When a fund sells part of a holding it bought in tranches, the cost it relieves decides the realised gain, the realised multiple and, under a deal-by-deal waterfall, when carry is paid.
On a partial exit the fund relieves the cost of the shares sold under its stated method, books net proceeds less that cost as realised gain, and reverses the unrealised gain it was carrying on those shares. Selling 20.0m of 52.0m shares, bought in two tranches at $1.00 and $2.00, for $55.0m net books a realised gain of $30.38m on average cost, $35.00m on first-in first-out and $23.00m highest-cost first. NAV moves by the same $5.0m under all three.
Worked in full in The Private Equity Fund Controller Playbook by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
A full exit is easy: all the cost goes. A partial exit, a secondary sale of a block at IPO, a sell-down to a co-investor, a partial recapitalisation, forces a choice the full exit never does: which cost goes out with the shares sold. When the holding was bought in one tranche the answer is a simple fraction. When there was a follow-on at a higher price, the answer changes the realised gain, the realised multiple reported to investors and, under a deal-by-deal waterfall, the timing of carry.
| Item | Shares, m | Price, $ | Cost, $m |
|---|---|---|---|
| Initial investment | 40.0 | 1.00 | 40.0 |
| Follow-on, three years later | 12.0 | 2.00 | 24.0 |
| Holding | 52.0 | 1.2308 | 64.0 |
| Fair value at the prior quarter end | 52.0 | 2.50 | 130.0 |
| Shares sold in the partial exit | 20.0 | 2.80 | 56.0 gross |
Average cost: 64.0 ÷ 52.0 = $1.2308 a share; 20.0 × 1.2308 = $24.62m
First in, first out: all 20.0m from the initial tranche at $1.00 = $20.00m
Highest cost first: 12.0m at $2.00 plus 8.0m at $1.00 = $32.00m
In Excel, average cost: =Shares_Sold*Total_Cost/Total_Shares. For tranche methods, a running balance of shares and cost per tranche, depleted in the chosen order.
| $m | Average cost | FIFO | Highest cost first |
|---|---|---|---|
| Net proceeds | 55.00 | 55.00 | 55.00 |
| Cost relieved | 24.62 | 20.00 | 32.00 |
| Realised gain | 30.38 | 35.00 | 23.00 |
| Reversal of unrealised gain on the shares sold | −25.38 | −30.00 | −18.00 |
| Effect on NAV in the quarter | 5.00 | 5.00 | 5.00 |
| Realised multiple on the shares sold | 2.23x | 2.75x | 1.72x |
The reversal is the prior fair value of the shares sold, 20.0 × $2.50 = $50.0m, less the cost relieved. Realised gain plus reversal always equals net proceeds less prior fair value, $5.0m, so the method cannot change NAV. That identity is the control: if the two lines do not net to proceeds less prior mark, the cost relief is wrong. The single-tranche version of these entries is in realised vs unrealised gain on a fund exit.
| $m | Average cost | FIFO | Highest cost first |
|---|---|---|---|
| Remaining shares, m | 32.0 | 32.0 | 32.0 |
| Remaining fair value | 80.0 | 80.0 | 80.0 |
| Remaining cost | 39.38 | 44.00 | 32.00 |
| Remaining unrealised gain | 40.62 | 36.00 | 48.00 |
| Unrealised multiple on the rest | 2.03x | 1.82x | 2.50x |
Average cost is the neutral method: the remaining shares keep the holding's cost of $1.2308 a share and the unrealised multiple stays at 2.03x, exactly where it was before the sale. The tranche methods move performance between the realised and unrealised halves of the same investment. FIFO makes the exit look excellent and the residual stake mediocre; highest cost first does the reverse.
Under a whole-of-fund waterfall the method changes nothing in cash: distributions go to the LPs until contributed capital and the preferred return are met, whatever gain the accounts call realised. Under a deal-by-deal waterfall, carry on a partial realisation is computed on the gain attributed to the shares sold. Before any preferred return, 20 per cent of the realised gain is $6.08m on average cost, $7.00m on FIFO and $4.60m highest cost first. FIFO pulls $2.40m of carry forward, which is exposure to clawback if later exits disappoint; the mechanics are in European versus American waterfall: the clawback.
Realised multiples also feed the track record that the next fund is marketed on. A 2.75x "realised" on a partial exit that sold only the cheapest shares is a figure a diligent LP will rebuild on average cost, and get 2.23x.
Suppose the remaining 32.0m shares are sold later at $3.00, for $96.0m. The total gain over the life of the investment is fixed by cash in and cash out, and the method only decides how it is split between the two exits:
| $m | Gain, partial exit | Gain, final exit | Total |
|---|---|---|---|
| Average cost | 30.38 | 56.62 | 87.00 |
| FIFO | 35.00 | 52.00 | 87.00 |
| Highest cost first | 23.00 | 64.00 | 87.00 |
The $12.00m spread between FIFO and highest cost first on the first sale is pure timing. It reverses on the second. For a fund that reports realised and unrealised gains separately, and for carry computed deal by deal, timing is the whole issue.
The usual error is not choosing the wrong method but switching between them: average cost in the accounts, FIFO in the carry calculation and a pro rata of the original tranche in the investor report. Each one is defensible; together they produce three realised gains for one sale, and an auditor will ask which is the policy. A second error is to book the $5.0m uplift on the prior mark as the realised gain, relieving the shares at their $50.0m fair value instead of their cost; it leaves the realised line $25.38m short on average cost and strands that gain in unrealised appreciation on shares the fund no longer owns. Write the method into the valuation and accounting policy, apply it in every report, and keep the tranche ledger even if the policy is average cost, because the LPA's deal-by-deal waterfall may require tranche dates for the preferred return.
On a partial exit the cost method decides the realised gain and the realised multiple, never NAV and never the total gain. Average cost is neutral; tranche methods shift gain and, under a deal-by-deal waterfall, carry between exits. The book's quarter-end close, with its journal entries and NAV bridge, is in the free workbook for this case, and the NAV build itself is in how to calculate the NAV of a private equity fund.
No. NAV changes only by net proceeds less the prior fair value of the shares sold, $55.0m less $50.0m, or $5.0m, in the illustrative case. The method moves gain between the realised and unrealised lines: average cost books $30.38m realised and reverses $25.38m of unrealised; first-in first-out books $35.00m and reverses $30.00m.
Either can be acceptable if the accounting policy states it and it is applied consistently; many funds use average cost or specific identification of tranches. Average cost is the most neutral: it keeps the remaining position's cost per share at $1.2308 and its unrealised multiple at 2.03x, the same as before the sale. FIFO flatters the realised multiple, 2.75x against 2.23x.
Under a deal-by-deal waterfall, carry on a partial realisation is computed on the gain attributed to the shares sold. Before any preferred return, 20 per cent of the gain is $7.00m on FIFO and $4.60m on highest-cost first: the method pulls $2.40m of carry forward or pushes it back. Over the full exit the total gain is the same $87.00m.
This article is one calculation from The Private Equity Fund Controller Playbook. 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 1682 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.