Every LP conversation this year arrives at the same question. Where is the cash?
The IRRs look fine. The marks look fine. The funds report unrealized gains that keep the headline numbers respectable. But the distributions are not landing. Capital went out, paper value went up, and the wire transfers back to limited partners slowed to a trickle. That gap between what a fund says it is worth and what it has actually returned is the whole story of private equity in 2026.
The cleanest way to see it is vintage by vintage. DPI, distributions to paid-in capital, tells you how much of an investor’s committed money has actually come back. Not the mark. The cash. Walk the recent vintages in order and the drought is obvious.
2016 was the last vintage to give the money back
Start with the number that should worry everyone. The last US buyout vintage to reach 1.0x DPI was 2016. Bain, McKinsey, and MSCI converge on this point. A 1.0x DPI means the fund has returned the capital that was put into it. Not a profit yet. Just the original money back.
So a vintage that is now a decade old is the most recent one that has cleared that bar. Everything raised after 2016 is still, on a cash basis, in the hole to its investors.
That is not how the asset class is supposed to work. A 2016 fund hitting 1.0x in roughly its tenth year is already slow by historical standards. The vintages behind it are slower still. I wrote about why this metric has become the one that matters in The DPI Reckoning. The short version is that LPs have stopped accepting IRR and TVPI as proof of anything. They want the cash, and the cash is not there.
2017 and 2018: bought high, holding long
The 2017 and 2018 vintages are where the squeeze gets specific.
These funds deployed into a market with rising entry multiples. They bought well-run companies at full prices on the assumption that multiple expansion and a liquid exit market would do the rest. Then the exit market closed. Rates moved, debt got expensive, and the buyers who would have taken these companies off their hands either disappeared or repriced.
The result shows up in the DPI. The 2018 vintages sit around 0.6x DPI against a benchmark closer to 0.8x. Read that carefully. A 2018 fund is now seven or eight years into its life and has returned about sixty cents on the dollar. The benchmark it is being measured against is itself depressed, and these funds are trailing even that.
So you have a vintage that bought at high prices, held longer than planned, and has given back well under what its own peer group expected. The companies are not necessarily bad. Many are performing. They are simply stuck. The GP cannot sell them at a price that clears the model, and every quarter of holding burns more of the fund’s life.
2019 to 2021: marked up, not cashed out
The vintages right behind them are the ones that look best on paper and feel worst in the bank account.
Funds raised between 2019 and 2021 deployed into the cheapest debt in living memory and then into the 2021 deal frenzy. The marks went up fast. TVPI on these vintages can look genuinely strong because the unrealized portfolio is carried at healthy multiples. But DPI on these vintages is minimal. The cash has not moved.
This is the cruelest part of the drought for an LP. The fund tells you it is up. The capital account shows gains. And yet nothing has come back, because almost none of it has been sold. A markup is a promise. A distribution is a fact. Right now the industry is rich in promises and poor in facts.
The 2021 cohort is the sharpest example. Those companies were bought at the top of the multiple cycle. To return capital, the GP needs an exit market willing to pay something near what they paid. That market does not exist yet. So the 2021 vintage waits, and the LP waits with it.
Why the cash is stuck
Pull the camera back and the cause is the same across every vintage. There are too many companies that need to be sold and not enough motion to sell them.
The exit market backed up. Sponsors who would normally have flipped a company at year four or five are now holding into year six, seven, and beyond, hoping for a better window. The result is a backlog of unsold portfolio companies sitting in funds well past their natural exit date. I put a number on it in The PE Exit Backlog, and the scale of it is the mechanical reason DPI has flatlined across the board. You cannot distribute cash you have not raised by selling something.
The instinct is to treat this as a market timing problem. Wait for rates to fall, wait for the IPO window, wait for strategic buyers to come back, and the distributions will follow. Some of that is true. Some of it is out of any GP’s hands.
But waiting is not a strategy. And the part of the exit that a GP actually controls has nothing to do with rates or windows.
The one lever a GP still holds
Here is the operator point, and it is the one I care about most.
When the exit market reopens, and it will, the funds that distribute first will be the ones that can move fast. The constraint on speed is rarely the decision to sell. The constraint is the gap between deciding to sell and being able to put a credible data room in front of a buyer.
That gap is data-readiness work. And the firms that have done it during the hold will exit in a different timeframe than the firms that start it when the banker is hired.
Think about what happens after a GP decides a company is ready to go to market. The buyer’s diligence team asks for revenue by customer by product by month, reconciled to the general ledger, for thirty-six months. They ask for churn defined consistently across the period. They ask for margin by segment that survives a challenge. If the company can produce that in days, the process moves. If it takes three weeks and the numbers do not tie out, the process stalls, the buyer reprices, and the GP either accepts a discount or pulls the deal and waits another cycle.
I have watched both versions of that exact moment. The difference between them was never the quality of the business. It was whether the data was ready before the decision to sell, or scrambled together after it. That is the dynamic behind why record multiples make data your edge, not your debt. In a slow market the same logic applies in reverse. When liquidity is scarce, the company that can move cleanly through diligence is the one that converts a soft window into an actual distribution.
This is why the timing of the work matters as much as the work itself. Data readiness built in year two guides the value creation plan and is already in place when the window opens. The same work attempted in the last six months before a sale is a fire drill the buyer will see straight through. You can map your own position on that curve with the Diligence Clock, which is built to show how much runway you actually have between where you are now and a clean, buyer-ready data room.
What this means for an LP and a GP
For an LP, the vintage picture is a portfolio-construction signal. The 2017 and 2018 vintages are where realized losses against expectation are most likely to crystallize, because those funds are running out of life and out of patience. The 2019 to 2021 vintages are where the gap between mark and cash is widest, which means the marks on those funds deserve the hardest questions. When you press a GP on any vintage, the question is no longer “what is it worth.” It is “what is the path to a distribution, and how fast can the company move once you decide to sell.”
For a GP, the message is narrower and more actionable. You cannot conjure an exit market. You can compress the time between deciding to sell and being able to prove the numbers. That compression is the difference between distributing in the first wave of a recovery and distributing in the third. In a DPI drought, being first to the wire is worth more than it has been in a decade.
The cash is stuck for reasons that are mostly structural and partly self-inflicted. The structural part you wait out. The self-inflicted part, the slow, scrambled, low-trust data room that adds months to every process, you can fix now. The vintages that have already done that work are the ones that will turn the next open window into actual distributions instead of one more quarter of promising marks.
Where is the cash? It is sitting in companies that have not been sold yet. The firms that get it back first will be the ones that did the unglamorous readiness work while everyone else was waiting for the market to save them.