The exit window did not reopen. The industry built a side door instead.
When trade buyers went quiet and the IPO market stayed shut, GPs needed a way to return capital and hold their best assets longer. The continuation vehicle became that way. You move a company out of the old fund and into a new vehicle, existing LPs cash out or roll over, and fresh secondary capital comes in to fund the next leg of the hold.
It has gone from a niche maneuver to the main event. Continuation vehicles now make up roughly 89 to 90% of GP-led secondary volume (CAIA). For mature funds, continuation vehicles have risen to around 20% of the contributions-versus-distributions picture, up from about 6% historically (MSCI). This is no longer the workaround. It is the structural release valve for the exit backlog.
Here is the part most management teams miss. A continuation vehicle is a re-diligence event. The asset you have been operating for three or four years gets graded again, by a new set of buyers, on data you may not have built for outside scrutiny. The CV does not lower the bar. It raises it.
A second buyer, with a sharper question
In a normal sale, one buyer grades your data once. They run diligence, they price the risk they find, you close, and the data conversation moves inside the new owner’s walls.
A continuation vehicle splits that into two audiences who do not trust each other by default.
The first audience is the secondary buyers funding the new vehicle. They are pricing a single asset, not a blind-pool fund. They cannot hide a weak company behind a strong portfolio. So they go deep on the one company in front of them, and they go deep on the numbers behind the GP’s value creation story.
The second audience is the existing LPs being asked to roll or cash out. They are watching a sponsor sell an asset to a vehicle the sponsor also controls. The conflict is built into the structure, and every LP in the room knows it. The only thing that resolves that tension is evidence. Clean, attributable, independently checkable evidence that the price is fair and the growth is real.
Two audiences. Both skeptical. Both pricing one company on the strength of its data.
The suspicion you are diligencing against
There is a reputational headwind sitting underneath every CV process, and you should name it before a buyer does.
Roughly 30% of LPs view continuation-vehicle assets as distressed. The quiet assumption in part of the market is that the CV is where you put the company that could not sell. The asset that failed to clear in a trade process, dressed up and rolled into a new vehicle to defer the reckoning.
That suspicion sets the tone for diligence. The secondary buyer is not arriving neutral. They are arriving with a prior that says prove to me this is a winner you chose to keep, not a problem you could not offload. Every gap in your data reads as confirmation. Every clean, fast, reconciled answer chips away at the prior.
This is the same skepticism I wrote about in the context of sponsor-to-sponsor exits, where the next PE owner assumes your numbers were built to sell rather than to run the business. In a CV the dynamic is sharper, because the buyer also suspects the asset is on the table for the wrong reason. You are diligencing against a story about why the company is even in front of them.
What a CV process actually demands of your data
The CV diligence list looks similar to a sale on the surface. The depth is different, because the buyer is concentrating all their risk on one asset and they have a reason to doubt the headline.
Five things get tested harder than most teams expect.
Value creation attribution that holds up. The whole CV pitch is that the best years are ahead. The buyer wants the receipts on the years behind. Which initiatives drove EBITDA growth, how much was organic versus acquired, and what evidence closes the case on each one. A narrative does not survive this. If you cannot decompose your own growth into specific, measured levers, the buyer prices the uncertainty, and they price it against you.
A defensible forward model. Secondary buyers are underwriting the next leg, so the projection matters as much as the history. The model needs to trace back to leading indicators the company actually tracks, not to assumptions that appear for the first time in the deck. When the pipeline data, the cohort retention, and the unit economics all reconcile to the forecast, the forward number becomes credible. When they do not, the forecast looks like an argument for the price the sponsor needs.
One version of the numbers. When finance, sales, and operations each carry their own version of revenue or margin, every diligence question becomes a reconciliation exercise the buyer watches in real time. Two leaders giving two answers to the same question is the single fastest way to confirm the buyer’s suspicion that the asset is not as clean as the sponsor claims.
Data lineage back to the source. Organic versus acquired, segment-level margin, customer-level profitability. These all require the ability to trace a reported number back to the system it came from. If your growth story leans on acquisitions, the buyer wants entity-level detail that proves the integration is real and the combined view is not a spreadsheet stitched together for the process.
Speed of answer. In a concentrated, conflicted process, response time is a signal. A company that returns an accurate margin-by-product-line number the same day reads as a company that runs on its numbers. A company that needs a few days and three people to assemble it reads as a company operating blind between board decks. The buyer notices, and they adjust their prior accordingly.
If you want a structured view of how a sophisticated buyer grades data across categories like these, the Buyer Scorecard lays out the dimensions they actually score, and a CV buyer applies them with the volume turned up.
Why the early-hold work is the whole game
Here is the pattern across the CV processes that go smoothly. The companies that sail through are not the ones that scrambled to assemble a clean data picture in the months before the vehicle launched. They are the ones that did the data work in the early hold, so that by the time the CV came around, the evidence already existed.
This is the same logic that governs the broader exit backlog and the LP pressure reshaping how GPs run their portfolios. I covered that downstream pressure in LP pressure reshaping PE portfolio operations. The reporting bar that LPs now set during the hold is the same bar a CV buyer sets during diligence. If you built to meet it early, the CV is a read. If you did not, the CV is a fire drill, and the buyer can see the difference.
The reason this matters more for a CV than for a standard sale comes down to who is asking and why.
A CV exists to prove that the asset deserves more time and that the price is fair to LPs being asked to roll. Both claims rest entirely on data the company has already accumulated. You cannot manufacture three years of clean attribution in a quarter. You cannot retrofit data lineage onto acquisitions that were never integrated properly. You cannot produce a defensible forward model from a company that has never tracked the leading indicators the model depends on.
The work has to be in place before the process starts. That is why the early hold is the whole game.
The metric that ends up under the microscope
There is one number in a CV that gets more scrutiny than any other, and it connects directly to the broader shift in how the industry measures success.
Distributions. A continuation vehicle is, at its core, a way to generate a distribution for LPs who want their capital back while keeping the asset alive for those who roll. That distribution is the headline of the deal. And as I argued in the DPI reckoning, cash returned has become the metric LPs actually judge GPs on, ahead of paper marks.
So the CV puts your data directly in the path of the number LPs care about most. The valuation that sets the distribution has to be defensible. The growth story that justifies the valuation has to be attributable. The forward model that supports the price has to reconcile to indicators the company tracks. Every one of those is a data problem before it is a deal problem.
A CV that is priced on shaky data does not just risk a worse outcome on this deal. It tells the LP base that the GP cannot substantiate its own marks, which is the exact doubt the DPI reckoning has put at the center of every fundraise.
Graded twice, scored on the same thing
The continuation vehicle is the most important structural change in how PE assets change hands right now, and it is going to keep growing while the exit backlog clears. The companies passing through it are being graded twice. Once by the secondary buyers funding the new vehicle, once by the existing LPs deciding whether to roll, and both grades come down to whether the data holds up.
The suspicion is real. A meaningful share of the market assumes a CV asset is the one that could not sell. You do not beat that suspicion with a better story. You beat it with data that answers fast, reconciles cleanly, attributes growth to specific levers, and traces every number back to its source.
That data does not get built in the run-up to the vehicle. It gets built in the early hold, by teams that treated investor-grade reporting as a standing requirement rather than a diligence sprint. If a continuation vehicle is anywhere on your horizon, the question to ask now is the one the buyer will ask later. Can this company prove its growth is real, in days, with numbers nobody has to relitigate? If the answer is not yet, the time to fix it is before the second grading begins.