General Partner: Who Runs the Fund and What They Are Paid For

The general partner runs the fund. The general partner decides what the fund buys and sells, signs on its behalf and answers for how it is managed. Payment comes in two quite different ways: a management fee charged on a contracted base whatever happens, and carried interest. Carried interest is a share of profit, and it arrives only if the distribution waterfall reaches it. The two behave nothing alike.

Most people meet this subject through one phrase, and the phrase does them a quiet disservice. Two and twenty. The phrase sounds like a single price, the way a broker's commission is a single price, and a reader who holds it that way has one number in their head for what a manager earns and no way at all to tell a good year from a bad one. The two figures in that phrase are not two parts of one payment. They are two payments with different bases, different triggers, different timing and different certainty. In the fund worked below, one of them produces Rs 70,20,00,000 and the other produced nothing whatsoever. Andrew Metrick and Ayako Yasuda, in The Economics of Private Equity Funds, published in the Review of Financial Studies in 2010, made exactly this separation the centre of their work. A manager's certain fee income and its contingent profit share behave as two different things, so they valued them as two different things. The split between the certain payment and the contingent one is what everything below rests on.

The worked case throughout is Nilgiri Growth Partners Fund II, invented, a closed-end fund of Rs 5,00,00,00,000 of commitments managed by Nilgiri Alternatives Advisors Private Limited, also invented. Every figure here belongs to that fund at its record date, the end of its Year 9 Quarter 2. That date falls 8.50 years after its final close. No number is usual, expected or characteristic of anything beyond this one invented arrangement. Fund II has never paid a profit share, and nothing about a profit share can be learned from a fund that has not paid one. A second fund therefore appears below: Nilgiri Growth Partners Fund I, invented, over its own completed ten years.

What does the general partner actually do?

Start away from funds altogether. A building of forty flats has an association, and the association has a managing committee. The committee decides which contractor repaints the block, signs the contract, watches the work, argues when the paint peels in eighteen months, and stands in front of the residents at the annual meeting to say what was spent and why. The residents pay. The residents do not choose the contractor. If a resident thinks the committee picked badly, the resident can complain loudly, can refuse to stand for the committee next time, and can vote at the annual meeting, but the resident cannot ring the contractor and cancel the job. The division between the people who put up the money and the small group who make and answer for the decisions is the shape that recurs at a much larger scale.

A general partnerThe party that runs the fund, decides what it buys and answers for it. in a private fund does five things, and it is worth having them as five rather than as a vague sense of running things. The general partner decides, meaning it chooses which companies the fund buys, at what price, on what terms, and when the fund sells them again. It signs, meaning it is the party whose signature binds the fund to a share purchase agreement, a shareholders agreement, a loan document or a notice calling capital. The same party manages, meaning it does the work between buying and selling: taking board seats, reading the monthly numbers, replacing a finance head, deciding whether to put in more money. It realises, meaning it chooses the route out and negotiates the second number. The exit price is the only number that ever turns a holding into cash. And it answers, meaning it reports to the investors on a contracted timetable and stands behind how the fund has been run.

Every one of those five is a decision the investors have handed over, and handing them over is the entire reason a pooled fund exists rather than twelve investors each buying companies for themselves. An investor in Nilgiri Growth Partners Fund II, invented, committed money to a set of decisions it had not yet seen. At the moment of committing, not one of the fund's nine holdings existed. The manager went on to review 412 opportunities, sign 31 confidentiality undertakings, issue 14 non-binding offers, go to exclusivity on 11 and complete 9 of them. The investor was not asked about any of the 412 and could not have vetoed the nine. None of that is an oversight in the arrangement. It is the arrangement.

Why does one side answer for everything and the other only for what it promised?

Now the part that explains why the role exists at all, and it is not the fee. The answer is liability. In the partnership form this whole vocabulary comes from, a limited partner and a general partner do not stand in the same place when something goes wrong. The limited partner's exposure is capped at what it promised to put in, and the general partner's is not capped at all, and that single asymmetry is the reason somebody has to hold the second role rather than everybody holding the first.

