The J-Curve: Why Private Fund Returns Start Negative

A private fund's reported return starts below zero and stays there for years, and nothing has gone wrong. Fees and expenses are drawn from the first quarter, holdings sit at what was paid, and no buyer has paid for anything yet. Nilgiri Growth Partners Fund II, invented, reported 0.88 times and minus 47.2 per cent at its own Year 1 year end, and passed 1.00 times and zero per cent during its Year 6.

Think about a couple who have just bought a small shop with borrowed money and their own savings. In the first month they pay the deposit, the transfer charges, the lawyer and the signboard painter. Nothing has been sold yet. If somebody asked them that evening what their venture is worth against what they have put in, the honest answer is less, and the shortfall is exactly the signboard and the lawyer. Nobody would call that a bad shop. A closed-end private fund reports its first years from exactly where that couple stood, and the shape its reported figures trace over its life has a name because almost every such fund traces it. One invented fund's whole record is set out below, year by year, and a single year of it settles nothing.

Why is the first number a private fund reports a bad one?

Because a fund spends before it earns, and it is required to report that. The J-curveThe shape a closed-end fund's reported return traces: down first, then up. is the picture that spending-first produces when the reported figures are plotted against time. The curve falls in the early years, flattens along a long bottom, then turns up somewhere in the second half of the fund's life. The letter it resembles is where the name comes from and it is the only thing about the name worth remembering.

Nilgiri Growth Partners Fund II, invented, is managed by Nilgiri Alternatives Advisors Private Limited, invented, and registered as an Alternative Investment Fund with the Securities and Exchange Board of India. 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 acquisition cost, drew Rs 4,80,00,00,000 from investors in seventeen capital calls, and returned Rs 4,38,00,00,000 in four distributions to its record date. That record date is the end of its own Year 9 Quarter 2 and sits 8.50 years after its final close. Every year quoted is counted from that final close and never from a calendar date.

The shape confuses because the fund is not doing anything unusual while the shape is being drawn. The fund is buying companies, paying its own bills, and waiting. In the early years there is not yet any judgement to report on, so the negative reading is not a report on the manager's judgement. There is only money that has gone out and value that has not yet been tested by anybody paying for it.

What exactly is pulling the early years down?

Three separate things, and the useful discovery is that all three push in the same direction at the same time. Explain away any one of the three and the reported figure is still below where a reader expects it to be. The shape therefore survives every individual excuse made for it.

Force one is the money that leaves before anything is bought. The very first capital call of Nilgiri Growth Partners Fund II, invented, was Rs 13,10,00,000 at its own Year 1 Quarter 1, and not a single rupee of it bought a company: it was the fund's organisational expenses, its first year of management fee, and its first year of operating costs. The Rs 13,10,00,000 sits in the denominator of every ratio the fund will ever report, and it never bought an asset.

Force two is that the companies actually bought are carried at costRecorded at what was paid, because nothing has happened that would justify moving it.. A company bought last quarter has not published anything, has not been sold, and has not done anything a valuer could point at. So the fund reports it at what it paid. Carrying at cost is not conservatism and it is not a policy choice made to look careful. It is what the evidence supports. The consequence is that the numerator sits still while the denominator grows.

Force three is that nothing has been sold, so the column that shows cash returned to investors is empty. Nilgiri Growth Partners Fund II, invented, distributed nothing at all in its own Years 1 to 5. Not a small amount. Nothing.

How long does it stay below, and what does the whole record look like?

Before the record itself, one habit worth building. A private fund's performance record has six columns and they are two different kinds of thing. Three of them are counted: how much has been called from investors, how much has been paid back, and what the unsold holdings are carried at. Three of them are divided: each is one of the first three over another of them. Almost every misreading of a private fund's report starts by treating a ratio as though it were a fact that somebody observed.

Here are the three counted columns for Nilgiri Growth Partners Fund II, invented, at each of its own year ends and then at its record date.

And here are the three computed from them, for the same invented fund at the same nine dates. The first is total value to paid inCash distributed plus carrying value, divided by capital paid in., the second is distributions to paid inCash distributed divided by capital paid in. The ratio ignores everything not yet sold., and the third is the net internal rate of return since inceptionMeasured from the fund's own final close rather than over a single year., computed on the dated cash flows with the net asset valueThe total carrying value of everything the fund still holds. treated at each date as a terminal valueThe carrying value treated as a final cash flow at the measuring date, as though everything were sold that day for its carrying amount..

Nilgiri Growth Partners Fund II, invented, spent five full years below both 1.00 times and zero per cent, and nothing had gone wrong in any of them. Plotted, those two rows are the shape the J-curve is named after.

