitive process without going back to thirteen parties for a vote.

The four matters are not trivial, and a reader who hears the word consent as a polite formality has swung too far the other way. Be very literal about what a consent right does reach. Consent on conflicts means the manager cannot quietly sit on both sides of a transaction. Consent on valuation policy means the method by which unsold positions are marked is not the manager's alone to change. Consent on the first extension means the fund cannot simply carry on past its contracted term because carrying on suits the manager. The three consents above are real teeth. The teeth sit on a fixed list, and the list does not include which company to buy.

Who funds it, and is the manager exempt from a capital call?

The manager is not exempt, and this is the single most missed fact about the pair. A great many readers picture the manager as the party that contributes work rather than money, and in this invented fund that picture is wrong by Rs 10,00,00,000. The manager commitmentMoney the manager itself puts into the fund alongside investors. here is Rs 10,00,00,000, being exactly 2.0 per cent of total commitments, funded in cash, and it takes the same treatment as an investor interest at every stage of the payout. The commitment is real money and it is called with everybody else's.

Every call this fund issues goes to all thirteen parties pro rataIn proportion, so each party pays the same percentage of its own commitment., meaning each party pays the same percentage of its own commitment. Take drawdown 2 of Nilgiri Growth Partners Fund II, invented, a call of Rs 55,00,00,000 in Year 1 Q3 to pay for holding 1. Rs 55,00,00,000 against total commitments of Rs 5,00,00,00,000 is exactly 11.0 per cent. Investor 1 promised Rs 1,00,00,00,000 and therefore pays Rs 11,00,00,000. The manager promised Rs 10,00,00,000 and pays Rs 1,10,00,000 under the identical rule, with no exemption anywhere in the notice. The rate is a property of the call, not of who is receiving it.

The aggregate check is the satisfying one. At the record date this fund has drawn Rs 4,80,00,00,000, or 96.0 per cent of commitments. The same 96.0 per cent applied to the investor side gives Rs 4,70,40,00,000. Applied to the manager side it gives Rs 9,60,00,000. The two shares add to Rs 4,80,00,00,000 exactly, with nothing left over. The sums close only because the manager was called on the same terms as everybody else the whole way through. If the manager had been exempt, the sums would not close.

There is a trap sitting directly under all of this and it catches careful readers rather than careless ones. The manager is the thirteenth party. The manager is not one of the twelve investors. So a share of the investor side and a share of the whole fund are different numbers, and a statement or a report that quotes one without saying which has not finished its sentence. The manager's Rs 10,00,00,000 is 2.0 per cent of the Rs 5,00,00,00,000 of total commitments and 2.0408 per cent of the Rs 4,90,00,00,000 committed by investors. Both are correct. The two percentages answer different questions.

The same thing happens to the capital call. Rs 55,00,00,000 is 11.0 per cent of total commitments and 11.2245 per cent of investor commitments. A rate with no denominator attached has not said what it is a rate of, so the denominator has to be named every time. This is not pedantry: the two rates here differ by about two per cent of themselves, and on a fund report that difference is exactly large enough to look like an error somebody made rather than a question somebody did not ask.

Why does the manager put its own Rs 10,00,00,000 in at all?

So that some of its own money sits alongside the investors' money, on the same terms, and moves the same way. In Nilgiri Growth Partners Fund II, invented, the manager side's Rs 10,00,00,000 takes the same treatment as an investor interest at every stage of the payout. The money is not paid first, it is not paid on better terms, and it does not escape a loss that everybody else takes.

The record date shows what that produces. The manager side has paid in Rs 9,60,00,000 on the same 96.0 per cent as everybody else. Its 2.0 per cent share of the Rs 4,38,00,00,000 this fund has distributed is Rs 8,76,00,000, and its 2.0 per cent share of the Rs 2,82,00,00,000 of value not yet sold is Rs 5,64,00,000. The two shares add to Rs 14,40,00,000 of total value against Rs 9,60,00,000 paid in, or 1.50 times. Investor 1, on Rs 1,00,00,00,000, reads exactly the same 1.50 times: Rs 87,60,00,000 of distributions plus Rs 56,40,00,000 of unsold share against Rs 96,00,00,000 paid in. Same fund, same day, same multiple, and the commitment exists to produce that match.