Put it in household terms first. Two people back a small workshop. One puts in Rs 5,00,000 and takes no part in running it: if the workshop collapses owing money to a supplier, that person loses the Rs 5,00,000 and the supplier cannot come to their door for the rest. The other person runs the workshop, signs the orders and is answerable for its obligations without a ceiling: if the workshop owes more than it has, that is a claim against them. The first person has bought a capped risk in exchange for having no say. The second has taken an uncapped one in exchange for having every say. A supplier will not deal with a workshop where nobody at all is answerable, so neither position is available without the other.

Scaled up, that is the partnership shape. The investor commits Rs 1,00,00,00,000, as investor 1 of Nilgiri Growth Partners Fund II, invented, actually did, and that commitment is the ceiling on what the arrangement can ever ask of it. The same commitment is also the price of silence on the five duties drawn above. The general partner takes the decisions and, in the pure form of the structure, answers for the partnership's obligations without a corresponding ceiling. The pay is downstream of that swap, not the cause of it: the payments exist because somebody has to be answerable, not the other way round. That matters because the fund worked here is a trust rather than a partnership, and what stands in place of the uncapped liability once the partnership form is left is set out below.

What is the general partner paid, and what is each payment for?

Two payments. Not one payment with two components, and this is where the phrase two and twenty does its damage. The management feeA charge on a contracted base, payable whatever the fund's holdings do. is paid for running the fund. The fee is charged on a base the contract names, it is charged whether or not anything is made, and it starts in the first year when the fund has bought nothing at all. Carried interestThe manager's share of profit, paid only if the distribution order reaches it. is paid for gains, and only for gains. Carried interest is a share of profit. The trigger is the order of distributions reaching the point where it applies, and if that point is never reached it is never paid.

The abstraction hides the difference, so say it in ordinary words. One of these two payments is a cost of operating and the other is a share of a result. A reader who calls them two kinds of pay has learned nothing. What separates them is not how much each is, but what has to happen before either arrives. A shop assistant's wage is paid at the end of the month whether the shop had a good month or a terrible one. A commission on sales above a target is paid only if the target is passed. Both are money reaching the same person. The wage and the commission answer completely different questions about the month. Nobody would add them into one figure and call it the cost of the assistant without saying which part was which.

The terms of the fund in view are these, and they are Nilgiri Growth Partners Fund II's own contracted terms, invented. The management fee is 2.00 per cent a year. The carried interest is 20.0 per cent, taken above a preferred return of 8.0 per cent a year compounded annually. The profit share sits in the fourth tier of the fund's distribution order, which is covered separately. Other funds contract other terms. To the record date the fee has produced Rs 70,20,00,000 and the carried interest has produced nothing at all. The fund has distributed Rs 4,38,00,00,000 against the Rs 4,80,00,00,000 it drew. A shortfall of Rs 42,00,00,000 is how far it stands from the point where any profit share can begin.

What does the management fee actually pay for?

Here is the question people skip, and skipping it is why the fee feels arbitrary. The fee does not buy anything the fund holds. Not one rupee of the Rs 70,20,00,000 charged by Nilgiri Growth Partners Fund II, invented, to its record date went into a company. The fund's nine holdings cost Rs 4,00,00,00,000 and that figure is separate from the fee and separate again from the Rs 9,80,00,000 of fund expenses. Adding the three gives exactly the Rs 4,80,00,00,000 the fund has drawn from its investors. Run that check every time these numbers appear: Rs 4,00,00,00,000 plus Rs 70,20,00,000 plus Rs 9,80,00,000 is Rs 4,80,00,00,000, to the rupee.

So what does the fee buy? The fee pays for the existence of the firm that does the five duties drawn above. Salaries for the people who read 412 opportunities to find nine. The office they sit in. The travel to see a factory in person. The lawyers and accountants who run diligence on the eleven that reached exclusivity. Two of those eleven fell away, produced no holding and still cost real money. The systems that produce a quarterly capital account statement for twelve investors. The fee buys the capacity to look. Looking has to be paid for whether or not it finds anything, and that is precisely why the fee cannot be contingent on a result.

There is a second half to this that readers routinely miss. Because the fee is drawn from the investors like any other capital call, it is capital the investors have paid in, and the fund's distribution order treats it as such. The first thing the order does is return every rupee of capital ever drawn, and the fee and the expenses are inside that every rupee just as much as the money that bought companies. So the fee is not a charge that vanishes: it is money the investors must get back before the manager sees any share of profit at all. The contracted order of payments is worked out fully under the distribution order. Treating the fee as returnable capital is what makes the Rs 70,20,00,000 raise the bar the manager has to clear before the second payment can begin.

