Final Close: When Fundraising Ends and the Clock Starts

Final close is the day a private fund stops accepting commitments and starts counting. Every deadline it has runs from that one day: the investment period, the term, the two extensions, and every performance figure the fund will ever publish. Nilgiri Growth Partners Fund II, invented, treats its own Year 1 Quarter 1 as the first three months after its own final close, and its record date sits 8.50 years after it.

There is a moment in the life of a private fund when nothing visible happens and everything changes. No money moves. No company is bought. Nobody announces anything. A signature goes onto a document, the manager stops accepting further commitments, and from that instant the fund has an origin: one single day that every period, every deadline and every figure it will ever report is measured from. Before that day the fund has no clock at all, and after it every date the fund uses is simply a distance from that one point. The day itself matters, and so does the arithmetic it makes possible. Two funds run by one manager can have clocks four years apart, and that gap produces a collision that costs real people real reconciliations.

Two things about private fund time shape the arithmetic that follows. The first is that it is not calendar time. Time is counted from each fund's own final close and written in that fund's own years and quarters, so a calendar year, a month name and a calendar date have no place in it. The second is that how long a fundraise should take, how long one usually takes, and what a manager must do before it ends are conditions somebody sets: they sit in a fund's own documents and in the current text published by the relevant authority.

What actually happens on the day a private fund holds its final close?

Almost nothing that could be photographed. One permission changes.

Before that day the manager is allowed to admit more investors. The manager can accept another commitment, and every time it does, the size of the fund goes up. After that day it cannot. A final closeThe day a fund stops accepting commitments; every deadline it has runs from it. is the point at which the door shuts on new money for that vehicle, and the interesting thing about a shut door is not the door but everything that becomes calculable once nobody else can walk through it.

Consider a wedding in an ordinary street. For weeks the list moves. An aunt is added, a colleague drops out, somebody's three cousins arrive from out of town. While that is going on, nobody can tell the caterer anything useful. Every number the caterer needs is a share of a headcount that keeps changing. Then there is a day when the household says: that is the list. From that afternoon the hall booking, the per-plate arithmetic, the number of chairs and the size of the tent are all fixed problems with answers. Nothing about the wedding itself has happened yet. The whole thing is still ahead. But the denominator has stopped moving, and a denominator that has stopped moving is what lets everybody else start working.

Nilgiri Growth Partners Fund II is managed by Nilgiri Alternatives Advisors Private Limited, sponsored by Nilgiri Financial Holdings Private Limited, and settled as a trust whose assets are held by Nilgiri Trusteeship Services Private Limited. In this arrangement the role the imported vocabulary calls the general partner is discharged by the manager and the trustee between them, and the contract is a trust deed and a contribution agreement rather than a partnership agreement.

At that fund's final close, three numbers stopped moving at once, and they are worth writing out in full rather than in shorthand. Twelve investors had committed Rs 4,90,00,00,000 between them. The manager had committed Rs 10,00,00,000 of its own, exactly 2.0 per cent of the total. Added together, total commitments are Rs 5,00,00,00,000. Total commitments cannot go up afterwards, and Rs 5,00,00,00,000 is the denominator underneath almost every other figure this fund will ever report.

Now the trap, and it is the single most persistent error anybody makes reading a private fund. There are two denominators sitting in that paragraph, not one. Rs 4,90,00,00,000 is what the twelve investors committed. The manager is a thirteenth party alongside the twelve and is not one of them, so Rs 5,00,00,00,000 is what the fund has in total. Investor 1, a domestic life insurance company, invented, committed Rs 1,00,00,00,000. Against total commitments that is exactly 20.0 per cent. Against the twelve investors' own Rs 4,90,00,00,000 it is 20.4 per cent. Both sentences are true, they are about the same rupees, and a share quoted without saying which denominator it used has not told the reader anything they can check.

The difference between the two denominators is not pedantry and it is not a rounding argument. The gap is 0.4 of a percentage point on one investor, and the same slip repeated across a schedule of thirteen parties produces a column that does not add to the total it claims to add to. Name the denominator every time. The cost is four words.

Why is the origin the final close rather than the first close?