What makes the early years readable, once the paid in column is split?

The cumulative paid in column splits into two parts: what went into companies, and what went on running the fund. Split that way for Nilgiri Growth Partners Fund II, invented, the early years stop being a mystery. The table below is the same nine-date record with the paid in column taken apart, and the net asset value set beside the first part of it.

The first two figures in any row add to that row's cumulative paid in from the earlier table, to the rupee. The four highlighted rows carry the point. For four full years the net asset value of Nilgiri Growth Partners Fund II, invented, equals the acquisition cost of everything it had bought, to the rupee: Rs 1,00,00,00,000 against Rs 1,00,00,00,000, then Rs 2,10,00,00,000, then Rs 2,75,00,00,000, then Rs 3,65,00,00,000. Not one holding had been marked up or down in any of those four years. The Year 4 figure already includes the Rs 15,00,00,000 the fund put back into its first holding at its own Year 4 Quarter 1, and even with that added the two columns still match exactly.

Which means the whole shortfall against paid in over those four years is the fee and the expenses and nothing else. Rs 13,10,00,000, then Rs 23,70,00,000, then Rs 34,30,00,000, then Rs 44,90,00,000. There is a clean identity hiding in that: over those four years the multiple this fund reported is the cost of what it had bought divided by that cost plus the fee. Nothing about performance enters it at all. The identity is why the multiple sat at 0.88, then 0.90, then 0.89, then 0.89 for Nilgiri Growth Partners Fund II, invented. It wobbled by a point or two because the fee's share of the paid in wobbled, and for no other reason.

The first mark away from cost appears at that fund's Year 5 year end, where Rs 4,34,00,00,000 of net asset value sits on a Rs 4,00,00,00,000 book. The Year 5 mark is the first moment in this fund's whole life at which any part of the reported figure is about the companies rather than about the cost of running the vehicle.

Where does the curve cross, and does everything cross at once?

A crossingThe moment a reported figure passes its reference point: 1.00 times for a multiple, zero per cent for a rate. is just the moment a reported figure passes its reference point. For Nilgiri Growth Partners Fund II, invented, the multiple went from 0.95 times at its own Year 5 year end to 1.06 times at its own Year 6 year end, and the rate of return went from minus 1.8 per cent to plus 1.7 per cent over the same twelve months. Both crossings therefore fall inside that fund's Year 6.

There is a limit to what the record can establish. The record reports at year ends, so it fixes the crossings inside a year and no more finely than that. Anybody naming the exact quarter a fund crossed is reading something the annual record does not contain. The two crossings landing in the same year is not a coincidence, though: the fund's first realisation and its first serious upward marks arrived in the same stretch of its life, and both measures are computed off the same three counted columns.

What actually crossed it, and what did one realisation really do?

Nilgiri Growth Partners Fund II, invented, made its first distribution of Rs 63,00,00,000 at its own Year 6 Quarter 4, being the proceeds of the second of its nine holdings. Watch what that did and what it did not do.

The distribution moved the distributions ratio of Nilgiri Growth Partners Fund II, invented, from 0.00 at its Year 5 year end to 0.14 at its Year 6 year end. Registering realised cash is exactly what that ratio is for. The multiple already counted that value, so the distribution barely moved it. While the holding was still held the value sat in the net asset value column, once sold it sat in the distributed column, and the multiple adds the two together. A realisation moves value from one column into another rather than creating any, so a reader watching only the multiple will barely see the largest single event of a fund's year.

So what did move the multiple of Nilgiri Growth Partners Fund II, invented, from 0.95 times to 1.06 times over its own Year 6? The numerator went from Rs 4,34,00,00,000 to Rs 4,94,00,00,000, a rise of Rs 60,00,00,000, and the distribution contributed nothing to that rise. The rise is what happened to the marks over that year, and it happened in a year that also contained the write-off of one of the nine holdings in full. Both of those are in the same Rs 60,00,00,000.

Can a multiple fall while nothing at all happens to the holdings?

Yes, and this fund shows it in the cleanest possible way. Between the Year 8 year end of Nilgiri Growth Partners Fund II, invented, and its record date two quarters later, its total value to paid in fell from 1.51 times to 1.50 times.

Nothing was sold in those two quarters, nothing was distributed, and not one carrying value moved: the net asset value is Rs 2,82,00,00,000 on both dates and the cumulative distributed is Rs 4,38,00,00,000 on both dates. The numerator of the multiple is therefore Rs 7,20,00,00,000 on both dates and it did not change by a rupee. The change is underneath the multiple, in its denominator. The fund made one more capital call, drawdown seventeen, of Rs 2,20,00,000 at its own Year 9 Quarter 1, for the fee and expenses of that year. Paid in went from Rs 4,77,80,00,000 to Rs 4,80,00,00,000. The same numerator over a bigger denominator is a smaller number, and that is the entire story of the fall from 1.51 times to 1.50 times.