Whether 2.0 per cent is enough of the manager's own money for that purpose is a judgement of a different kind. Its answer varies with who is asking, how large the manager is, where the Rs 10,00,00,000 itself came from, and what else the same people have at stake. A reader can find the number, name its denominator, and ask the next question.

Whose obligation has a ceiling written into it, and whose does not?

Risk is the third axis, and risk is the reason the pair of words exists at all. Everything else about the arrangement could be written into an ordinary contract between equals. The asymmetry of what each side carries could not, and that is what the two words were coined to name.

In the imported structure the vocabulary comes from, a limited partner's obligation is limited to what it has agreed to put in, and a general partner's is not capped at a sum it promised in advance. The ceiling is the difference, and every other feature of the pair hangs off it. Nobody accepts an obligation with no ceiling on decisions somebody else makes, so the side that carries the uncapped obligation is the side that gets to decide. A wider set of consent rights would start to look like deciding, and deciding is precisely what the ceiling was granted in exchange for not doing, so the side whose obligation has a ceiling accepts a narrow set.

Nobody's legal position should be taken from a description of the mechanism. Nilgiri Growth Partners Fund II, invented, is a trust, not a limited partnership. Each contributor's promise is fixed by its own contribution agreement, and for investor 1 that promise is Rs 1,00,00,00,000, of which Rs 96,00,00,000 has been called and Rs 4,00,00,000 remains callable. The reader who wants the actual position of any actual party reads the actual instrument.

One arithmetic trap sits right here and it has caught three separate readings of this fund. There are three different unfunded figures on the same day and all three are correct at their own level. The fund's unfunded commitment is Rs 20,00,00,000, being Rs 5,00,00,00,000 promised less Rs 4,80,00,00,000 called. The investor side's unfunded is Rs 19,60,00,000. The manager side's is Rs 40,00,000. And investor 1's own is Rs 4,00,00,000, being Rs 1,00,00,00,000 less Rs 96,00,00,000. The last two add back into the first, Rs 19,60,00,000 plus Rs 40,00,000 making Rs 20,00,00,000, and quoting any of them as though it were another is the commonest arithmetic error in this whole subject.

What is each side paid, and has either side actually been paid it?

The manager side is paid two things, and how each one is calculated is a large subject covered separately. The manager receives a management fee, payable under the contract whatever happens to the investments. The manager also receives a conditional share of profit called carried interest, and that share arrives only if the fund clears a threshold its own documents fix.

Here is the state of both at the record date of Nilgiri Growth Partners Fund II, invented, at the end of its Year 9 Quarter 2. Management fee drawn: Rs 70,20,00,000. Carried interest paid: nil, not one rupee. Nil carried interest is not a fund in trouble and not a manager being generous. This fund is nine years old and has distributed Rs 4,38,00,00,000 against Rs 4,80,00,00,000 drawn. Nine years in, the fund has not yet finished returning the capital it called, and until it does, the conditional payment is not due. One payment has been made and the other has not, and both facts are the arrangement running exactly as written.

The investor side is not paid at all in the sense the manager side is. The investor side receives distributions, meaning its own money coming back plus whatever the fund made on it. Nilgiri Growth Partners Fund II has distributed Rs 4,38,00,00,000 in four payments across Years 6 to 8, and it carries a further Rs 2,82,00,00,000 of value in five positions that have never been sold to anybody. The Rs 2,82,00,00,000 is an estimate of what unsold things are worth, not cash anybody has received, and a reader who treats the two as the same kind of number has misread the whole report.

Why does the Indian version of this pair have three parties, not two?

Because the vehicle is not a partnership. Nilgiri Growth Partners Fund II, invented, is settled as a trust under an indenture of trust, the form an Indian pooled private vehicle most commonly takes. A trust separates two things a partnership keeps together: holding the assets and deciding what to do with them.