Because a fund at its first close is not finished being assembled, and a measure cannot count from a point that is still moving.

A first closeThe earlier point at which enough is signed for the fund to begin. is the earlier moment at which enough has been signed for the fund to exist and begin working. A first close is a real event with real consequences, and how a fund gets there, what happens to the investors who sign later, and how those later investors are brought level with the earlier ones are all covered separately. One thing matters here: between first close and final close, the size of the fund is still going up. So is the investor list. So is every denominator built on either of them.

Hold that against what a reported figure actually is. Any rate of return a fund quotes is arithmetic performed on cash flows that each carry a date. Feed the same rupees into that arithmetic with different dates and a different answer comes out. Dates are in the calculation for exactly that reason. A fund that cannot say when something happened cannot report a rate of return that anybody else is able to reproduce. Counting from a point that was still moving would mean that two people reading the same fund, one of whom joined at first close and one at final close, would compute two different answers from the same underlying cash and both would be arithmetically correct. Two correct answers drawn from the same cash is not a disagreement anybody can settle.

So the convention picks the later of the two, and picks it for a plain reason: it is the first moment at which the fund is a finished object. The commitments are all in. The investor list is closed. The manager's own commitment is in the total. Everything that is going to be a denominator has stopped being a moving target. From that instant the arithmetic is reproducible by anybody holding the same documents. Reproducibility is the only property a reported figure genuinely needs.

There is a second convention living nearby and it is worth naming so that the two are not confused. A fund's vintageThe year a fund is treated as having started under whichever convention is used. is the year it is treated as having started when it is compared with other funds raised around the same time. Different comparisons pick that year in different ways. Vintage and origin are separate questions with separate answers, and how a fund is placed against a cohort is covered separately. The narrower and harder point is this: whichever origin a fund uses, every single figure downstream is measured from it, so the origin has to be stated and it has to be the same one every time.

How is a date written on this clock, and how does it become a number of years?

A date is written as a year and a quarter of that fund's own life, and it becomes a number of years by one line of arithmetic that never varies.

The fund clockThe counting of years and quarters from one fund's own final close. works like this. Year 1 is the twelve months beginning at final close. Each year is split into four quarters, Q1 through Q4, in order. A date is written as Year 4 Q2 and never in any other form, and in particular never compressed into initials. The compressed form is exactly what tempts a writer into dropping the fund name in front of it. Then the conversion: Year n Quarter m sits at n minus 1, plus 0.25 times m, years after that fund's final close.

The four dates this guide keeps returning to work out as follows, every one of them on Nilgiri Growth Partners Fund II's own clock. Its Year 1 Q1 is zero plus 0.25 times one, or 0.25 years, so the very first quarter of that fund's life ends three months after the door shut. The end of its Year 5 Q4 is four plus 0.25 times four, or 5.00 years exactly. The end of its Year 9 Q2 is eight plus 0.25 times two, or 8.50 years. The end of its Year 10 Q4 is nine plus 0.25 times four, or 10.00 years. The fourth quarter of any year always lands on a whole number, and that is a useful check: a Q4 date that is not a round year means a mistake somewhere in the multiplication.

One of those four figures has a name attached to it. The end of Fund II's own Year 9 Q2 is the record dateThe single moment as at which every figure for this fund is stated. of Nilgiri Growth Partners Fund II, invented, meaning the single moment as at which every carrying value, every cumulative figure and every measure for that fund is stated. Nothing after it is stated anywhere, and nothing after it is projected. Where 8.50 years appears written about this fund, that is where it comes from: not a rounding of nine years, but a quarter that converted to 8.50 by the same rule every other quarter converts by.

Somebody will ask why this bothers with quarters at all rather than counting in years. The honest answer is that a quarter is the unit the documents and the reporting actually run on, so a quarter is what appears in the document in front of the reader, and turning it into a number is therefore a conversion done constantly rather than once. The quarter is also the reason the record date is not a whole year. A fund does not stop reporting on a birthday. A fund stops at the end of a quarter, at the moment the administrator strikes a value, and eight and a half is simply where that quarter fell.

