Exit: How a Private Investment Turns Into Cash

An exit is the moment a private position is turned into something an investor can be paid with, and until it happens a holding's value is an estimate. A position can leave by five routes: sale to a buyer in the same industry, sale to another fund, an offering followed by a sell-down, a secondary sale of part of the position, and a write-off.

One fact makes this subject different from anything met in listed markets. A private fund publishes a value for every holding it has not sold, and that value is a figure somebody produced by looking at information on a timetable. The figure is careful, the figure is signed off, and nobody has paid it. An exit is the only event that replaces a number somebody produced with a number somebody paid. Everything a private fund reports before that moment is, in the most literal sense, a proposal awaiting a counterparty.

Ang, in Asset Management, 2014, treats illiquidity as a structural feature of an asset rather than as an inconvenience attached to it. Taken seriously as a feature, illiquidity makes the way out not an afterthought bolted on to the end of an investment but the test of everything reported before it. So what follows is not a tour of transaction types but one question asked five different ways: what has to be true before an investor sees money?

The worked case throughout is Nilgiri Growth Partners Fund II, an invented closed-end fund managed by Nilgiri Alternatives Advisors Private Limited, sponsored by Nilgiri Financial Holdings Private Limited, and settled as a trust whose trustee is Nilgiri Trusteeship Services Private Limited. Twelve investors and the manager committed Rs 5,00,00,00,000 between them. The fund bought nine companies for Rs 4,00,00,00,000 of cost. By its record date, the end of its Year 9 Quarter 2, five exit events had happened across five of those nine holdings, and each one used a different route. Every figure below belongs to that one invented fund over that one stated period, and a different fund with a different portfolio would produce a different set.

What is an exit, and what has to happen before an investor sees any money?

A house a household has lived in for twenty years makes the point. Everybody on the street has a view on what it is worth, the neighbour who sold last spring has a number, and a broker will offer a range over tea. None of that is a price. The price exists on exactly one day, when somebody signs, the transfer is registered and the money lands in the account. Until that day the house is worth an opinion, and opinions do not pay a school fee.

A private fund is that situation, repeated nine times, run with other people's money, and under a duty to report a number every quarter anyway. The word exitThe event that turns a private position into something an investor can actually be paid with. covers the day the opinion is replaced. Exit is a plain word for a plain thing, and because it marks the boundary between two completely different kinds of number, the word does far more work than it looks like it does.

Two things then have to happen, in order, and readers routinely collapse them into one. First the fund receives money, the realisationThe moment the fund itself actually receives the money for a holding it has sold.. Then the fund pays money out, the distributionThe payment the fund makes out to its own investors from money it has received.. A realisation is money arriving at the fund and a distribution is money leaving it, and anybody who treats them as one event will get the timing wrong every single time. In Nilgiri Growth Partners Fund II they are always one quarter apart, and that gap is a term of this fund's own documents rather than a rule about funds anywhere.

Now look at when this actually happened. The fund had a five-year investment period and a ten-year contracted term. Across the first five years the fund bought nine companies and sold none of them. Selling nothing for five years is neither a failing nor a delay. A fund that buys unlisted businesses and then works on them simply looks like that from the outside for half its life. Then, between Year 6 Quarter 3 and Year 8 Quarter 3, five positions produced five exit events in nine quarters.

Two readings of that picture are worth carrying forward. The first is that a fund which has sold nothing for five years is not necessarily idle. The second is that once positions do start leaving, they can leave quickly and close together. Both are properties of this one invented fund at these stated dates, and neither is a pattern to expect anywhere else.

What are the five routes a private position can leave by?

How Private-Market Exit Routes Work

There are five, and the whole set is easier to hold as five answers to the same four questions than as five kinds of transaction. Who is on the other side? How much of the position leaves? When does the cash arrive? And what does the fund still hold once it is done? Every route out of a private position is just a particular set of answers to those four questions, and no two of the five answer all four the same way.

Nilgiri Growth Partners Fund II used each of the five exactly once. One of each is convenient for teaching and is not how a real portfolio would be expected to distribute itself. Here they are against the four questions, and then each one is worked on its own holding underneath.

