Private Fund Return Multiples: Which Denominator Was Used

A private fund return multiple is total value divided by something, and the something changes the answer. Nilgiri Growth Partners Fund II, invented, held Rs 7,20,00,00,000 of total value at its record date, the end of its Year 9 Q2. Divided by acquisition cost that is 1.80 times. Divided by capital paid in it is 1.50 times. Divided by commitments, 1.44 times. One fund, one afternoon, three numbers.

Consider a vegetable seller anybody could walk past this evening. She pays a wholesaler Rs 400 for a crate. She also pays Rs 50 to hire the handcart and Rs 30 to the boy who helps her carry it to the corner. By nightfall the crate has sold for Rs 720. Asked what she made, she can answer correctly in at least two different ways, and both of them are true. Against the crate, Rs 720 on Rs 400 is 1.80 times. Against everything that actually left her purse, Rs 720 on Rs 480 is 1.50 times. Nothing about her day changed between those two sentences; only the number underneath the division changed. The choice of denominator is the whole subject, and a private fund does exactly the same thing with nine more zeroes on the end.

There is a third answer available to her too, and it is the one nobody ever asks for. She had told the wholesaler she would take up to Rs 500 of stock that week and in the end took only Rs 480 of it. Divide Rs 720 by the Rs 500 she had promised rather than the Rs 480 she spent and the answer is 1.44 times. Three figures, one crate, one evening: 1.80, 1.50 and 1.44. Nilgiri Growth Partners Fund II, invented, produces exactly those three figures at its record date, the end of its Year 9 Q2, and for exactly the same reason.

What is a return multiple actually dividing, and by what?

Every multiple in this subject is one number over another number, and the number on top is the easy half. The top number is called total valueCash already distributed to investors plus the carrying value of everything the fund still holds., and for a private fund it is built from two pieces added together: the cash the fund has already sent back to its investors, and the carrying value of everything it is still holding. Nothing else goes in. No expected sale, no pipeline, no what-it-might-fetch.

The number underneath is where the trouble lives. Three plausible candidates exist, and the fund gets to pick. The fund can divide by what the holdings cost to buy. The fund can divide by every rupee the investors actually handed over. The fund can divide by every rupee the investors ever promised, called or not. A multiple never carries a record of which of the three it was divided by, so the figure travels perfectly well while the meaning falls off somewhere behind it. The travelling figure is not a scandal and nobody is lying. A multiple is a division sum that arrived without its own instructions.

The discipline that follows is a small and boring one. The denominator gets said out loud, every single time, in the same breath as the multiple. Not "the fund is at 1.50 times" but "the fund is at 1.50 times capital paid in". The sentence is four words longer and it is the difference between a measurement and a rumour.

How much of the numerator is money, and how much is somebody's estimate?

Every worked figure below belongs to Nilgiri Growth Partners Fund II, invented. The fund is a closed-end growth and buyout fund managed by Nilgiri Alternatives Advisors Private Limited, also invented, and every figure attached to it belongs to its own clock. At its record date, the end of its Year 9 Q2, its total value is Rs 7,20,00,00,000, and that figure is two very different things stapled together.

Rs 4,38,00,00,000 of it is cash. Four holdings left the portfolio, one of them only partly, and the proceeds went out to investors in four payments. The distributed cash is in bank accounts belonging to twelve investors and a manager. Nobody can revalue it, argue with it or restate it next quarter. The other Rs 2,82,00,00,000 is a carrying value sitting on five holdings that are still held, and not one of those five has been sold to anybody. The carrying value is a number arrived at by a process, reviewed annually by the independent valuation agent, Palani Valuation Advisors LLP, an invented limited liability partnership (LLP), and struck into the accounts by Kolar Fund Services Private Limited, invented, as the administrator. The process is careful. The carrying value is also not cash.

Added together, they make the Rs 7,20,00,00,000. All three multiples have that same Rs 7,20,00,00,000 on top, so every one of them is 39.2 per cent an estimate before the division is even reached. For Nilgiri Growth Partners Fund II, invented, at its record date, the end of its Year 9 Q2, Rs 2,82,00,00,000 over Rs 7,20,00,00,000 is 0.391666, rounding to 39.2 per cent.

Why is one fund on one afternoon 1.80, 1.50 and 1.44 times?

Because the same Rs 7,20,00,00,000 is being set against three different things, and all three of those things are real. Here they are, in the order a reader usually meets them.

