Family offices are not leaving private markets. They are rerouting them.

The lazy reading of this year's family office surveys is that families are cooling on private markets after a hard stretch of distributions. The numbers describe something narrower and more structural.

What the surveys actually say

The UBS Global Family Office Report 2026 (307 family offices, average net worth $2.7 billion) puts allocations to alternatives at roughly 42% of portfolios. That is not a retreat from private markets by any definition. What is shrinking is the traditional private equity slice: 22% of portfolios in 2023, 17% in 2025, and planned flat at 17% for this cycle. Meanwhile a record 60% of offices say they plan to change their strategic asset allocation within twelve months.

Set that against the Citi Wealth 2025 Global Family Office Report: around 70% of family offices are engaged in direct private investments, and 40% of those increased their activity in the past year alone.

The money is not leaving private companies. It is changing route on the way to them.

What the fund was actually for

It is worth being honest about what the blind-pool fund solved. It was a bundle: origination, underwriting and governance, sold to investors who had capital but no deal capability. That was a fair trade for as long as it was true.

For a growing number of families it is no longer true. They have hired the deal people, and many of those deal people came out of the funds and the advisory firms themselves. We covered one such move earlier this month: a single-family office COO crossing to a multi-family platform is the same migration seen from the other side. Once a family can originate and underwrite, the fund is selling something the family already owns.

The solo detour, and where it ended

Here the data complicates the popular story, and it is worth keeping the complication. UBS's own time series shows that pure direct allocations peaked around 2021 at roughly 13% of portfolios in directs against 8% in funds, and have since fallen back, a reversal flagged in coverage of the 2026 report. Running proprietary deals solo turned out to be operationally expensive: sourcing, diligence, board seats, the years of ownership that follow.

Families learned the difference between wanting the deal and wanting the whole job.

Co-investment is where it settled

Which is why the most revealing figure in the set is the structure, not the volume. Blended industry tallies, drawing on Citi, PwC, UBS and Dentons survey data, put roughly four in five family direct positions in co-investment or club structures rather than solo cheques.

Co-investment is not a rejection of sponsors. It is a renegotiation. The family keeps the sponsor's underwriting and operating work, takes a direct position alongside the fund, and pays materially less on that slice. It is a family saying: we want your judgement, not your fee schedule on all of our capital. And every cycle of it teaches the family a little more about how the deal actually gets done. That education does not reverse.

Why it matters

The buyer across the table has changed. In the lower and middle market, the counterparty is increasingly a principal or an in-house deal lead, not a fund associate. No portfolio clock, no year-five exit pressure. Often slower, usually freer, and far more willing to structure creatively.

Relationships outweigh processes. A family that is not automatically shown every bankered auction cares intensely about who brought them the deal. Provenance is part of the underwriting.

Origination is the last thing they buy. Capability in underwriting and operations has moved in-house; the scarce import is still the introduction: the right company, the right co-investor, the right room. That is the seat we build for.

Sources: UBS Global Family Office Report 2026 · Citi Wealth Global Family Office Report 2025 · Citi Ventures: The new architecture of private capital · Modus on the UBS 2026 findings · Value Add VC survey aggregate