The table below sets out the same five routes as a schedule, the form they would take inside a fund's own records rather than as a diagram. The last column repays care: it is a multiple on the cost that each exit actually released, not on the fund's whole Rs 4,00,00,00,000, and those are different denominators.

A total nobody has checked is just a claim, so check the bottom row yourself. Proceeds: 203 plus 63 plus 150 plus 22 plus nothing is 438 crore. Cost released: 70 plus 45 plus 60 plus 10 plus 35 is 220 crore. And 438 divided by 220 is 1.99. The five exits together returned 1.99 times the cost they released, for Nilgiri Growth Partners Fund II, over the nine quarters from its Year 6 Quarter 3 to its Year 8 Quarter 3, and that sentence is the whole of what the figure says.

Route 1: a sale to a buyer already in the same industry

Holding 1 was Sahyadri Diagnostics Private Limited, invented. The fund paid Rs 55,00,00,000 for existing shares in Year 1 Quarter 3 and a further Rs 15,00,00,000 in Year 4 Quarter 1, a total cost of Rs 70,00,00,000. In Year 7 Quarter 2, after 5.75 years, the whole position went to a buyer already operating in the same industry for Rs 2,03,00,00,000. The proceeds are 2.90 times the cost and a profit of Rs 1,33,00,00,000.

The reason this route exists as a category at all is that the buyer is not only buying an investment. The buyer is also buying something that fits alongside a business it already runs. In this transaction the buyer expected to remove Rs 4,20,00,000 a year of duplicated cost once the two operations sat together, and on the multiple it was paying that stream was worth roughly Rs 29,00,00,000 to it. The highest bid from a buyer with no existing operation was Rs 1,75,00,00,000. The gap is Rs 28,00,00,000, being 16.0 per cent of that lower bid, and it is almost exactly the value of the duplicated cost the industry buyer could remove and the other bidder could not.

The gap has a stated reason, and the reason is not a rule. A buyer in the same industry does not always pay more, and any statement that it does would be false. Sometimes there is nothing to remove. Sometimes the industry buyer already has what the target has and wants only a piece of it. Sometimes there is no industry buyer willing to look at all. How a bid ceiling gets built, and why two buyers reach different ones, is worked in full under bid ceilings, and the result is used here in a sentence rather than re-derived.

Route 2: a sale of the whole position to another fund

Holding 2 was Konark Polymers Private Limited, invented, bought in Year 1 Quarter 4 for Rs 45,00,00,000 and sold in Year 6 Quarter 3, after 4.75 years, for Rs 63,00,00,000. The sale is 1.40 times cost, a profit of Rs 18,00,00,000, and it was the first thing this fund ever sold.

The buyer here was another fund: an investor buying a position as an investment, with a plan to hold it for a few years and sell it again. Look back at the comparison figure and notice how little separates route 2 from route 1 on the grid. All of the position leaves, in one transaction, on completion, and the fund holds nothing afterwards. The single row that differs is the first one, and the first row decides the price. A buyer who is only ever going to sell the company again is pricing that company on its own performance and nothing else.

Route 3: an offering, and then selling whatever is left

Holding 3 was Tungabhadra Logistics Private Limited, invented, bought in Year 2 Quarter 2 for Rs 60,00,00,000. In Year 8 Quarter 1 the company made an initial public offering. The fund received Rs 60,00,00,000 at the offering itself, and Rs 90,00,00,000 later, from selling the shares it had kept, once the shares could be sold. Rs 60,00,00,000 plus Rs 90,00,00,000 is Rs 1,50,00,00,000 in total, being 2.50 times cost over 5.75 years.

Route 3 is the one readers most often misread, and the misreading is always the same one: that a listing is the exit. It is not. A listing is the event that makes a later exit possible. At the offering the fund sold part of what it held and kept the rest, and the part it kept was not cash, was not immediately saleable, and moved in price every day until it was sold. The second stretch is the sell-downSelling the shares a fund still holds in a business after that business has listed., and in this holding it produced Rs 90,00,00,000, being 60.0 per cent of everything the holding ever returned.

Between the two events the fund held listed shares that it was not free to sell. A lock-in applied. How long any lock-in runs, who it binds and what conditions attach to it are matters for the Securities and Exchange Board of India at sebi.gov.in, and they change. The shape of the route survives every change in those rules: two cash events, separated by a period in which the position was quoted, visible, and not yet money.

