Sixty-four per cent of family office decision-makers said they planned six or more direct investments in the year ahead, according to the BNY Wealth 2025 Global Family Office Study, which covered 282 offices. Six in twelve months is a scheduling commitment before it is an allocation decision. Every one of them requires a company found, a management team met, a set of terms argued over, and a judgment defended to whoever else in the family has a claim on the money.
None of that arrives in the post. Somebody has to know somebody. Then somebody has to be willing to say so out loud, in writing, with their own name attached.
Tommy Shields does that for a living. He is Head of Investor Relations at Onyx Reserve, a private investment firm operating in South Florida, and the willingness gets tested in front of him most weeks.
“An introduction is a loan taken out against somebody else’s reputation,” Shields said. “The person making it has put their own standing behind a stranger, and they do not get that standing back if the stranger turns out badly. A transaction closes and the file goes in a drawer. An introduction sits on the books of the person who made it for as long as both parties are still in the business.”
The direct-investment surge runs on private knowledge
The BNY finding is not an outlier. Citi Wealth’s 2025 Global Family Office Report, based on 346 respondents across 45 countries surveyed in June and July 2025, found 70% of family offices making direct investments, with four in ten of those having increased that activity.
Read the two studies together and they describe the same behaviour from different distances. A large majority now participates. A substantial minority participates at volume.
Direct investing has a precise operational meaning that gets lost when it is discussed as an allocation preference. A commitment to a fund is an act of delegation. The office signs documents, wires capital, and pays a manager to do the finding, the meeting and the arguing. Direct means the office does the finding itself, which means the office has to be findable in return.
The allocation data shows how much weight now rests on that. The UBS Global Family Office Report 2025, drawn from 317 single family offices with an average family net worth of $2.7bn, found North American family offices holding 54% of portfolios in alternatives against 46% in traditional assets, with private equity at 27% and real estate at 18%. Roughly half of the average balance sheet in this segment now sits in assets that have no ticker, no daily price and no way in except through a person.
Deloitte Private’s Family Office Insights Series counted 8,030 single family offices worldwide in 2024, up 31% from 6,130 in 2019, and projected 10,720 by 2030, with global family office assets under management at $3.1trn in 2024 heading for $5.4trn by 2030. More offices, larger, all wanting to do the finding themselves.
The opening narrowed while the traffic grew
Here the optimistic reading runs into a problem.
Goldman Sachs, surveying 245 family offices in May and June 2025, found alternatives at 42% of portfolios, down from 44% in 2023, with private equity falling from 26% to 21%. The RBC and Campden Wealth North America Family Office Report 2025, covering 141 respondents with $285bn in collective wealth, found North American offices holding 29% of the average portfolio in private markets in 2025, down from 30% the year before, though 88% remained invested in private markets in some form.
So the direct-investment count is rising inside a private-markets allocation that has stopped growing and in places has contracted. More activity, more offices, more people in the room, and a slightly smaller pot to fight over.
Goldman also found 39% of its respondents planning to increase private equity over the following twelve months, which points the other way. The two numbers can both be true, and the reconciliation is uncomfortable for anyone hoping the door has swung open: intentions rose while positions fell, and intentions are cheap.
The practical consequence is that access got harder rather than easier. When capital was expanding faster than the supply of credible counterparties, a competent stranger with a decent proposition could get a hearing. When the pot flattens, attention is rationed first, and the rationing is done by whoever already has the calendar.
“Everybody talks about access as though it were a door that is either open or shut,” Shields said. “It is closer to a queue with no visible order to it. Somebody who vouches for a person has not opened any door. What they have done is move that person forward in a line that nobody running it will ever admit exists.”
What the obligation actually costs
The distinction Shields draws is about liability rather than warmth.
A transaction is bounded. Both sides know what is being exchanged, both sides have counsel, and the arrangement is designed so that when it concludes, it concludes. Whatever happens afterwards belongs to the parties and to the documents.
An introduction has no such boundary. Nothing is exchanged, nothing is signed, and no lawyer reads it. What passes is an assessment: this person is worth your afternoon. If the assessment proves wrong, the person who made it absorbs the cost, and they absorb it quietly, without recourse, in the form of the next call they make being taken slightly less seriously.
“The asymmetry is what people miss,” Shields said. “The person receiving an introduction risks an hour. The person making it risks every future introduction they were ever going to make. That is why the good ones are careful to the point of seeming rude about it, and why anybody who hands them out freely stops being worth anything to anybody within about two years.”
There is a version of this argument that collapses into sentiment, and the data does not support the sentimental version. Nobody is choosing counterparties on the strength of a handshake when 8,030 family offices worldwide have professional staff, external counsel and diligence checklists.
What the numbers do support is narrower and harder. When direct investing becomes the dominant behaviour of a segment, the first filter stops being the deal and becomes the person carrying it. UBS found North American offices with succession plans placing the next generation on the board in 59% of cases, on strategic asset allocation in 41% and on investment management in 39%, which means the filter is being handed to people who did not build the original network and will have to build their own.
“There is a reason the obligation gets taken seriously by people who have been here a long time,” Shields said. “Everyone eventually needs somebody to vouch for them, including the person doing the vouching. The system only works because it is expensive to abuse and everyone can see who is abusing it.”
Whether it stays expensive to abuse is the part the data cannot answer. The surveys count intentions and allocations. They do not count how many of the six direct investments a family office plans this year will come from somebody the principal has known for a decade, and how many from somebody who was in the right room in April.