The Club Deal Decade: Why Family Offices Co-Invest — and When the Model Breaks

Over the past decade, the world's wealthiest families have carried out one of the larger reallocations in private capital, and they have done it almost without commentary. PwC's Global Family Office Deals Study — which tracks announced transactions rather than survey intentions — found that direct investments into companies, whether start-ups, M&A or private equity stakes taken without a fund in between, accounted for around 70% of family office deals in 2024. In 2015 the picture was inverted: real estate and fund commitments made up 56% of activity.

And families are not doing it alone. The same study shows that club deals — transactions made alongside other investors — rose from 58% of family office investments in late 2015 to a peak of 75% in early 2023, and still stood at 69% in the first half of 2025. In the American market, the 2024 edition of the study put the club share above 70%. The lone family writing a solo cheque into a private company is now the exception, not the archetype.

So the interesting question is no longer whether family offices do direct deals. It is why the club format has won so decisively — and under what conditions the features that make clubs attractive quietly turn against the families inside them.

Why families left the fund queue

Start with the arithmetic, because it is genuinely compelling. A traditional private equity fund charges a management fee on committed capital and takes roughly a fifth of profits above a hurdle — the familiar two-and-twenty. A co-investment offered at reduced or zero fee removes most of that drag at a stroke. On a position held for six or seven years, the difference between gross and net compounds into something a principal can see without a spreadsheet. The structure also spares the investor the blind-pool problem: in a fund you underwrite a manager and hope; in a direct deal you underwrite the asset itself.

The shift shows up in allocations. UBS's Global Family Office Report 2025 — a survey of 317 offices with an average family net worth of $2.7 billion — found private equity running at 21% of the average portfolio, with direct investments making up more than 40% of that private equity allocation. Direct deals are no longer a hobby line; for many offices they are the core of the illiquid book.

But fees alone do not explain the migration. Control does — a family office answers to no limited partners, faces no fund life, and can hold a good business for twenty years or exit in two. And identity does. The principals building these offices are disproportionately founders who sold operating businesses; they believe, often correctly, that their edge is operating rather than allocating.

The most encouraging statistic in the genre suggests families know where that edge ends. Goldman Sachs's 2025 Family Office Investment Insights — 245 offices surveyed, two thirds of them with a net worth above $1 billion — found that 44% of respondents invest primarily directly in private real estate, where families often have decades of operating history, while continuing to rely mainly on managers across other alternatives. Direct where the family has provable expertise; intermediated where it does not. The same survey found technology accounting for 25% of family office direct investment activity — which is where that self-knowledge gets tested, since far fewer families built their fortunes writing software than buying buildings.

What the evidence actually says

The academic record is more divided than either the banks or the sceptics admit — and the division is instructive. The foundational caution comes from Lily Fang, Victoria Ivashina and Josh Lerner, whose study of two decades of direct investing by seven large institutions, published in the Journal of Financial Economics in 2015, found that co-investments underperformed the very funds alongside which they were made. The mechanism they identified was adverse selection: managers tend to share deals with outside investors when they need the capital — deals too large for the fund, done at frothy moments, in hot sectors. Solo transactions in the same dataset outperformed, but only where the investor had the capability to resolve information problems independently.

The rebuttal came five years later. Reiner Braun, Tim Jenkinson and Christoph Schemmerl, examining a large sample of buyout and venture co-investments in the Journal of Financial Economics in 2020, found no evidence of adverse selection on average: gross return distributions of co-investments and the remaining fund deals were similar. Net of fees, their simulations showed that reasonably sized portfolios of co-investments significantly outperformed fund returns — the fee saving is real, and on average it is not given back through worse deals.

Both findings can be true, and for a family office the reconciliation matters more than the headline. The average co-investment may be fine; the particular co-investment being shown to you may not be, depending on why you are being shown it. And the phrase doing quiet work in the optimistic paper is "reasonably sized portfolios". The result assumes diversification across many deals. A family office holding four direct positions is not running the strategy the academics tested. It is making four bets.

