Private equity is becoming a family business
Family offices have made private equity their largest allocation and are going direct. The reason is a fund model whose exit machinery has jammed.
Private equity is now the single largest asset class in family office portfolios, ahead of public equity, and seven in ten of those families are putting money directly into companies rather than only into funds. That shift is not a fashion. It is a response to a fund model whose exit machinery has visibly seized up, and it points at something more interesting than an allocation change.
The numbers behind the shift
Family offices used to be a rounding error in this conversation. They are not any more. Deloitte counts roughly 8,030 of them worldwide, up from about 6,130 in 2019, and projects around 10,720 by 2030. Estimated assets under management sit near 3.1 trillion dollars and are forecast to reach 5.4 trillion over the same period.
The geography is broader than the stereotype suggests. North America holds about 3,180 of them, roughly 40 percent. Asia-Pacific has around 2,290, about 29 percent, and Europe about 2,020, a quarter. This is not a Swiss and American phenomenon with a few outposts. It is a global reallocation of who owns operating companies.
Deloitte's survey work puts private equity at roughly 30 percent of family office investments, up from about 22 percent in 2021, which is enough to have overtaken public equity as the largest single allocation. Citi surveyed 346 single family offices across 45 countries and found 70 percent engaged in direct investments, with four in ten of those having increased that activity over the previous year.
Why they are stepping around the funds
The obvious explanation is fees, and fees are part of it, but they are not the interesting part. Families have been paying two and twenty for decades without deciding to build their own deal teams. Something changed on the other side of the trade.
What changed is that the fund model stopped returning money. Bain's 2026 global private equity report puts roughly 32,000 portfolio companies, worth about 3.8 trillion dollars, sitting unsold on sponsors' books. Distributions as a share of net asset value have stayed below 15 percent for four consecutive years. Average holding periods at exit have stretched to around seven years, against five to six for much of the previous decade, and close to 40 percent of all portfolio companies have now been held longer than five years, up from 29 percent in 2019.
What that looks like from a family's side of the table
You committed capital to a ten year vehicle. Year seven arrives and the companies have not been sold. Your paper returns look fine, because they are marks. Your actual cash back has been under 15 percent of net asset value for four years running, and the manager's proposed solution is a continuation vehicle, which is a request to hold the same asset for longer in a new structure with new fees.
Put yourself in that position and the direct route stops looking adventurous and starts looking obvious. If the capital is going to be locked up in operating companies for seven years or more anyway, the question becomes why it should be locked up in someone else's operating companies, on their timetable, with a layer of fees for the privilege.
The illiquidity was always the deal. What broke was the promise that it ended on schedule.
The measurement problem underneath it
There is a deeper reason the two structures are diverging, and it is not about the current market. It is arithmetic, and it is permanent.
Bain's analysis of buyout vintages from 2000 to 2015 found that internal rate of return stagnates around year seven and declines after that. That is not a statement about business quality. A well-run company at year eight is usually better than it was at year four. It is a statement about the metric: internal rate of return is time-weighted, so an identical gain earned over a longer period scores worse, and the same excellent business becomes a worse investment every year you keep it.
That sounds like a technicality until you put actual numbers on it, at which point it stops being a nuance and becomes the whole problem.
| What the owner did | Money returned | Years held | Internal rate of return |
|---|---|---|---|
| Sold quickly | 2x | 3 | 26.0% |
| Held the same company | 2x | 7 | 10.4% |
| Held and kept compounding | 3x | 10 | 11.6% |
| Held for a generation | 5x | 15 | 11.3% |
Read the first row against the last. One owner doubled their money. The other made five times their money. On the metric the industry is judged by, the first one scored more than twice as well. Nothing in that table is a trick and none of it depends on the market: it is what time-weighting does, and it does it every year, to every fund, forever.
A fund is judged on that number. Which means a fund manager holding a genuinely improving business faces a structural incentive to sell it, not because selling is right for the company, but because the scoreboard punishes patience. That is the clock, and no amount of alignment language removes it, because it is built into how the industry keeps score.
It also explains why the continuation vehicle arrived when it did. It is the industry engineering a way to keep a good asset without the original fund's clock scoring it, which is a sensible response to the problem and also a quiet admission that the metric, not the business, was the thing that needed managing.
- Time-weighted return
- A measure that penalises how long you took as well as how much you made. Under it, doubling your money in three years beats doubling it in seven, even though the second outcome may involve a much better company.