How an offering is arranged, priced, subscribed and allotted is a separate subject and is covered separately. An offering matters only as a way out and for what it costs in time: for holding 3, a first instalment at Year 8 Quarter 1 and a second one after that, rather than one number on one day.

Route 4: a secondary sale of part of a position

Holding 9 was Indravati Packaging Private Limited, invented, bought in Year 5 Quarter 3 for Rs 25,00,00,000 and the last of the nine to be bought. In Year 8 Quarter 3, three years later, the fund sold 40 per cent of the position for Rs 22,00,00,000. Because 40 per cent of Rs 25,00,00,000 is Rs 10,00,00,000, that sale released Rs 10,00,00,000 of cost at 2.20 times. The remaining 60 per cent, cost Rs 15,00,00,000, stayed exactly where it was and is carried at Rs 33,00,00,000, also 2.20 times.

Everything distinctive about this route is in that last sentence. A part sale is the only route on the list where the fund is on both sides of its own position afterwards: it has been paid for some of the holding and is still marking the rest as an estimate. Rs 22,00,00,000 of holding 9 is a number a buyer paid. Rs 33,00,00,000 of it is a number Palani Valuation Advisors LLP, an invented limited liability partnership (LLP), and the manager between them produced. Same company, same day, two completely different kinds of figure, and a reader who adds them without noticing has mixed cash with opinion in a single total.

Who buys such a thing? An investor whose business is buying positions that already exist rather than making new ones. The market in positions that already exist, and the manager-initiated versions of it, are a subject in their own right and are covered separately. Here the point is structural: part of a position can leave without the rest of it going anywhere.

Route 5: the write-off, and why it belongs on this list at all

Holding 5 was Palar Foods Private Limited, invented, bought in Year 3 Quarter 1 for Rs 35,00,00,000 and written off in full in Year 6 Quarter 4, after 3.75 years. Proceeds nil. Multiple 0.00 times. No buyer, no negotiation, no completion.

A write-offRecording a holding as worth nothing, which removes it from the portfolio without producing any cash. feels like the absence of an exit rather than one of the five, and putting it on the list is a deliberate choice with a reason behind it. The position left the portfolio. The fund stopped carrying a value for it, stopped paying a management fee on its cost from the following year, and stopped having anything to sell. Everything that is structurally true of an exit is true here except the money.

Leaving the write-off off the list quietly teaches a reader that exits are good news, and that lesson is the single most expensive misreading in this whole subject. Five things left this portfolio and one of them returned nothing at all. A mental picture of an exit as a cheque arriving fits four of the five and is wrong about the fifth.

What is a traditional exit, and what is that word actually doing?

Traditional Exit

The phrase appears constantly in this subject, and it sounds like praise. Praise is not what the phrase carries. The phrase is a contrast, and once the other term of that contrast is clear, the word stops sounding like a compliment and starts being useful.

A traditional exitSelling a holding outright to somebody outside the fund, as distinct from moving it into another vehicle the same manager runs. means selling a holding outright to somebody who is not the fund and not the manager. The money comes in from outside, the position goes out to outside, and the relationship ends. All five routes worked here are traditional in that sense, including the write-off, where nothing comes in but the position still goes out and the relationship still ends. The word marks where the counterparty sits, not how well anything went.

The phrase is set against arrangements where the holding does not actually leave the manager. A fund near the end of its term can move a holding into a new vehicle that the same manager runs, with the existing investors offered a choice between taking cash and rolling into the new vehicle. A fund can also borrow against the holdings it still has, taken together, and pay the proceeds out, without selling anything at all. Neither of those has been done by Nilgiri Growth Partners Fund II. Both are worked in full, as clearly labelled counterfactuals, under the continuation vehicle and under borrowing against unsold holdings.

Why does the distinction matter enough to have its own word? Because in a traditional exit the price is set by somebody with no interest in the manager. In the arrangements it is contrasted with, the manager can end up on both sides of the same transaction, setting the price it sells at and the price it buys at. Sitting on both sides is a structural conflict, no amount of good intention solves it, and what answers it is process. The contents of that process are worked through separately. Here, the only thing to carry forward is that the word is describing who is on the other side.