The first is acquisition costWhat the fund paid for its holdings, before any management fee and before any fund expense.. Nilgiri Growth Partners Fund II, invented, put Rs 4,00,00,00,000 into nine companies, and Rs 7,20,00,00,000 over Rs 4,00,00,00,000 is 1.80 times at its record date, the end of its Year 9 Q2. Acquisition cost is the deal-level denominator. The multiple on cost answers the question "did the investment decisions work", and it answers it while ignoring every rupee it cost to run the fund that made them. Ignoring the running cost is not a flaw. The deal-level question was never about running cost in the first place. The multiple on cost just is not what an investor got.

The second is capital paid inEvery rupee actually called from investors, for investments, management fee and expenses alike.. Nilgiri Growth Partners Fund II called Rs 4,80,00,00,000 across seventeen drawdowns, and Rs 7,20,00,00,000 over Rs 4,80,00,00,000 is 1.50 times at that same record date. The 1.50 times has a proper name, total value to paid inTotal value divided by capital paid in; this is the figure an investor's own statement shows., and it divides by what the investor actually parted with rather than by what the manager chose to spend on companies. An investor's own statement shows this one and not the other two.

The third is total commitmentsEvery rupee investors promised the fund, whether the fund has called it or not.. Twelve investors and the manager between them promised Rs 5,00,00,00,000, and Rs 7,20,00,00,000 over Rs 5,00,00,00,000 is 1.44 times, again at the record date. Almost nobody quotes the commitment multiple. Knowing it exists is worth more than knowing either of the other two. The commitment multiple is the only one of the three a reader will never be handed, and therefore the only one nobody thinks to ask about.

Look at what the picture refuses to do. The picture does not put a tick beside one of the three. A reader who wants to be told which multiple is the correct one has asked a question that has no answer. Each of the three is the exact and complete answer to a different question. The deal-level figure answers what the buying and selling did. The paid-in figure answers what happened to the money an investor sent. The commitment figure answers what happened to the money an investor set aside. The three questions are separate, and it would be strange if they had the same answer.

What exactly sits in the gap between 1.80 times and 1.50 times?

Something very specific, and it can be pointed at rather than waved at. Capital paid in exceeds acquisition cost by Rs 80,00,00,000 for Nilgiri Growth Partners Fund II, invented, at its record date, the end of its Year 9 Q2. The Rs 80,00,00,000 is two things and only two things: Rs 70,20,00,000 of management fee drawn from investors to that date, and Rs 9,80,00,000 of fund expenses. Rs 4,00,00,00,000 of cost plus Rs 70,20,00,000 plus Rs 9,80,00,000 is Rs 4,80,00,00,000 exactly. The check is worth running whenever both figures are given for the same fund. If the two ends do not tie, one of the two numbers is measuring something that has not been disclosed.

The handcart and the boy carrying the crate are the same item at a smaller scale. The Rs 80 the seller spent on them bought no vegetables. The Rs 80 bought the ability to get the vegetables to the corner and sell them. The handcart and the boy are a real cost, they came out of the same purse, and a figure that divides by the crate alone has quietly stepped over them.

Two things must not be assumed about that Rs 80,00,00,000. None of it is carried interest. Nilgiri Growth Partners Fund II, invented, has paid none at its record date. Its cumulative distributions of Rs 4,38,00,00,000 are still Rs 42,00,00,000 short of the Rs 4,80,00,00,000 it has called, and under a whole-of-fund arrangement nothing beyond the return of capital gets paid until that gap closes. And none of it is a judgement on the manager's fee. The fee and its basis were settled in the fund's documents before a rupee moved, and they are covered separately.

And what sits in the smaller gap, between 1.50 times and 1.44 times?

Money that never moved at all. Total commitments exceed capital paid in by Rs 20,00,00,000 for this fund at its record date, being 4.0 per cent of the Rs 5,00,00,00,000 promised. The Rs 20,00,00,000 is unfunded commitmentThe part of a promise contracted for but never called, so it has never left the investor's own account.: contracted for, sitting in twelve investors' own accounts, available if the fund asks. The unfunded commitment has never been in the fund, has never bought anything, and has never cost anybody a rupee of return.

And yet the multiple on commitments puts it in the denominator anyway. The uncalled Rs 20,00,00,000 drags the answer from 1.50 times down to 1.44 times. Is that unfair? No, and this is worth being careful about. An investor who set Rs 5,00,00,00,000 aside for this fund had to hold it available. The set-aside could not be spent twice. Asking what the whole set-aside produced is a perfectly sensible question, and the honest answer is 1.44 times for Nilgiri Growth Partners Fund II, invented, at its record date. The figure is not more honest or less honest than the other two; it is the answer to a question about the set-aside rather than a question about the spending.