- Continuation vehicle
- A new fund created to buy an asset from the manager's own older fund, so the position can be held longer. A reasonable tool, and also an admission that the original timetable did not fit the business.
A family holding a company directly is not scored that way. If the business compounds for twenty years, that is the whole point rather than a drag on a ratio. The advantage is not that families are smarter or more patient by temperament. It is that nothing in their structure forces a sale at year seven, and structure beats temperament every time.
What families are actually good at, and what they are not
It would be easy to write this as families good, funds bad. That is not what the evidence supports, and pretending otherwise would leave out the part that decides whether any of it works.
Genuinely better: the holding period
No fund life, no vintage year, no pressure to show a realisation before the next raise. A family can buy something and simply keep it, which is the correct answer for a large share of good businesses and an option a fund does not have.
Genuinely better: what the seller sees
An owner selling a company they built often cares who gets it. A family that intends to hold, keep the name and keep the people is a different proposition from a sponsor with a five year plan, and that difference wins deals that price alone would not.
Usually worse: deal flow and screening
A fund sees hundreds of opportunities a year through an established network. A family office frequently sees what its advisors and friends bring it, which is a much smaller and much more biased sample.
Usually worse: what happens after completion
This is the real gap. A sponsor has an operating partner bench, a playbook, and people who have done the first hundred days before. A family that has bought its first company has a board seat and good intentions.
Frequently worse: saying no
Committing to a fund and being talked out of one deal are different emotional events. Direct investing concentrates decisions in a small group who see fewer comparisons, and it is harder to reject something when it is the only thing on the table.
Screening and post-completion capability are what decide outcomes. Buying well is a screening problem and owning well is an operating problem, and permanent capital only helps with the second if somebody can actually run the business. Patient money attached to no operating capability is just a slower way to find out you overpaid.
What this means if you are not a billionaire
The structural insight here generalises well below the family office threshold, which is what makes it worth writing about rather than admiring.
The advantage is not the size of the balance sheet. It is the absence of a forced exit. Anybody who buys a company with their own money, or with money that has no fund life attached, has the same edge over a sponsor bidding on the same asset: they can pay for what the business will produce over decades rather than what it can be resold for in five years, and they can accept an improvement that takes three years to show up.
It also changes what you should be willing to buy. On a five year clock, a business needs a story that a future buyer will pay up for. On no clock, it needs to produce cash reliably for a very long time, which is a different and more boring test, and one that a great many unglamorous companies pass easily. The full argument for that is in permanent capital versus the fund clock, and why so many consolidation attempts fail regardless of structure is in why roll-ups fail.
Where I think this ends up
The direction of travel seems clear enough. More capital owned by people with no obligation to sell, chasing companies that reward being kept, in an environment where the traditional exit route is congested with 32,000 companies waiting their turn. That is a structural shift rather than a cyclical one, and it favours the holder over the trader.
The thing to watch is whether the operating capability follows the capital. Owning a company forever only compounds if it gets better while you hold it, and getting better is work done by people, not by a structure. The families that build a genuine operating bench will do extremely well out of this. The ones that simply lengthen their holding period will discover that patience applied to a mediocre business produces a long, quiet, expensive nothing. That is the same test good company, bad ops sets, and it does not get easier with more zeroes attached.
The short version
- Private equity is now roughly 30 percent of family office portfolios, up from 22 percent in 2021, and has overtaken public equity as the largest allocation.
- Seventy percent of single family offices surveyed by Citi across 45 countries invest directly, not only through funds.
- The push factor is a jammed exit market: around 32,000 unsold portfolio companies worth 3.8 trillion dollars, and distributions under 15 percent of net asset value for four straight years.
- Underneath that sits a permanent arithmetic problem. Time-weighted return decays past year seven, so the scoreboard punishes a manager for keeping a business that is still improving.
- The scale of it: 2x in three years scores 26.0 percent, while 5x over fifteen years scores 11.3. Five times the money, less than half the score, and no market conditions required.
- The edge is the absence of a forced exit, not the size of the balance sheet, and it is only real if you genuinely never sell.
Questions I get on this
How much of a family office portfolio goes into private equity?
Why are private equity funds holding companies longer?
Is a family office better than a private equity fund at owning a business?
Family office counts, assets and allocations from Deloitte Private's Family Office Insights Series; direct investment share from Citi's 2025 survey of 346 single family offices across 45 countries. Holding periods, unsold inventory, distribution rates and the vintage return analysis from Bain & Company's Global Private Equity Report 2026.