How does the cash actually reach an investor once a holding is sold?

Not immediately, and not automatically. Between a signed agreement and money in an investor's bank account there are five distinct steps. A reader who thinks the fund controls all five will misread every delay they ever see, so it is worth being precise about who can make each step happen.

Now the specific rule that governs step 5 in this fund. Nilgiri Growth Partners Fund II distributes in the quarter after the cash arrives. Every time. The rule is one line in the fund's own documents, and it is the entire reason its five realisations became four distributions sitting in four different quarters.

Apply that one rule to all four cash realisations and the whole distribution schedule falls out of it without any further information.

Five realisations, four distributions, and the missing one is not an error in the schedule. Holding 5 produced no cash, so there was nothing to pay out, and no payment appears. Note also that the first two realisations are in adjacent quarters and produced completely different things: Year 6 Quarter 3 produced Rs 63,00,00,000 and a payment the following quarter, and Year 6 Quarter 4 produced nothing and no payment at all.

A notice and a payment are what actually land in an investor's hands at step 5. Because what such a notice leaves out surprises people, the notice is worth reading closely.

The absence of any such figure is the point. The notice tells an investor what arrived and where it came from. Whether the investment has done well is a separate calculation, made against everything ever paid in rather than against this one payment, and it is nowhere on the notice an investor actually receives.

Does an exit have to produce cash at all?

No, and the two events that prove it happened in this fund one quarter apart, close enough together to hold side by side. Whether an exit produces a payment to investors turns on a single question, and it is not a question about how good the transaction was or how large the business was or what anybody thought of the outcome.

The test is not whether the transaction was good, it is whether money moved. The test sounds obvious written down. In practice it is routinely lost because the word exit carries a whiff of completion that makes people assume a payment. Four of this fund's five exit events produced payments. The fifth removed a position, changed the portfolio, changed the fee the fund charged from the following year, and paid nobody anything.

The everyday version is a shop that closes. If the owner sells the lease and the fittings to somebody who wants to run a shop there, money changes hands and the household has something to spend. If the shop simply shuts and the stock is worthless, the shop has still gone, the rent has still stopped, and there is nothing to spend. Both are the end of the shop. Only one of them is money.

Why can one exit out of five decide most of what a fund returns?

Because the five are nothing like the same size, and no amount of counting will show that. Of the Rs 4,38,00,00,000 this fund has realised, holding 1 alone produced Rs 2,03,00,00,000. Dividing, 203 over 438 is 0.4635, so one holding out of nine produced 46.3 per cent of everything Nilgiri Growth Partners Fund II has ever realised to its record date at the end of Year 9 Quarter 2. Holding 3 produced another 34.2 per cent. Between them, two holdings produced 80.6 per cent of all the cash this fund has ever received back.

Taking either half of that picture on its own produces a wrong reading, so hold both at once. The same nine holdings that produced a single Rs 2,03,00,00,000 realisation also produced one position worth nothing and another carried at 0.70 times its cost, and both facts belong to the same portfolio and the same manager. Holding 6, Vaigai Edutech Private Limited, invented, cost Rs 30,00,00,000 and is carried at Rs 21,00,00,000. Holding 5 cost Rs 35,00,00,000 and returned nil.

There is nothing to celebrate and nothing to condemn in that shape. The shape is arithmetic belonging to one invented fund at one stated date. The shape does show something practical about how to read any such portfolio: an average across nine holdings would have described almost none of them, and the useful questions are about the sizes at the ends rather than the middle.

What is still unsold, and what does that mean at Year 9 Quarter 2?

Five exit events sounds like a fund that has nearly finished. Look at what is actually left and a different picture appears, and it is not a worse picture. The position is simply the honest one for a ten-year fund in its ninth year.

