Blog/Private Equity
The 2026 evidence on family-office direct investing shows a market that is large, active and still difficult to rank with precision. Citi’s 2026 Global Family Office Report found that 75% of the 351 families it surveyed across more than 40 countries invest directly in companies. Dakota Marketplace data cited by Angel Investors Network counted 73 family-office direct investments globally in June 2026, worth $19.97 billion in disclosed value, followed by 61 deals in July. Named July participants include ICONIQ Capital and Doerr Capital, each logging three deals. No credible global league table of direct-investment volume exists, because private offices do not disclose complete deal data. The evidence does support a structural shift in how private capital reaches operating companies, and a persistent gap between families that can underwrite a deal and families that can only follow one.
Direct investing means putting capital straight into an operating company, private business, startup, real asset or private-market transaction rather than committing to a third-party blind-pool fund. It covers minority stakes, growth equity, venture rounds, buyouts, secondaries and co-investments.
“Top” carries at least four meanings in this market: largest assets under management, highest disclosed deal count, largest share of portfolio allocated to direct deals, and most accessible to external capital raisers. Directory providers frequently conflate them. Public rankings also mix single-family offices, multi-family offices, family-controlled holding companies and hybrid wealth platforms, while some list public-equity holdings alongside private direct stakes as if they were the same exposure.
For this article, relevance means evidence of scale, disclosed direct activity or practical importance to dealmakers within the 2026 sources below. That narrower approach is less satisfying than a clean ranking, but it is more faithful to the opacity of the market.
| Family office or platform | Evidence in 2026 sources | Type of relevance | Caveat |
|---|---|---|---|
| ICONIQ Capital | Three direct deals in July 2026, based on Dakota data cited by Angel Investors Network | Disclosed deal activity | Single-month snapshot, not annual volume |
| Doerr Capital | Three July 2026 deals, with participation in Antora Energy’s $550m Series C and CuspAI’s $450m Series B | Named large-round participation in energy and AI | Participation only, not lead role or check size |
| Bezos Expeditions | Oratomic $300m Series A in quantum computing | Named deep-tech transaction | One disclosed example |
| Walton Enterprises | Estimated $225bn AUM according to Altss, with stakes listed in Apollo Global Management, Snowflake and Pinduoduo | Scale | Listed positions are public equity, not private direct deals |
| Cascade Asset Management | Large holdings and direct exposure listed by Unbiased | Scale | Public holdings dominate the disclosed picture |
| Pritzker Group | Estimated $10bn under active management with a direct track record according to Altss | Platform capability | No named 2026 transaction in the source |
| Stetson Family Office and Elysium Management | $11bn and $9bn AUM respectively, described as direct investors in Dakota’s New York list | Regional access point | New York list, not a global ranking |
Altss, AllFamilyOffices, OpenVC and FamilyOfficeHub are commercial data platforms. Their figures are estimates rather than audited disclosures and should be treated as such. They are useful for market mapping, but they cannot turn partial disclosure into a verified global league table.
Citi’s 2026 figure of 75% compares with 70% in its 2025 survey, although the samples differ and the two numbers are not a clean like-for-like series. The more useful detail is ticket size. Citi found that the most common preferred direct-investment cheque is $1m to $5m, chosen by 44% of respondents. Nineteen percent prefer below $1m, 18% prefer $5m to $10m, around 11% target $10m to $25m and fewer than 9% write above $25m.
That distribution matters for anyone building a buyer list. A family office with $2bn of net worth may still prefer a $3m cheque, which makes it a syndicate participant rather than an anchor investor.
Allocation data varies by dataset. Addepar’s Q2 2026 analysis, covering more than 650 family offices with nearly $1.4 trillion, put alternatives at 46% of average portfolios as of 30 June 2026. UBS 2026, as cited by Aleta, puts alternatives at 42%, with private equity at 18%, split 8% direct and 10% through funds. These are different populations and classification rules, rather than contradictory findings.
FINTRX reported that 92.7% of family offices newly added to its database in Q2 2026 list direct investments as a primary interest, compared with 10.4% for hedge funds and 6.3% for private credit. Newly added offices skew towards those actively marketing for deal flow, so the figure is best read as a signal about stated appetite among new entrants rather than the behaviour of established multi-billion-dollar offices.
