How should a fund ops team forecast DPI for LPs when the big exit is an IPO still under lockup?

Q: When does an IPO actually count toward DPI?

Only when cash or shares reach the LPs. Until then, a higher mark on the public stock raises TVPI and RVPI, not DPI. ILPA's reporting definitions count in-kind stock distributions as distributions, valued when they're made, so a distribution of shares counts once it happens, not the day the company lists.

Q: What should the DPI forecast you send LPs include?

One row per tranche of shares, with the share count, the earliest date the fund can sell or distribute, the price range you assume and the source for each number. Add a base case and a down case, the carry and recycling assumptions, and a note on what would move the dates. Keep it short.

Treat the IPO as a schedule, not a distribution. Forecast the cash in tranches: when the lockup ends, when the fund can actually sell or distribute, and at what range of prices. Show LPs a base case and a down case, net of carry, and tie every date and share count to the prospectus.

I'll admit my first reaction when a portfolio company files to go public is the same as everyone's: look at that TVPI. The ops question comes a beat later, usually from an LP. When does any of this turn into cash? That question is the one this post is about.

Why LPs are asking now

Distributions have been thin for years. Carta's data on 2021-vintage funds found that only about a quarter of them had returned any cash at all through Q3 2025. So when a big holding lists, LPs read it as the moment the fund finally pays out, and they want a date.

The honest answer is "not yet". The IPO day is the start of a schedule, and the fund's own position usually can't move for months. A forecast that says so, with dates and sources, does more for LP trust than an optimistic sentence in the quarterly letter.

When does an IPO actually count toward DPI?

Only when cash or shares reach the LPs. DPI is distributions divided by paid-in capital, and a listing distributes nothing by itself. Until the fund sells or hands out shares, a higher mark on the public stock raises TVPI and RVPI. DPI stays where it was.

In-kind distributions do count. ILPA's reporting template definitions treat stock distributions as distributions, valued when they're made. So DPI moves on the day the shares go out, not the day the company lists. (That also means the LP's eventual sale price can differ from the value you reported, which is worth saying out loud.)

ILPA's Performance Template, released in January 2025 for data starting in Q1 2026, standardizes how DPI, TVPI and IRR get reported. If your LPs are asking for it, the forecast should use the same definitions.

Step 1: Build the tranche table from the prospectus

Start from documents, not from memory. The registration statement discloses the lockup terms, so pull them from there and save the page reference next to each number.

For each block of shares the fund holds, record:

Share count, from the fund's ledger and the cap table, reconciled to the prospectus.
Lockup end date. The SEC's investor guidance says most lockups run 180 days, but check the actual terms. Some lockups include early-release clauses tied to the stock price or to earnings, and some release shares in stages.
Other selling limits. If the fund counts as an affiliate (a board seat often makes it one), Rule 144 limits how much it can sell in each 90-day window. Shares bought before the IPO are restricted securities, with their own holding-period rules.
Insider trading windows. A partner on the board means the fund may only be able to trade in open windows after earnings.
The source for each line: prospectus page, ledger entry, counsel's note.

This is the boring part, and it's the part that makes the forecast trustworthy. Every number in the table should trace to a document someone can open.

Step 2: Turn the schedule into cash, with a range

Now put prices on it. Don't forecast a single number. Use at least two cases:

Case	Price assumption	Timing assumption
Base case	A discount to today's price, held flat	Sell or distribute soon after each lockup end
Down case	A meaningful drop by the first lockup end	Sales spread over several 90-day windows

Stocks often slip ahead of a lockup expiry as the market prices in the extra supply, and the SEC's guidance says so directly. So the base case shouldn't assume today's price either.

Then decide, per tranche, whether the fund plans to sell and distribute cash or distribute shares in kind. That decision belongs to the partners and the LPA, not to the forecast. The forecast just needs to say which path each tranche assumes, because the timing differs.

How do carry and recycling change the DPI forecast?

LPs care about the cash they receive, so forecast DPI after carry. If the fund is past its hurdle, part of each distribution goes to the GP, and the LP's DPI rises by less than the gross proceeds suggest. If the fund isn't past its hurdle yet, the first distributions may go entirely to LPs. Run the waterfall from the LPA's actual terms. A rule of thumb ("it's 20%, call it 0.8x") gets the timing wrong in the cases that matter most.

Check recycling too. Some LPAs let the fund keep part of its proceeds to make new investments or to cover fees. That cash never shows up in DPI, and an LP who hears "big exit" will notice.

Your fund admin has the waterfall model. Ask them to run the forecast scenarios through it, so the numbers in the LP letter match the numbers on the capital account statements later.

What should the DPI forecast you send LPs include?

Keep it to a page. One row per tranche, with:

the share count and its source
the earliest date the fund can sell or distribute, and why
the price range assumed in each case
the expected DPI after carry, in each case

Add a short note on what would move the dates: an early release, a follow-on offering, a trading window that closes. Then say plainly that it's a forecast. (LPs have seen plenty of them. What they remember is whether you updated it when things changed.)

A model can help draft the explanation around the table. It shouldn't produce the numbers. The share counts come from the ledger, the dates come from the prospectus, and the waterfall comes from the fund admin. A person checks each line against its source before it goes out.

What this looks like in practice

The funds that handle this well don't treat the IPO as one event. They treat it like any other change to a position: it has dates, share counts and a document behind each one. When the lockup ends or a tranche goes out, they update the row and the forecast moves with it.

At Cura, each position's rounds and share counts trace back to a signed source document, which is the input this table needs. The forecast itself, and the decision to sell or distribute, stay with the people who answer to the LPs.