Notice how small it is. The unrounded readings of Nilgiri Growth Partners Fund II, invented, are 1.5069 times at its Year 8 year end and 1.5000 times at its record date, a difference of about seven thousandths, and the only reason it shows up at all is that the two round to different figures at two decimal places.

Can a rate of return fall while nothing at all happens to the holdings?

Also yes, and for a different reason. Over exactly the same two quarters, the net rate of return since inception of Nilgiri Growth Partners Fund II, invented, fell from plus 8.7 per cent at its Year 8 year end to plus 8.3 per cent at its record date.

A multiple has no clock in it. A rate of return is nothing but a clock: it asks at what annual rate the money that went out would have had to grow to produce what came back, by the date being measured to. The same Rs 2,82,00,00,000 of carrying value that was being counted at 8.00 years is now being counted at 8.50 years, and the arithmetic has to spread the same result over a longer stretch. Holding an unchanged value for longer lowers a rate of return by construction. It would do so even if the fund had done absolutely nothing in those two quarters, and doing nothing is very nearly what the fund did.

A reader who has understood the multiple can still get caught here. The two measures fell in the same two quarters for two entirely different reasons, and neither reason has anything to do with any company. One was the denominator. One was the calendar.

How is a fall that is arithmetic told apart from a fall that is a loss?

Two questions, in order, and they work every time because a ratio only has two places a change can come from.

First: did the numerator fall, or did the denominator rise? For a multiple the numerator is cash distributed plus carrying value, and the denominator is capital paid in. If the numerator is unchanged and the denominator has grown, the fall is arithmetic and the question is answered. Second, if the numerator did fall: was any holding actually marked down? A numerator can fall for exactly one reason. Something in the carrying value came off, and a fund's own report names which holding and by how much. If the denominator rose and nothing was marked down, then nothing was lost.

Applied to the two falls set out here, both resolve in one step. For the multiple of Nilgiri Growth Partners Fund II, invented, between its Year 8 year end and its record date, the numerator was Rs 7,20,00,00,000 on both dates, so the answer is denominator and the test ends there. For its rate of return over the same two quarters, both the numerator and the denominator of the multiple were unchanged, so the fall is neither of them. A rate of return has only one other input, and that input is elapsed time.

What does this do in a review meeting?

More people read a private fund's report than the subject suggests: somebody on an investment committee deciding whether to commit again, a monitoring team inside a large investor, an analyst covering an institution that has made several commitments, an auditor, and a student who will be doing one of those jobs in three years. The shape buys each of them something specific, and none of it requires an opinion about anybody.

First, the shape settles what a single reported figure is worth on its own. Without a date attached to it, a reported figure is worth very little. A reading of 0.89 times means one thing at a fund's third year and something entirely different at its ninth, and the figure itself carries no clue which. So the first thing to write beside any private fund number is how far into that fund's own life it was taken. Second, it shows where to look when a figure moves. The three counted columns are short and a report prints all three; two subtractions between two dates settle whether the movement came from a call, a distribution or a mark, and that is a five minute job rather than a conversation.

Third, it changes what an early comparison between two funds can honestly say. One of two funds at different points on their own clocks has more of its fee already in its denominator and less of its life behind it, so the two are not comparable on a reported multiple at all. The fix is not a cleverer ratio; it is refusing to put the two figures side by side until each fund's own elapsed years have been written next to its own number. Fourth, and most usefully, it settles which questions the report can answer and which it cannot. A fund's report can say exactly why a multiple fell. The report cannot say what any holding will eventually sell for, and no shape anywhere can either.

What does the shape promise about what comes next?

Nothing whatsoever, and that is not a hedge, it is the honest content of the record. The record is nine dated readings from one invented fund over its own 8.50 years. The last one is Rs 4,80,00,00,000 paid in, Rs 4,38,00,00,000 distributed and Rs 2,82,00,00,000 of carrying value across five holdings that nobody has bought.

The carrying value is the reason no shape carries a promise. Rs 2,82,00,00,000 of the reported total value of Nilgiri Growth Partners Fund II, invented, at its record date is a carrying value rather than cash, and the only way it becomes a fact is that somebody pays for it. The turn in this fund's curve after its own Year 6 is a description of what happened, in a record that stops on the day it stops, and it is not a shape that anything is required to continue. Six quarters of that fund's contracted ten-year term remained at its record date.