The 61 July 2026 deals broke down with information technology at 16, industrials at 12 and healthcare at 10. FamilyOfficeHub identifies artificial intelligence, longevity medicine, defence technology and energy transition as four themes attracting deliberate large-scale US family-office capital in 2026.
Doerr Capital illustrates the pattern. Antora Energy’s $550m Series C sits in industrial decarbonisation, whereas CuspAI’s $450m Series B sits in AI-driven materials discovery. Both are large, syndicated, late-stage private rounds rather than the small proprietary deals families are sometimes assumed to prefer.
Citi found that stage preference varies by AUM and region, with Asia Pacific respondents notably focused on growth and pre-IPO opportunities at 79% and 68% respectively. The same regional tilt reinforces the need to separate direct-investing appetite from a generic family-office label.
Families want direct economics. Most cannot build the sourcing and diligence apparatus required to earn them alone. In practice, the resolution is the co-investment or club deal, where the family invests alongside a lead sponsor or syndicate and relies on that lead for origination, diligence depth and post-close governance.
The economics can still improve. A co-investment sleeve alongside a fund commitment reduces blended fee load and gives the family more control over pacing and concentration. It does not, however, create proprietary access by itself.
Adverse selection remains a live risk. A GP that shows its best deals to strategic partners and its hardest ones to the broader co-investment list creates a portfolio the family did not choose. Two questions separate credible co-investors from passive cheque writers: why is this deal being syndicated, and what would the family have done differently had it seen the asset first?
The fee argument also needs discipline. Avoiding 2-and-20 does not eliminate cost. Legal, tax structuring, special purpose vehicle formation, commercial diligence, quality of earnings work, monitoring and annual valuation reporting all sit on the family’s own budget.
A family office reviewing a direct investment needs a process that is closer to institutional private markets than public-market portfolio selection. The following workstreams are the minimum, especially where the family is taking concentrated single-name risk.
Assume a family commits $5m to a growth round, sized within Citi’s most common bracket, with a five-year hold and a 2.5x gross return.
Through a fund paying 2% management and 20% carry, the blended drag on a $5m commitment over five years runs roughly $500,000 in management fees plus carried interest on the $7.5m gain, or about $1.5m at 20%. Gross proceeds of $12.5m become roughly $10.5m net.
Done directly, the family keeps the full $7.5m gain but absorbs its own costs: commercial and financial diligence, legal negotiation of the shareholder agreement, SPV formation, ongoing valuation and reporting. Call that $150,000 to $300,000 on a deal of this size. Net proceeds land near $12.2m.
The saving is real. It is also entirely contingent on the family picking a deal at least as good as the fund would have. A single write-off inside a concentrated direct portfolio erases several years of fee savings, which supports reserve discipline and position limits rather than a higher volume of direct deals.
GPs raising co-investment capital should size family participation against the Citi cheque distribution rather than against headline net worth. A family that writes $3m needs a syndicate structure, not a bilateral negotiation.
Bankers running sell-side processes can treat family offices as credible bidders where the family has an operating history in the sector, but they should test capacity to fund on a compressed timetable and willingness to accept process deadlines. Relationship-led coverage matters more here than a broad distribution list.
Private credit desks should avoid reading the 6.3% FINTRX figure as evidence that families have abandoned private debt. That number covers new database entrants only and says nothing about the allocation behaviour of large established offices.
Corporate finance teams gain flexibility on hold period and structure in many family-office transactions, but they should expect requests for information rights and board observation that a passive fund LP would never make. Those rights can be valuable alignment tools, although they need to be documented with the same care as other shareholder agreement protections.
Direct investing is now a standing feature of family-office portfolios rather than an opportunistic sideline, and the 2026 data supports that on adoption, deal count and allocation. It does not support the claim that families are replacing funds. The UBS-cited 8% direct versus 10% fund split inside private equity describes coexistence.
The families that will compound through this cycle are the ones that treat access as the starting point rather than the edge. A family office that cannot answer why it was shown a deal, and what it would have priced differently, is buying syndicate allocation at full risk with none of the diversification a fund commitment provides.
P.S. If you are raising from or investing alongside family offices, check out our Premium Resources for investor and fund databases, transaction decks and more tools to help you advance your career.
Join 80,000+ readers
One email a week: markets and finance news, the latest M&A deals and reports, investor biographies and deep dives.
Free forever. Unsubscribe anytime.