There is a conversation I keep having, in slightly different words, with GPs across Europe and the United States this quarter. It tends to start with a portfolio review and end somewhere else entirely.
The review itself is familiar. The company is fine. The numbers are roughly where they were in the last board pack. The founder is focused. The runway is adequate. What is new, and what quietly unsettles the conversation, is the creeping awareness that the company most of us underwrote two, three, four years ago is not quite the company we are now holding — and that the distance between those two was travelled during the months between our scheduled check-ins.
A little over a year ago I wrote a piece on these pages called Portfolio Harvest Season. The argument was that the first quarter of each year is the moment to triage — to decide, with discipline, which names to back, which to hold, and which to run a process on. I stand by the diagnosis. I am less sure the cadence that underwrote it still holds.
Two forces have tightened in parallel, each of them alone familiar, together rather less so.
The first is the liquidity math, which has not loosened. Bain's 2026 Global Private Equity Report puts distributions at 14% of NAV through Q3 2025 — a level not seen since 2008–09, and below historical averages for four consecutive years. Cambridge Associates estimates that continuation vehicles will represent at least 20% of 2026 distributions as LPs increasingly elect to sell rather than roll. What began on a T-shirt — DPI is the new IRR — is now how re-up committees actually behave. The re-up decision is being made on realised cash, not on marks.
The second is the pace at which the underlying businesses are being tested. Crunchbase reports that the first quarter of 2026 saw close to $300 billion in global venture deployment — nearly 70% of all of 2025, in a single quarter — with roughly 80% flowing into AI, and four frontier labs absorbing about 65% of global VC dollars between them. Much of the capital is routing around the incumbents and into the models that threaten them. Anthropic's release of Claude Mythos Preview this month is one visible marker of the rate. It is not the point.
The point is that secondary buyers are already pricing the convergence. Bloomberg reported in late February that discounts on tech-heavy portfolios had widened to as much as 20%, from roughly 5% only weeks earlier, explicitly citing AI disruption risk in the underwriting. That is the quiet signal in the tape. It is not the market telling us valuations are soft. It is the market telling us that the competitive landscape those valuations were built on is no longer the landscape.
Quarterly cadence was calibrated for a world in which a portfolio company's thesis could be trusted to hold between check-ins. That world is no longer the reliable backdrop.
What follows is not a case for heavier-handed interference. Founders are already aware, often painfully, of the ground shifting under them. What has shifted is the burden on the investor side — to be present more often, with sharper questions, in a way that helps the company adapt rather than simply reports back to the firm.
In practice, that tends to look like moving the core founder conversation from quarterly to monthly, with the cadence structured rather than ceremonial. It looks like running an honest AI pre-mortem on each meaningful position at least twice a year — asking what would need to be true for the company's wedge to become a feature inside a frontier model, and treating the answer with the seriousness it deserves. And it often looks like acknowledging, quietly, that the investment team at its current headcount cannot carry this weight alone.
The gap is real, and it is also addressable. An operating partner with genuine category depth, a fractional or interim CEO, an independent advisor close to the sector — each of these extends the firm's surface area into the company without crowding the founder or stretching the partnership. The title is not the point. The point is that someone with real domain depth is inside the weekly rhythm of the company, and not the quarterly one.
Harvest season has not ended. It has simply stopped being a season. The discipline that decides which names to back, which to hold, and which to exit is the same discipline it has always been; the window in which it remains effective has narrowed. The firms that will reraise through the next two vintages will not be the ones that worked hardest at year-end. They will be the ones whose portfolios were legible to an outside buyer at any point in the year, because the work never quite paused.
The weeks ahead are a mix of home and away, with the on-the-road time mostly built around conversations on M&A and growth capital. If our paths cross, even briefly, it would be good to connect.
April 27 — Home. Liliana turns seven. All other matters respectfully deferred.
May 11–12 — London/Dublin
May 17 — Stablecon, Amsterdam
June 1 — Money20/20, Amsterdam. If you're going, ping me.
July 1–2 — iGB L!VE, London
If you are quietly pressure-testing a portfolio this quarter — on readiness, secondary optionality, or continuation structure — I'm always open to a thoughtful conversation.
Securities Offered through Wellesley Hills Securities. Member FINRA/SIPC