So there are three parties where the borrowed vocabulary has two. The trusteeIn the Indian form, the party that holds the assets., Nilgiri Trusteeship Services Private Limited, invented, holds the assets of the fund and has duties to the beneficiaries, and makes no investment decision at all. The investment managerIn the Indian form, the party that makes the decisions., Nilgiri Alternatives Advisors Private Limited, invented, makes every investment decision and holds none of the assets it decides about. The sponsorThe party standing behind the manager., Nilgiri Financial Holdings Private Limited, invented, stands behind the manager and carries the manager side's own Rs 10,00,00,000 of commitment.

And yet everybody involved still says general partner and limited partner. The habit is not laziness. The economics of this arrangement, the commitment, the call, the fee, the conditional profit share and the order in which money comes back, were designed in the partnership form somewhere else and imported wholesale into a trust, and the vocabulary came with them because it names the economics rather than the legal form. So the honest sentence is that in this fund the role the words general partner name is discharged by the manager and the trustee between them, and what an investor actually signs is a trust deed and a contribution agreement rather than a partnership agreement. There is no limited partnership here, and no single party corresponds to the imported phrase.

What goes wrong when a reader treats this pair as a chain of command?

What is worth asking in one hour with a private fund's documents?

Far more people read a private fund's documents than ever sign one. An analyst at an institution writing a paper for an investment head, somebody sitting on an advisory committee for the first time, a treasury officer working out what a commitment obliges the treasury to do, a student trying to make sense of a quarterly report: all four are reading, and all four are usually looking for a number when the answer they actually need is a structural one. The three questions are the fastest way in.

The first question is who decides, and it is a question about names rather than about principles. Which body approves an investment, how many people sit on it, and how many of them work for the manager. In this invented fund the answer is an investment committee of five with four from the manager, and that answer alone establishes that no investor approved any of the nine companies. Next comes what the investor side gets instead, and the actual list is the thing to insist on. Four named matters is a very different answer from a general right to be consulted, and only one of those two is enforceable.

The second question is who funds it, and specifically whether the manager funds alongside. Rs 10,00,00,000 out of Rs 5,00,00,00,000 is a fact that sits in one line of a document, and its absence would be a different fact worth the same one line. 2.0 and 2.0408 are the same money, so name the denominator before writing the percentage down. A report that mixes them is a report to be read twice.

Ask what each side bears third, and ask it of the instrument rather than of anybody's summary. Ask what each contributor has promised, what is still callable, and where that is written. In this fund investor 1 promised Rs 1,00,00,00,000, has paid Rs 96,00,00,000 and has Rs 4,00,00,000 outstanding, and the three unfunded figures at fund level, investor level and manager level are three different numbers that must never be swapped for one another. An hour spent on those three questions reveals more about a private fund than an hour spent on any single performance figure it reports.

One more thing is worth knowing about: the two payments the manager side receives behave so differently from each other that the difference is itself a field of study, examined by Metrick and Yasuda in their work on the economics of private equity funds. A fee that arrives whatever happens and a share of profit that may never arrive at all are not two versions of the same thing, and reading them as one number is how a reader ends up surprised by a fund that has paid a manager Rs 70,20,00,000 and nothing else at all in nine years.

What does this pair of words not say?

The pair of words does not say which party in the actual structure is doing what. The honest limit of the vocabulary is exactly there. The two words name an economic arrangement with two sides, and any particular vehicle may have three parties, or four, or a different set of names in a different country, and the words will still be used because they name the economics rather than the entities. Anybody who wants to know who holds the assets of a particular fund has to go and read who holds the assets of that particular fund.

The pair also says nothing about size, quality, conduct or outcome. Every private fund in existence has both sides of this pair, including the ones that did well and the ones that did not, so knowing that a vehicle has a general partner and a set of limited partners says exactly as much as knowing a company has shareholders. The pair is a starting shape, not a finding. Three questions fill that shape in for one particular fund, and the same three work on any fund at all.