A household version of the same idea, in case the arithmetic made it feel more special than it is. The warranty on a fridge runs from the day it was bought, not from the day it was manufactured and not from the day it was delivered to the shop. Year 1 of that warranty is the twelve months after purchase. If somebody says the fridge is in its third year, that sentence means nothing until it is clear which of those three days is being counted from, and the three can easily sit months apart. A fund is the same shape of problem, only with more money and more people who need the same answer.

Where does the first capital call sit, and what about the other sixteen?

After the origin, like all the others. The answer sounds trivial until it is seen as a definition rather than an observation.

A drawdownA capital call, dated on the fund's own clock. is capital the manager calls from investors against the commitments they have already made, and every one of them carries a date on the fund's own clock. How a call is issued, what notice it carries and how it lands in an investor's account are covered separately and used here rather than explained. Only the geometry matters here: where the calls sit relative to the origin, and what that establishes.

Nilgiri Growth Partners Fund II, invented, made seventeen calls between its own final close and its own record date, and they add to Rs 4,80,00,00,000. Against the Rs 5,00,00,00,000 of commitments fixed on the day of that final close, that is 96.0 per cent. The first sits at the end of its Year 1 Q1 and is Rs 13,10,00,000. The conversion puts it at 0.25 years after the origin. The last sits at the end of its Year 9 Q1 and is Rs 2,20,00,000, or 8.25 years after the origin, one quarter before the record date. Eight years separate the first call from the last.

Every single one of the seventeen is on the same side of the origin, and none of them could be anywhere else. Year 1 Q1 is defined as the first three months following final close, so there is no earlier quarter on this clock for a call to sit in. There is no Year 0 and there is no negative quarter. An origin is not a marker placed somewhere in the middle of a fund's life. An origin is the edge of the counting itself, and everything the fund does lies to one side of it. That is exactly why the choice of origin is not a presentational decision: it fixes what direction time runs in for every figure the fund produces.

The shape of the seventeen is worth a moment because it makes the periods visible without anybody having to describe them. Thirteen of the calls fall inside the first five years and add to Rs 4,55,50,00,000, or 94.9 per cent of everything this fund has ever drawn. The remaining four fall across the three and a quarter years after that and add to Rs 24,50,00,000. The tall bars are early and the slivers are late, and the boundary between the two is not a habit of the manager: it is the end of a period whose length was settled on the day of final close and which the next section takes up.

One more figure belongs here and it is the one this subject gets wrong most often. Rs 5,00,00,00,000 was committed and Rs 4,80,00,00,000 has been called, so Rs 20,00,00,000 of commitment remains uncalled at the record date. The Rs 20,00,00,000 is 4.0 per cent, and it is the fund's own uncalled commitment, sitting at the level of the whole vehicle. The figure is not any single investor's uncalled commitment. Merging the two is how a schedule ends up five times out, and the subtraction is set out below.

What does final close fix that nobody can change afterwards?

Seven things, and they are the reason the day matters at all.

The first is the total of commitments, Rs 5,00,00,00,000. Once the permission to accept more has ended, that total cannot rise. The second is the investor list: twelve investors, and no thirteenth investor may be admitted to this vehicle. The third is the manager's own Rs 10,00,00,000, sitting inside that total as a thirteenth party alongside the twelve and taking the same treatment as an investor interest.

The fourth and fifth are periods. The termThe contracted life of the fund, counted from final close. of this fund is ten years from final close, ending at the end of its own Year 10 Q4, or 10.00 years. The investment periodThe stretch during which capital may be called to buy something new, also counted from final close. is five years from final close, ending at the end of its own Year 5 Q4, or 5.00 years. Each period's permissions and what happens inside it are covered separately. Final close settles the length of each period. Counting that length from the origin settles the end dates too.

The sixth is the two extension rights. Fund II's documents carry two extensions of one year each. The first is at the manager's election with the prior written consent of the investor advisory committee. The second requires the consent of investors holding more than half of commitments by value. Neither has been taken by the end of Fund II's Year 9 Q2. Final close does not fix whether an extension happens; it fixes that there are two of them, that each is one year, and that a named party consents to each. Other funds are written other ways.