Why does the basis matter far more than the rate?

Everybody quotes the rate. Almost nobody asks what it is charged on, and the thing it is charged on is doing most of the work. The fee basisThe amount the percentage is charged on. In this fund it changes partway through. of Nilgiri Growth Partners Fund II, invented, is not one number for the fund's whole life. Its investment periodThe opening stretch of a fund's life during which it may make new investments. runs five years from final close. Through those five years the basis is aggregate investor commitments of Rs 4,90,00,00,000, so 2.00 per cent of that is Rs 9,80,00,000 a year, five years running, Rs 49,00,00,000 in all. From Year 6 the basis becomes something else entirely: the acquisition costWhat the fund paid for a holding, before any change in its value. of the holdings the fund has not yet sold, measured at the start of each year.

Getting this wrong is the most repeated arithmetic fault in the whole subject, so notice which number the fee is charged on. The manager does not charge itself a fee on its own Rs 10,00,00,000, so the basis during the investment period is the Rs 4,90,00,00,000 of investor commitments and not the Rs 5,00,00,00,000 of total commitments. The gap is small and the discipline is not. The same Rs 9,80,00,000 is 2.00 per cent of investor commitments and 1.96 per cent of total commitments, both exact, and either share quoted without naming the figure it was divided by is something nobody can check.

The second basis behaves in a way the first never could. Once the fee is charged on the cost of what is still held, it falls whenever something is sold. Selling removes that holding's cost from the base. The base is cost and not value, so the charge does not fall when a holding rises in value and does not rise when a holding is written down. An unrealised holdingA holding the fund still holds and has not sold. written down to a fraction of what it cost stays in the base at its full cost until it is actually disposed of. Holding 6 of this invented fund cost Rs 30,00,00,000 and is carried at Rs 21,00,00,000. Nobody has sold it, so it is still in the base at Rs 30,00,00,000.

What changes at the end of the investment period?

The step-downThe contracted change of fee basis once the investment period ends. is the least obvious thing about a management fee, and it is not a renegotiation. Nobody sat down at the end of Year 5 and asked for a lower fee. The contract said from the beginning that when the investment period ended the base would change from commitments to the cost of what was still held, and at the end of Year 5 that day arrived. The rate never moved, not by a single basis point, and the annual charge still fell by nearly two thirds. The rate was never what did the work.

Why would a contract be written that way at all? Look at what the fee buys at each stage. In the first five years the manager is looking at 412 opportunities to find nine, and the work is proportional to the money it has been asked to deploy, so the base is the money promised. After Year 5 no new investment may be made, and the work is proportional to what is still being held and still has to be managed and sold, so the base becomes the cost of what is still there. The base follows the work. When holdings 2 and 5 left the portfolio in Fund II's Year 6, the base fell by their Rs 45,00,00,000 and Rs 35,00,00,000 of cost, and the next year's charge fell with it, without anybody negotiating anything.

The fee of this invented fund, year by year, to the rupee

Everything below belongs to Nilgiri Growth Partners Fund II, invented, over its own first 8.50 years.

Two checks are worth running yourself. Neither is obvious, and both tie this table to the rest of the fund. First, the Year 9 basis of Rs 1,80,00,00,000 can be rebuilt from the other side: holdings 4, 6, 7 and 8 cost Rs 60,00,00,000, Rs 30,00,00,000, Rs 30,00,00,000 and Rs 45,00,00,000, being Rs 1,65,00,00,000, and holding 9's remaining 60 per cent carries Rs 15,00,00,000 of cost, and Rs 1,65,00,00,000 plus Rs 15,00,00,000 is Rs 1,80,00,00,000 exactly. Second, the fee of Rs 70,20,00,000 plus fund expenses of Rs 9,80,00,000 plus the Rs 4,00,00,00,000 that actually bought companies is Rs 4,80,00,00,000, precisely the capital this fund has drawn. Every rupee this manager has been paid to run the fund is inside the capital the investors put in. The total drawn is therefore larger than the total invested, and the two figures are never interchangeable.

What is the second payment, and why has it never arrived here?

