One fund's since-inception IRR recomputed every year with the NAV 20 per cent lower and higher, to see when the number starts to mean something.
A fund's since-inception IRR becomes meaningful when the NAV no longer drives it, which on an illustrative fund is around year eight. Recompute the IRR with the NAV 20 per cent lower and 20 per cent higher: the answer swings by about 21.7 points either way in year 2, 7.9 in year 4, 3.3 in year 6 and 1.1 in year 8. The fund finishes at 1.70x and 10.15 per cent; its year-4 IRR, 7.8 per cent, could honestly have been anything from −0.7 to 15.1.
Worked in full in Private Markets Performance by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
An interim IRR is an ordinary IRR with one artificial cash flow: the NAV, entered as if the investor sold its interest on the report date at the manager's valuation. That is a reasonable convention. The question is how much of the answer comes from that one entry, and the simplest way to find out is to perturb it.
| Year | Calls | Distributions |
|---|---|---|
| 0 to 5 | 20, 25, 25, 15, 10, 5 | |
| 3 to 11 | 5, 10, 15, 25, 30, 30, 25, 20, 10 | |
| Total | 100 | 170 |
Final result: TVPI 1.70x, IRR 10.15 per cent. Each year's NAV is set as the value of the fund's remaining net cash flows discounted at 12 per cent, a deliberately prudent valuer who marks below the fund's eventual return. The test applies a 20 per cent error to that NAV in each direction, illustratively the scale of difference between a valuation and an exit price on a single company, here applied to the whole portfolio.
Interim IRR at age t = IRR(calls and distributions to t, plus NAVt × (1 + e) as a final inflow), for e = −20%, 0 and +20%.
In Excel, copy the flows to date into a helper column, add =C6*0.8 (the NAV less 20 per cent) to the last year's flow, and take =IRR(D2:D6); repeat with 1.2 for the upper case.
| Year | DPI | RVPI | TVPI | NAV share of value | IRR | NAV −20% | NAV +20% | ± points |
|---|---|---|---|---|---|---|---|---|
| 2 | 0.00 | 0.99 | 0.99 | 100% | −1.1% | −24.2% | 19.1% | 21.7 |
| 3 | 0.06 | 1.03 | 1.09 | 95% | 5.4% | −7.7% | 16.8% | 12.2 |
| 4 | 0.16 | 1.03 | 1.19 | 87% | 7.8% | −0.7% | 15.1% | 7.9 |
| 5 | 0.30 | 1.00 | 1.30 | 77% | 8.9% | 3.2% | 13.8% | 5.3 |
| 6 | 0.55 | 0.87 | 1.42 | 61% | 9.5% | 6.0% | 12.6% | 3.3 |
| 7 | 0.85 | 0.67 | 1.52 | 44% | 9.8% | 7.7% | 11.7% | 2.0 |
| 8 | 1.15 | 0.45 | 1.60 | 28% | 10.0% | 8.8% | 11.1% | 1.1 |
| 9 | 1.40 | 0.26 | 1.66 | 16% | 10.1% | 9.5% | 10.6% | 0.5 |
| 10 | 1.60 | 0.09 | 1.69 | 5% | 10.1% | 10.0% | 10.3% | 0.2 |
Two forces shrink the band. The NAV becomes a smaller share of total value as distributions arrive, from 87 per cent in year 4 to 28 per cent in year 8. And the holding period lengthens, so a given error in value is spread over more years of rate. Before year 5 both work against the reader: the NAV is nearly everything and the clock has barely started, which is why the year-2 IRR can sit anywhere between −24.2 and 19.1 per cent on an error no valuer would call unusual.
On these assumptions the band first falls within one point either way in year 9. By year 6 the central IRR is already close, 9.5 against a final 10.15, because the prudent valuer's bias matters less as the NAV shrinks; but a 20 per cent error either way still leaves 6.0 to 12.6 per cent.
The same test shows why TVPI feels steadier. In year 4 the 20 per cent error moves TVPI by 0.21 of a turn on 1.19x and the IRR by 7.9 points on 7.8 per cent: the multiple has no clock to amplify the error. By year 8 the TVPI swing is 0.09.
The perturbation is cheap enough to run every quarter on every fund. Report the IRR with its band beside it, for example "9.5 per cent (6.0 to 12.6 on a 20 per cent NAV range)" in year 6, and the reader knows how much weight the number can bear. The size of the error is a choice: 20 per cent is a round illustration, and a credit fund with observable loan prices would justify a smaller one than a venture fund. What matters is applying the same band across funds so that a young fund's IRR is never read with the same confidence as a mature one's.
An interim IRR is only as good as the NAV inside it, and for most of a fund's first half the NAV is most of the answer. Stress it: if a 20 per cent change in NAV moves the IRR by more than a few points, report the multiples and the realised share first. Private Markets Performance measures one fund from the cash flow series to the reporting line, and the free workbook for this case lets you run the same test on your own fund's cash flows.
Because the NAV is entered as a final cash flow, and early on it is almost all of the value. In year 4 of the illustrative fund the NAV is 87 per cent of total value: a 20 per cent error in it moves the IRR between -0.7 and 15.1 per cent. The cash flows to date say very little; the valuation says nearly everything.
It is less sensitive, not more reliable. A 20 per cent NAV error moves the year-4 TVPI by 0.21 of a turn, about 17 per cent of its total value, while it moves the IRR by 7.9 points either side of 7.8 per cent. Both rest on the same NAV; the IRR amplifies the error because the holding period is short.
Conventions vary, and many investors look at quartile rankings only from around year six. On this illustrative fund the interim IRR is within 0.7 of a point of the final 10.15 per cent by year 6, but the uncertainty band from a 20 per cent NAV error is still 3.3 points either way; it narrows to 1.1 by year 8.
This article is one calculation from Private Markets Performance. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: private equity and private markets → · All 1340 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.