There are two different counts in play and they are easy to run together, so be careful with the counting. Five exit events happened, across five holdings. But only four holdings have left the portfolio entirely, being holdings 1, 2 and 3 which were sold and holding 5 which was written off. Only 40 per cent of holding 9 was sold, so holding 9 produced an exit event and is still in the portfolio. Four gone plus five still held is nine, and any sentence that says five gone and five held has counted holding 9 twice. The distinction is worth the trouble because the two counts carry different money. The four holdings that are gone entirely produced Rs 4,16,00,00,000 between them, being 203 plus 63 plus 150 plus nothing. The remaining Rs 22,00,00,000 of the Rs 4,38,00,00,000 came from holding 9, and holding 9 has not gone anywhere.

The Rs 2,82,00,00,000 is residual valueWhat a fund still reports as the value of the holdings it has not sold., and every rupee of it is the kind of number this guide opened by warning about. The residual value is produced annually by Palani Valuation Advisors LLP, the independent valuation agent, and marked quarterly by the manager in between. How that figure is struck, and what an investor's own statement of it looks like, are defined and worked separately. The figure has one purpose: to say plainly that 39.2 per cent of everything this fund is said to be worth has never been paid by anybody. The arithmetic checks: 282 over 720 is 0.3917.

Six quarters is not much time to sell five businesses, and the fund's documents allow two extensions of one year each, neither of which has been taken. A running clock beside unsold positions is exactly what creates the end-of-term pressure covered separately. The combination is not a defect in this invented fund. Every closed-end vehicle with a fixed life ends up with the same geometry.

One fund, one clock and one record date lie behind every figure in this worked case. Rs 5,00,00,00,000 was committed, Rs 4,80,00,00,000 has been drawn and Rs 20,00,00,000 is still uncalled, against Rs 4,00,00,00,000 of acquisition cost spread across nine holdings, each with its own entry date, cost, proceeds and multiple. Holding 1 sold for Rs 2,03,00,00,000 where the best bid from outside the industry was Rs 1,75,00,00,000, a gap of Rs 28,00,00,000 resting on Rs 4,20,00,000 a year of duplicated cost worth roughly Rs 29,00,00,000 to the buyer that could remove it. Holding 2 sold for Rs 63,00,00,000. Holding 3 returned Rs 60,00,00,000 at its offering and Rs 90,00,00,000 in the sell-down, making Rs 1,50,00,00,000 in all. Holding 9 gave up Rs 22,00,00,000 for 40 per cent of itself and leaves Rs 33,00,00,000 carried against the rest. Holding 5 returned nil. Rs 4,38,00,00,000 has reached investors in four payments and Rs 2,82,00,00,000 stands as residual value. Time runs from this fund's own final close, so Year 7 Q2 means the second quarter of its seventh year and never a calendar date.

How does somebody actually use any of this when a report lands on their desk?

The reader here may be an analyst at an institution that has committed to a fund like this one, a student handed a quarterly report in an interview, or somebody at a bank looking at a manager for the first time. The report will lead with a multiple and a rate of return. Four moves change what happens to that report, and none of them needs an opinion about anybody.

Move one: split the reported value into what was paid and what is estimated. For this fund that is Rs 4,38,00,00,000 against Rs 2,82,00,00,000 out of a stated Rs 7,20,00,00,000 of total value. Nothing about the second figure is dishonest. Residual value is just a different kind of number, and a report that presents one total has blended two.

Move two: the exits taken one by one rather than as a count. Five exits and Rs 4,38,00,00,000 says almost nothing until it is clear that Rs 2,03,00,00,000 of it came from one holding and nothing at all came from another. The carrying valueThe figure a fund reports for a particular holding it has not sold. of what remains matters for the same reason: five unsold holdings at Rs 2,82,00,00,000 could be five similar positions or one large one and four small.

Move three: each realisation read against the route it used. The routes show where the remaining risk sits. A position sold outright to a buyer in the same industry is finished. A position that has listed but not been fully sold down is not finished because what is left moves in price. A partly sold position is half finished by definition. In this fund, at the moment holding 3 listed at Year 8 Quarter 1, Rs 90,00,00,000 of its eventual Rs 1,50,00,00,000 had not yet been received.

Move four: the clock set beside the list. Six quarters of contracted term against five unsold holdings is a fact worth writing down, not because it predicts anything, but because it settles which questions the next report will have to answer. A lender assessing the manager, an investor deciding whether to commit to a later fund, and an analyst writing a review all end up asking the same thing: what has to happen between here and the end of the term, and who has to agree to it?