What happens to the multiple when only the denominator moves?

The control below can be dragged, and the answer is worth predicting before it is touched. The numerator is nailed down at Rs 7,20,00,00,000, the total value Nilgiri Growth Partners Fund II, invented, actually had at its record date, the end of its Year 9 Q2. The acquisition cost is nailed down at Rs 4,00,00,00,000. The only thing that moves is the fee and expense drawing that sits on top of the cost.

Notice the shape of what just happened. The two markers meet at the Rs 0 end and then walk apart, and the distance between them at any setting is not a return, a loss or a performance. The distance is a cost of operation expressed in turns of a multiple. A fund with a heavier drawing on top of the same investments will show a lower multiple on paid in while showing exactly the same multiple on cost, and both of those figures are correct at the same moment. Both figures being correct at once is the single most useful point in the whole subject.

How do the cash half and the estimate half fit back together?

Everything so far has changed the denominator and kept the numerator whole. The opposite move fixes the denominator at capital paid in and splits the numerator into its two honest halves. The split produces the two other names a reader meets constantly and rarely has explained.

Distributions to paid inCash returned to investors divided by capital paid in. Everything in it has actually been paid. is cash returned over capital paid in. For Nilgiri Growth Partners Fund II, invented, at its record date, the end of its Year 9 Q2, that is Rs 4,38,00,00,000 over Rs 4,80,00,00,000, or 0.9125 exactly, written 0.91. Residual value to paid inThe carrying value still held divided by capital paid in. Nothing in it has been sold to anybody. is carrying value still held over the same capital paid in: Rs 2,82,00,00,000 over Rs 4,80,00,00,000, or 0.5875 exactly, written 0.59.

Added together, 0.9125 plus 0.5875 is 1.5000, and 1.5000 is the total value to paid in the fund reported at its record date. Not approximately. Not to two places. Exactly. The addition works only because all three fractions are divided by the very same Rs 4,80,00,00,000, and if it ever fails to work in a report, the three figures were not built on one denominator and somebody has some explaining to do.

The three are best held as one fraction rather than as three acronyms. There is one denominator, capital paid in, and there are three numerators sitting over it: cash returned, value still held, and the two of them together. A reader who can say which numerator and which denominator a figure used can read any multiple anybody hands over. A reader who has only memorised three names has learned three names.

The split is doing real work, and the work is this. Nilgiri Growth Partners Fund II, invented, is at 1.50 times paid in at its record date, and 0.9125 of that has been paid to somebody while 0.5875 of it rests on five marks. Put another way, of every rupee of that headline multiple, roughly sixty one paise is bank balance and thirty nine paise is a carefully struck opinion about five companies nobody has yet bought. Both halves are legitimate. Cash and mark are not the same kind of thing, and the addition forces a reader to notice the difference rather than glide over it.

The identity is not a fund-level curiosity either. The identity survives all the way down to a single investor. Investor 1 of this fund, a domestic life insurance company, invented, committed Rs 1,00,00,00,000 and has paid in Rs 96,00,00,000 to the record date. The insurer has received Rs 87,60,00,000 of distributions and holds a Rs 56,40,00,000 share of the residual value. Rs 87,60,00,000 over Rs 96,00,00,000 is 0.9125. Rs 56,40,00,000 over Rs 96,00,00,000 is 0.5875. Its own total value to paid in is 1.50 times. Every drawing is strictly pro rata and no side letter moves its economics.

What is the 0.5875 actually made of?

Five holdings, and they are worth seeing broken out. A single figure of 0.5875 hides the fact that the pool is not evenly spread. Nilgiri Growth Partners Fund II, invented, still holds five companies at its record date, and their carrying values are Rs 1,08,00,00,000, Rs 21,00,00,000, Rs 39,00,00,000, Rs 81,00,00,000 and Rs 33,00,00,000, adding to Rs 2,82,00,00,000. One of those five is carried below what the fund paid for it. None of the five has been sold.

The picture carries a second lesson hidden in its own arithmetic. The five shares of the pool, rounded to one decimal place, are 38.3, 7.4, 13.8, 28.7 and 11.7 per cent. Added together, those five printed figures make 99.9, not 100.0. The unrounded shares sum to exactly 100.0000, and the missing tenth is an artefact of rounding five numbers independently rather than a hole in the record. The correct response is to say so in a note. The incorrect response, and it happens in real reports, is to nudge one of the five up by a tenth so the column looks tidy. The nudge puts a figure into a report that no arithmetic produced, and a nudge is exactly how a small untruth gets into a document that is otherwise honest.