Why the club format won

The club is, at its best, a machine for fixing exactly those two weaknesses. Sharing a deal across several families lets each take a sensible ticket, which is how a handful of positions becomes a genuine portfolio rather than a concentration risk. And the right co-investors are a diligence multiplier. The most underpriced asset in any club is the peer who spent thirty years operating in the sector under review — the family that built a hospital chain reading a healthcare services deal will ask questions no banker's associate would think to ask, and no data room can deflect.

There is also a structural difference between the two kinds of co-investing that gets blurred in the statistics. In a sponsor-led co-investment, the family is shown the deal because the sponsor needs the equity; the invitation is downstream of the manager's fundraising position, which is precisely where Fang, Ivashina and Lerner located the adverse selection. In a peer club, the lead family typically has its own capital committed first and its own name attached to the outcome. The alignment is different in kind, not merely in degree. It is no accident that surveyed families consistently express more comfort investing behind another principal's conviction than behind a manager's allocation decision — the lead family eats its own cooking by construction.

None of which makes clubs safe. It makes them conditional.

Where clubs break

Adverse selection wears a friendly face. The version that damages family offices is rarely the cold-eyed GP rationing a hot deal. It is the deal a friend brings. Social capital substitutes for underwriting: declining feels like an insult, questioning the numbers feels like questioning the friendship, and the diligence meeting becomes a dinner. The hardest transaction for any principal to refuse is one carried into the room by someone he trusts — which is exactly why those deals deserve more process, not less.

The diligence bench does not exist. Most family offices employ fewer than ten people, and the same key-person fragility that undermines the office's own bench undermines its underwriting. A two-person investment team cannot replicate the work of a mid-market buyout firm with forty professionals, an operating-partner roster and a litigation budget — yet a direct deal demands precisely that work, plus board service, plus the eventual exit. A fund commitment is an allocation. A direct deal is a job.

Concentration compounds quietly. Direct positions are lumpy, illiquid and unrebalanceable; they do not drift back to target, they wait. The cycle has already administered one lesson: UBS's 2025 report found that offices planning allocation changes intended to cut private equity from 21% to 18%, with the reductions driven mainly by direct investments, as slow exit markets and expensive financing trapped capital in positions families had expected to recycle. The fee saved on entry is small consolation for the exit that never comes.

Nobody owns the failure. When a fund's deal sours, there is a documented process and a paid professional whose job is the workout. When a club deal sours, there is often a WhatsApp group. Who funds the bridge round, and at what price to those who decline? Who takes the board seat through the restructuring? Were drag-along rights, tag-along rights and information rights actually papered, or did the families rely on goodwill? Clubs are built on relationships, and relationships are precisely what a down round taxes.

The discipline that separates outcomes

The offices that do this well treat direct investing as a programme rather than an appetite, and the elements are unglamorous. A written mandate stating which sectors are in scope, what the family's claimed edge actually is, and the maximum size of any single position and of the direct book in aggregate. A pre-committed external diligence budget — legal, financial, commercial — spent on every deal, with the relationship deals explicitly held to the same standard. Lead economics, governance rights and workout responsibility agreed in writing before money moves, while everyone still likes each other. And a post-mortem habit: marking the book honestly each year, including the positions a friend brought.

None of this is sophisticated. All of it is rare, for the same reason office succession plans are rare — small teams, founder gravity, and the comfortable assumption that judgement substitutes for process. The evidence suggests the opposite: judgement is what the process protects.

The peer question

The dilemma is sharpest where the model is newest. PwC counts roughly 300 family offices in India in 2024, up from 45 in 2018, managing an estimated $30 billion — most of them young, founder-led, and presented with more club deal flow than any of them can independently underwrite. Every one of those offices faces questions that no adviser with a transaction fee attached will answer straight. What does a fair co-investment term sheet between families look like? How do peers decline a friend's deal without ending the friendship? What share of the direct positions made five years ago would members honestly mark below cost today?

These are conversations SoHo exists to host — principal-only, anonymised where it matters, and solicitation-free by design. SoHo does not broker transactions and earns nothing from any deal members may or may not pursue together, which is precisely what makes honest answers possible. The room where nobody is selling the deal is the only room where the deal can be properly judged.

The clubs that endure are not the ones with the best deal flow. They are the ones in which someone can still say no — and remain a member.