For a long time the industry was run on a number you could not spend.
IRR was the headline. It is elegant, it is time-weighted, and it can look spectacular years before a single dollar comes back. A fund could carry a strong IRR while every LP in it was still waiting on cash. That worked because everyone agreed to wait. The markups felt real, the exits were assumed, and the next distribution was always around the corner.
The corner stopped arriving.
LPs have spent four years watching paper performance stay high while their bank balances stayed flat. They have drawn the obvious conclusion. The number that matters is the one that shows up in the account. DPI. Distributions to paid-in capital. Cash back, divided by cash in.
This is the reckoning. IRR told you how the fund was doing. DPI tells you how you are doing. After four years of weak distributions, LPs have decided those are not the same question, and they have started ranking the second one first.
The metric LPs now lead with
The shift is not anecdotal. It shows up in how LPs say they evaluate funds.
In the Allianz survey, roughly 21% of LPs now rank DPI as their most critical metric, up from about 8% three years ago. Over the same window, IRR slipped from about 42% to about 35% as the top metric. IRR is still important. It is no longer untouchable. The fastest-growing answer to “what matters most” is the one that measures realized cash.
That movement did not happen in a vacuum. It happened because the cash stopped coming.
Bain’s data shows distributions have stayed below 15% of NAV for four straight years. Read that slowly. For four years, LPs have been getting back less than fifteen cents on the dollar of their holdings annually, in an asset class they signed up for partly on the promise of liquidity events. McKinsey frames the same drought from the other side. DPI relative to AUM hit roughly 6% for the twelve months to mid-2025, against a historical average near 16%. The distribution engine is running at a little over a third of its normal pace.
When the engine runs that slow for that long, LPs stop trusting the dashboard and start watching the fuel gauge. DPI is the fuel gauge.
Why IRR was always a paper number until you exit
Here is the thing about IRR that the good years let everyone forget. It is a forecast wearing the costume of a result.
A high IRR on an unrealized portfolio is a statement about what you believe the companies are worth and when you believe you will sell them. Both of those are estimates. The valuation is a mark. The timing is an assumption. Until the company actually changes hands, the return is a position, not a fact.
DPI has no such ambiguity. You either distributed the cash or you did not. There is no mark to argue about and no exit date to assume. It is the one number in the fund’s reporting that a buyer’s check has to validate before it can move.
For years the gap between the two numbers did not matter, because exits were frequent enough that paper marks converted to cash on a predictable cadence. The marks got tested constantly, and they mostly held. LPs could treat IRR as a reliable proxy for DPI-to-come.
Four years of thin distributions broke that proxy. When marks stop converting, the distance between “what it is worth” and “what I have received” becomes the whole story. LPs are now living inside that gap, and they have decided to measure it directly. I wrote about how that pressure transmits from LPs through GPs down to the companies themselves in DPI Pressure and the Data Gap. The short version is that it does not stop at the fund level. It lands on the portfolio.
You cannot distribute what you cannot exit
DPI is a cash number, but it is gated by an operational one. You can only return capital you have realized, and you can only realize capital by selling the company. So the entire weight of LP liquidity pressure ends up resting on a single question. Can this company actually be sold, cleanly, in a market that has options.
That is not a financial-engineering question anymore. The era when you could manufacture DPI through multiple expansion and cheap leverage is the era that produced the current backlog. There are roughly 32,000 unsold portfolio companies waiting for an exit window, a problem I covered in The PE Exit Backlog. Every one of them represents capital that an LP committed and has not seen returned. Every one of them is a drag on a fund’s DPI.
The backlog is not evenly distributed. In a slow market, buyers are scarce and selective, and they concentrate their attention on the assets that are easy to underwrite. That selectivity is reshaping how the best operating partners run their portfolios, which is the subject of How LP Pressure Is Reshaping PE Portfolio Operations. The firms paying attention have stopped treating exit-readiness as a thing you arrange in the final quarter. They treat it as a standing condition of the asset.
The logic is simple once you follow the chain. LPs want DPI. DPI requires exits. Exits, in this market, require being the cleanest asset in front of a selective buyer. So the path to DPI runs straight through exit-readiness, and exit-readiness, when you get under it, runs straight through data.
A clean exit is a data event
Walk through what a buyer actually does in diligence and it becomes obvious why.
The buyer does not take the seller’s numbers on faith. They test them. They ask for revenue by segment and by cohort. They ask for margin by product line. They ask the company to reconcile what the finance system says against what the operational systems say. They ask the management team the same question twice in different rooms and listen for whether the answers match. Every one of those tests is a data question, and every confident, fast, consistent answer is a reason to keep going. Every slow or contradictory one is a reason to discount the price or walk.
This is where most mid-market companies are quietly exposed. They can produce financial results. They struggle to produce the operational data underneath those results in a form a buyer will trust. The numbers live in three systems with three definitions. Two leaders give two answers to the same question. The growth story is a narrative because the company cannot decompose its own performance into the initiatives that drove it.
A buyer reads that as risk, and risk is priced. The deal slows, the diligence period stretches, and in a market where the buyer has other options, the deal can simply not close. The company joins the backlog. The fund’s DPI stays flat. The next raise gets harder.
You cannot return capital you cannot cleanly exit, and you cannot cleanly exit on data the buyer will not trust. That is the whole chain, stated plainly. DPI sits at the top of it and data sits at the bottom, and the LPs now reading DPI first have, whether they name it or not, started caring a great deal about the bottom of that chain.
What this changes for the hold
If DPI is the number, then exit-readiness is not a phase. It is a property of the asset that you build in early and maintain throughout, because you no longer control when the window opens. A fund under distribution pressure can decide to take a company to market on short notice. The companies that transact in that situation are the ones that were already ready.
That reframes the data work. It is not a transformation program you schedule for year four. It is a set of standing conditions you establish early and keep current. A handful of metrics the value creation plan actually depends on, each with one source and one definition that finance, sales, and operations all sign. Thirty-six months of clean, consistent monthly data. Operational drivers connected to financial outcomes, so the value creation story is evidence rather than assertion. The ability for someone other than the CFO to produce the key reports inside forty-eight hours.
None of that is exotic. It is unglamorous, and it is the difference between being the asset a selective buyer chooses and the asset that waits. If you want a sense of how little runway a typical exit timeline actually leaves for fixing data once a process is live, the Diligence Clock makes the squeeze concrete.
The number that counts now
The good years let the industry confuse a forecast with a result. IRR was the forecast. The exit was assumed, the marks were trusted, and the cash followed closely enough that nobody had to separate the two.
Four straight years of thin distributions separated them. LPs got tired of holding paper and started counting cash, and the surveys now show DPI climbing the priority list while IRR slips. That is not a fashion. It is what happens when the proxy breaks and people go back to measuring the thing itself.
For a portfolio company, the message is direct. The fund’s ability to return capital depends on your ability to be sold, and your ability to be sold depends on whether your data survives a buyer’s scrutiny. DPI is decided in the data room long before it shows up in an LP’s account.
The companies that understand that are getting ready now. The rest are counting on a window they do not control and a buyer who has better options.
For a weekly brief on PE data strategy and exit readiness, subscribe to Inside the Data Room.