The seventh is the origin itself, and it is the one that carries all the others. Fixing the day fixes every date the fund will ever count from. Move that day and the end of the investment period, the end of the term, the outer edge of both extensions and the position of every cash flow in every measure all move with it, at once.

What does final close leave wide open for years?

Almost everything a reader actually wants to know, and saying so plainly is more useful than it sounds.

Nothing about which companies this fund will buy was settled on that day. Nothing about what any holding will be carried at was settled. Nothing about what will be distributed, or in which quarter, was settled. Whether either extension is ever taken was not settled, and at the record date neither had been. Whether the provision that suspends the investment period if the two named key people stop devoting substantially all of their time to the fund is ever triggered was not settled either, and it had not been triggered by the end of Fund II's Year 9 Q2.

Now the subtraction. The fund has Rs 20,00,00,000 of uncalled commitment at its record date, being Rs 5,00,00,00,000 less the Rs 4,80,00,00,000 drawn. The Rs 20,00,00,000 is a fund-level figure. Every investor of this fund is drawn strictly pro rata. Investor 1 committed Rs 1,00,00,00,000 and has paid in 96.0 per cent of it, being Rs 96,00,00,000, so investor 1's own uncalled commitment is Rs 1,00,00,00,000 less Rs 96,00,00,000, or Rs 4,00,00,000. Rs 20,00,00,000 and Rs 4,00,00,000 are five times apart, both are correct, and each is correct only at its own level. Investor 1's figure is also 20.0 per cent of the fund's Rs 20,00,00,000. The fraction never changed, so its share of the whole is its share of the remainder, and that identity is the check. Writing the fund's figure against an investor overstates that investor's remaining obligation by a factor of five.

Sit with the two lists together for a second. The gap between them is the shape of a private fund's whole life. The left list is settled in an afternoon and never moves again. The right list takes ten years to resolve and at the record date is still not finished, with six quarters of the contracted term remaining and the outcome of five holdings unknown. A reader who merges the two lists will make one of two errors: they will think a manager can still change something it cannot, or they will think it cannot change something it can. Both errors read as confidence.

Why do two of this manager's funds land on the same afternoon under different names?

Because each of them counts from its own final close, and those two closes are four years apart.

Nilgiri Growth Partners Fund I, invented, is the earlier vehicle run by the same invented manager. Fund I has been wound up. Its final close falls four years before Nilgiri Growth Partners Fund II's, and that single sentence generates every difficulty that follows. The offset means Fund I's Year n is the same real moment as Fund II's Year n minus 4. The offset also puts Fund II's own final close, its 0.00, at 4.00 years on Fund I's clock, the end of Fund I's own Year 4 Q4. And it means the two funds run alongside each other for years, each reporting in a numbering the other does not share.

Work the collision through. Fund I wound up at the end of its own Year 10 Q4, or 10.00 years after Fund I's final close. Subtract the four-year offset and that same afternoon is 6.00 years after Fund II's final close. The conversion puts it back at the end of Fund II's own Year 6 Q4. Fund I's Year 10 Q4 and Fund II's Year 6 Q4 are one afternoon, and neither name is wrong. On that afternoon one fund finished its life and the other made the first distribution it had ever made, Rs 63,00,00,000 to its investors. Two funds, one manager, one afternoon, two completely different sentences describing it.

The collision gets sharper. Both funds also have a Year 6 Q4 of their own, and the two are four years apart in real time. Fund I's own Year 6 Q4 is 6.00 years after Fund I's final close, and in that quarter Fund I called Rs 6,00,00,000 from its investors. Fund II's own Year 6 Q4 is 6.00 years after Fund II's final close, four years later in real time, and in that quarter Fund II distributed Rs 63,00,00,000. So the characters Year 6 Q4 name two different moments and two opposite directions of cash, and the only thing that separates them is which fund was meant.

Think about two children in the same household attending two different schools. Both say they are in their third year. One started two years before the other, so the two third years are two different stretches of real time, and no amount of care in how the sentence is said fixes it. The school's name in front of the year fixes it. The lesson is not a smaller lesson for being obvious.

Why is a bare year and quarter a reconciliation fault rather than a style preference?