Carried interest is the manager's share of the profit, and on this fund it is 20.0 per cent above a preferred return of 8.0 per cent a year compounded annually. The profit share is the fourth thing in the order in which distributions are applied. That order, with the return of capital, the preferred return and the catch-up before it, is worked in full separately. The trigger is what matters. The management fee is triggered by the calendar and the profit share is triggered by cash actually going back to investors, so one of them has arrived nine times and the other has never arrived at all. Nilgiri Growth Partners Fund II, invented, has distributed Rs 4,38,00,00,000 against Rs 4,80,00,00,000 drawn, so every rupee it has paid out is still return of capital and its carried interest to the record date is nil. Nilgiri Growth Partners Fund I, invented, is the other side of that: over its completed ten years it distributed Rs 4,80,00,00,000 on Rs 2,40,00,00,000 paid in, and the manager's carried interest came to Rs 48,00,00,000, exactly 20.0 per cent of that fund's Rs 2,40,00,00,000 of profit.

So how do the two sides actually work together across a fund's life?

How General Partners and Limited Partners Work Together

The arrangement between the two sides is usually described as a relationship. The word tells a reader nothing they can use. What the documents actually set up is four channels written into a contract, each carrying a different kind of traffic in a different direction. Once the four can be named, almost any question about who may do what has an answer.

The first channel is decision, and it runs one way and stops. The manager chooses what the fund buys, at what price, on what terms, and when to sell. The decision channel does not reach the investors at all. The second is consent, and it runs the other way, from investors to manager, on a short and defined list. In Nilgiri Growth Partners Fund II, invented, an investor advisory committee of seven, drawn from seven of the twelve investors and chaired by Meera Sathe for investor 1, consents on conflicts, on valuation policy, on the first of the fund's two one-year extensions and on any change to the investment policy. The committee does not approve investments and it cannot reject one. That advisory committee is covered separately. The third is information, running manager to investors on a contracted timetable. The fourth is vote, running investors to manager, taken by value of commitments rather than by head, on a short list of questions such as the second extension or the approval of a replacement key person.

Notice what is missing. No channel at all lets an investor reverse an investment decision after it has been taken. The absence is the arrangement working rather than the arrangement failing. An investor who dislikes what the manager did with holding 6, which cost Rs 30,00,00,000 and is carried at Rs 21,00,00,000, can read about it in the quarterly reporting, can raise it at the advisory committee if it holds one of the seven seats, and can vote where the contract gives it a vote. The investor cannot unwind the position. If it could, twelve investors would each be running the fund. Avoiding exactly that is why the structure exists.

Think about the residents' committee again. The residents receive the accounts, consent to a special levy above a certain size, and vote at the annual meeting. Not one of those is the power to ring the painter. The private fund version writes the same division down with more precision and a longer time horizon. The money is locked for ten years rather than until the next annual meeting, and precision matters more the longer an investor cannot leave.

What does the manager owe the investors that is not money?

A great deal, and it is contracted rather than voluntary. An investor of this invented fund receives six things, and they are worth listing because they are the entire content of the third lane. A capital account statement each quarter. An unaudited quarterly report within a number of days of quarter end that the fund's own documents fix. An audited annual report. A letter from the manager alongside the quarterly numbers. A notice for every capital call and every distribution. And an annual valuation report from an independent valuation agent. Here that agent is Palani Valuation Advisors LLP, a limited liability partnership, invented. Kolar Fund Services Private Limited, invented, is the administrator that strikes the net asset value.

Reporting is a duty on a timetable, not a service on request, and the difference decides what an investor can expect when it wants to know something on a Tuesday afternoon. The manager does not have to answer every question the moment it is asked. The manager has to produce what the contract says, when the contract says. The timetable is why an investor's operational review of a manager, run before committing, spends its time on the machinery rather than on the numbers: who strikes the net asset value, how independent the administrator is, how many people it takes to move cash, whether the auditor has ever issued anything other than a clean opinion. Investor 5 of this invented fund, a fund of funds, ran exactly that review on Nilgiri before committing to Fund II. None of it is a promise about outcomes. The review is a check on whether the machinery that produces the reporting actually exists.

What is the manager's own money doing in an account of how the manager is paid?