What do gross and net mean here, and why do they move around?

Two words, and they are the least informative pair in this whole subject unless somebody finishes the sentence. GrossBefore something. The something has to be named every time, or the word has said nothing at all. means before something. Net means after that same something. Neither word names the something, and the something is different on different documents. In one report it means before and after the management fee. In another it means before and after carried interest. In a third it means before and after both. The word is a container, and somebody has to name what is in it.

The difference shows properly in a completed record. Nilgiri Growth Partners Fund I, invented, wound up at its own Year 10 Q4. Over its ten years it called Rs 2,40,00,00,000 from its investors and distributed Rs 4,80,00,00,000. Rs 4,80,00,00,000 over Rs 2,40,00,00,000 is 2.00 times. Of that Rs 4,80,00,00,000, the amount that reached investors was Rs 4,32,00,00,000, and Rs 4,32,00,00,000 over Rs 2,40,00,00,000 is 1.80 times. The 0.20 times between those two figures is carried interest paid to the manager and it is nothing else at all.

Set beside the other 0.30 already worked above, the two are not the same species of number, even though both get called a gap in conversation. Nilgiri Growth Partners Fund II's 1.80 times against 1.50 times at its record date is one numerator over two different denominators. The fund has paid no carried interest, so none appears in either figure. Nilgiri Growth Partners Fund I's 2.00 times against 1.80 times at its wind-up is two different numerators over one denominator, and carried interest is the whole of it.

Every date in that comparison is written with its fund named in front of it, and that is not fussiness. The two funds run on two different clocks. Nilgiri Growth Partners Fund I reached its final close four years before Nilgiri Growth Partners Fund II did, so Fund I's Year 10 Q4 and Fund II's Year 6 Q4 are the same afternoon in the world. A bare year and quarter in a document that names both funds is not a style choice, it is a mistake waiting to be made by whoever reads it next.

What can a multiple not tell anybody at all?

How long it took. Not approximately, not badly: at all. There is no clock anywhere in any of the three fractions. Total value over capital paid in is two rupee amounts and a division sign, and it can be computed on a fund four years old or a fund eleven years old and the arithmetic would not notice which. Two funds sitting at 1.50 times are not the same thing if one of them took four years and the other took eleven, and no multiple in existence can tell them apart.

The missing clock has two consequences, and both of them catch people out. The first is that a multiple can fall while nothing whatever happens to the holdings. Nilgiri Growth Partners Fund II, invented, was at 1.51 times total value to paid in at its Year 8 year end and at 1.50 times at its record date, the end of its Year 9 Q2. In those two quarters it sold nothing, wrote nothing up and wrote nothing down. The fund called Rs 2,20,00,000 more for fee and expenses, the denominator got bigger, and the multiple came down. The fall is arithmetic. The fall is not a loss and nothing went wrong.

The two figures are two points of that fund's own path, and only two. The full year by year shape a private fund's value traces from its first drawdown to its last distribution is a subject of its own, covered separately.

The second consequence is the bigger one. A measure that does contain time exists. The measure is a rate of return rather than a multiple, it answers a genuinely different question, and comparing the two against each other is covered separately. The first thing to establish is this: if the number in question is a multiple, the years are simply not in it, and they have to be asked for separately or accepted as unknown.

What actually goes wrong when two multiples are put in a list?

How should a multiple that somebody hands over be read?

With three questions, and they are always the same three, in the same order, whatever the fund and whatever the market. The three questions take about fifteen seconds and they are the whole of the craft.

Ask what is inside the numerator, and specifically how it splits between cash already paid and value still held. Ask for the denominator, named in full: cost, capital paid in, or total commitments. Ask as at what date, on whose clock, and how many years have elapsed since the money first went out. A multiple that can answer all three is a measurement that can be set beside another one; a multiple that cannot answer even one of them is a number, and it stays a number however impressive it looks.

A last habit is worth building, and it costs nothing. Whenever both the headline and one of its two parts are given, the subtraction is worth doing. Given 1.50 times and 0.91, the 0.59 falls out, and the next question is what those five holdings are. Given the deal-level figure and the paid-in figure, the distance falls out, and the question is what filled it. The arithmetic is trivial. The questions it produces are the ones that turn a reported figure into something actually understood.