There is a habit that makes the fault far more likely, and it is worth naming because it looks like tidiness. The moment somebody compresses Year 6 Q4 into initials to fit a column, the fund name is the next thing to go. The compressed form already reads as a code rather than as a sentence. Writing the date out in full is slightly inconvenient and that inconvenience is the point: a form that has to be typed out properly is a form whose incompleteness gets noticed.

What can a reader compute the day after final close that they could not compute the day before?

The list is longer than most readers expect, and running through it is the useful test of whether the day mattered.

The last day of the investment period becomes computable. Its length is five years and it now has something to be five years after: the end of Fund II's own Year 5 Q4, at 5.00 years. The last day of the term becomes computable in the same way: the end of its own Year 10 Q4, at 10.00 years. The outer edge of each extension becomes computable, at 11.00 and 12.00 years, without anybody deciding whether either will be taken. Every one of those was a length before final close and became a date at it.

Once the total stopped moving, every investor's share of everything became computable. Investor 1's Rs 1,00,00,00,000 is 20.0 per cent of the Rs 5,00,00,00,000 of total commitments and 20.4 per cent of the twelve investors' Rs 4,90,00,00,000, and neither of those fractions can move afterwards, whatever else happens to the fund. The denominator the management fee is charged on during the investment period stops moving too, and what that fee is and how its basis changes later are covered separately and are used here rather than explained.

And any dated measure at all becomes possible, the change that matters most. Before final close, a cash flow could be given an amount but not a position. There was no agreed zero to measure a position from. After it, every rupee in and every rupee out carries a distance from the origin, and arithmetic that depends on when things happened has something to work on. The day of final close does not make a fund good or bad at anything. The day makes the fund measurable, and measurable is a different and more basic property.

Now the other half. An account that only lists what became knowable has taught half a lesson. Nothing about outcome became computable that day. The fund's purchases, the worth of any of them, what would come back and when, and whether the fund would need either extension were all exactly as unknown the afternoon after final close as the afternoon before. And nothing after the record date is knowable now: this record stops at the end of Fund II's own Year 9 Q2 and states nothing beyond it. Fixing the origin fixes the ruler, not the reading.

Given an hour with two of this manager's funds side by side, what comes first?

Build a third column, and put real time in it.

Somebody sitting inside a manager's finance team, an investor's private markets team or a diligence file does exactly this with everything above. Two reports arrive. Each one is internally consistent and each is written on its own clock. The two clocks do not line up, and neither report mentions the other. Here is the sequence that makes them comparable, and it takes about ten minutes.

First, the offset goes down once, at the top of the working sheet, in words: Fund I's final close is four years before Fund II's, so Fund I's Year n is the same real moment as Fund II's Year n minus 4. The offset is not carried in the head. Every error in this exercise comes from somebody carrying the offset in their head and then being interrupted.

Second, every date on both reports is converted into years after that fund's own final close, using n minus 1 plus 0.25 times m. Fund I's Year 10 Q4 becomes 10.00. Fund II's Year 6 Q4 becomes 6.00. The conversion is done mechanically and the number is written beside the date rather than replacing it. The original form is needed again when the source document is consulted.

Third, the offset is added to the younger fund and both are put onto one column of real time. Fund I's 10.00 stays at 10.00. Fund II's 6.00 becomes 10.00 once its four-year offset is added. The two land on the same row, and that row is the answer: one afternoon. Anything that then appears twice on that row is a single event seen from two sides, and anything that appears once is a single event full stop.

Fourth, before any share from either report is quoted, its denominator has to be established. A share of Fund II's Rs 4,90,00,00,000 of investor commitments and a share of its Rs 5,00,00,00,000 of total commitments are different numbers about the same money, and the manager sits inside the second and outside the first. Where a report does not say, the figure stays unquotable until somebody supplies the answer.

The four steps are also the four questions worth asking of any single private fund report. Which fund's clock is this date on. How many years is that from that fund's own final close. What is the denominator underneath any share quoted. And what is the record date, meaning the last moment anything in the report is stated as at. A report that cannot answer all four has offered nothing that can be checked, however confident the numbers look.