Nilgiri Alternatives Advisors Private Limited, invented, committed Rs 10,00,00,000 of its own to Nilgiri Growth Partners Fund II, invented, alongside the Rs 4,90,00,00,000 the twelve investors committed, taking total commitments to Rs 5,00,00,00,000. The manager's own commitmentThe manager's own money committed to the fund on investor terms. is exactly 2.0 per cent of the total, it is funded in cash rather than waived against the fee, and it takes the same treatment as any investor interest in every tier of the distribution order. Like everybody else's, that commitment has been drawn 96.0 per cent. Rs 9,60,00,000 of the Rs 10,00,00,000 has actually been called.

Two mechanical consequences follow, and they are the reason this belongs with payment rather than with governance. The first is that the manager does not charge itself a fee on its own money. The fee basis during the investment period is therefore the Rs 4,90,00,00,000 the investors committed and not the Rs 5,00,00,00,000 the fund committed in total. The second is that the manager is a payer as well as a payee: when a capital call goes out for the fee, the manager pays its own 2.0 per cent share of that call like every other holder of an interest.

What goes wrong when two and twenty is read as one number?

Who is the general partner in an Indian fund settled as a trust?

Everything above uses the vocabulary the documents and the investors actually use, and that vocabulary comes from the limited partnership. Now the honest complication. Nilgiri Growth Partners Fund II, invented, is not a limited partnership. The fund is a trust, settled under an indenture of trust. Indian pooled private vehicles most commonly take that form. There is no general partner in it as a matter of law, and there is no limited partner either. Every person involved will still use both words all day long. The economics were designed in the partnership form and then imported into this one.

So the question is not what the words mean. The question is which party does which job. The role a general partner discharges in a partnership is split here between two parties, and knowing which one does what is the difference between reading these documents correctly and guessing. Nilgiri Alternatives Advisors Private Limited, invented, is the investment manager: it takes the investment decisions, signs for the fund, reports, and is the party paid the management fee and the carried interest. Nilgiri Trusteeship Services Private Limited, invented, is the trustee: it holds the fund's assets and has duties to the beneficiaries, and it is paid no profit share. Between the two of them they do what a general partner would do. The uncapped liability drawn earlier does not appear here in that form at all. There is no partnership for anybody to answer for. The trust deed and the contribution agreement set out what each party answers for instead.

There is a third party and it is not either of those two. Nilgiri Financial Holdings Private Limited, invented, is the sponsor. The sponsor stands behind the manager and holds the manager's own Rs 10,00,00,000 commitment. Sponsor, manager and trustee are three separate parties with three separate jobs, and collapsing any two of them is the most reliable way to misread this structure. The sponsor is covered separately.

How does anybody actually use this, on a Monday morning?

Three people read these terms for three different reasons, and none of them is reading for an opinion about whether the pay is reasonable. The same clause in the fund's documents produces three different jobs.

The analyst at an institution considering a commitment is building a picture of what the arrangement costs over its life, and the basis is the entire job. The base is a fixed number, so given the terms of Nilgiri Growth Partners Fund II, invented, that analyst can write down Rs 9,80,00,000 a year for five years without knowing anything about the portfolio. From Year 6 the analyst cannot write anything down at all. The base is then the cost of what has not been sold, and that depends on realisations nobody can predict. Half of this fee schedule is knowable at signing and the other half is not, and an analyst who models the second half as though it were the first has produced a number that looks precise and is not.

The operations person inside the manager reads it as a cash calendar. Every one of those charges is drawn from investors through a capital call notice, signed here by Farida Contractor as chief operating officer, and every call has a notice period and a pro rata split across twelve investors. The Rs 70,20,00,000 is not one payment; it is a series of them, mixed in with the calls that bought companies and paid expenses.

The investor's own accountant reads it as a reconciliation. The fee charged to the fund appears in the capital account statement as capital contributed, not as an expense sitting somewhere outside. That treatment is why total capital drawn of Rs 4,80,00,00,000 exceeds total invested of Rs 4,00,00,00,000 by exactly the Rs 70,20,00,000 of fee and the Rs 9,80,00,000 of expenses, and why an investor who compares distributions to the amount invested rather than to the amount drawn will always overstate how far the fund has come. Investor 1 of this invented fund, with a Rs 1,00,00,00,000 commitment, has contributed Rs 96,00,00,000 and received Rs 87,60,00,000, all of it return of capital, and its share of the fee is inside that Rs 96,00,00,000 rather than beside it.