Family Offices Trade Startup Bets for Credit and Property

North American family offices keep 29% of the average portfolio in private markets, with new money aimed at private credit, buyouts and property, not early-stage startups. The Campden Wealth and RBC Wealth Management survey of 141 offices, taken between April and August 2025, captured expected 2025 returns of 5% and a 48% push to improve liquidity.

That mix recasts some of the region’s richest families as lenders and landlords. A year on, the companies paying those coupons are the ones starting to mark down.

Private Markets at 29% of the Average Book

Campden Wealth, the London family-office research firm, and RBC Wealth Management polled 317 single- and multi-family offices worldwide for the 12th global study. Of those, 141 sat in North America, most of them in the United States. On average those families held $2.0 billion of wealth, including operating businesses, and managed $1.5 billion. Collective wealth in the North American sample was $285 billion.

Private markets (direct private equity, private equity funds, venture capital and private credit) were the largest sleeve at 29%, down from 30% a year earlier. Some 88% of the offices had some private-market exposure. Public equities were next at 27%. Real estate, counted separately, was 18%.

The survey’s own table is blunter than the “ditch startups” headline. Venture capital is still 5% of the book. What slipped is the queue for new money. Net demand, the share planning to add a sleeve minus the share planning to cut it, favored direct private equity, private credit, and real estate.

WHERE NEW MONEY WAS HEADED IN 2025

Direct private equity was the most popular destination for new investment, named by 48% of offices. Private credit’s 4% weight is small next to property, but the +15 net score put it second only to direct buyouts among the private sleeves. The report said the draw was “the high nominal interest rates that sub-investment grade borrowers are prepared to offer non-bank lenders.”

Year-to-date results, scored as the share of offices that beat their own forecast minus the share that missed it, explain the ranking. Gold printed +61. Cash printed +20. Private credit printed +14. Direct private equity printed -16, private equity funds -24, and venture capital -33, the worst print in the table.

Borrowers Paying Up Are the Ones Starting to Crack

The coupon that made private credit look like a haven is the same coupon a stretched borrower has to find every quarter. Houlihan Lokey, the investment bank that marks thousands of these loans, said on Sept. 10, 2026, that stress has piled up at the small end of the market, the slice that is most of direct lending.

PRIVATE CREDIT MARKS IN Q2 2026

• Smallest loans: Among borrowers with less than $20 million of EBITDA, loans priced below 90 cents on the dollar rose from about 1% in 2023 to 12%.
• Core middle market: For borrowers with $20 million to $100 million of EBITDA, 6% of loans sit below 90, the highest reading in three years.
• Upper middle market: For borrowers above $100 million of EBITDA, 3% of loans sit below 90.
• Whole book: 7% of all loans now price below 90, more than double the historical average, while defaults on the full market were 0.8% of loan principal and 2.5% of borrowers by count.

For borrowers under $100 million of EBITDA, Houlihan’s loan-level private credit valuations put the Q2 2026 default rate at 3.0% by size and 3.6% by count. More than two-thirds of borrowers still grew both revenue and EBITDA. Healthcare was the only sector elevated on both default measures. The split is the story: the largest credits are holding, and the small ones that pay up for non-bank money are not.

Family-office trackers in late August 2026 already showed some offices trimming real estate and private-market sleeves after private-credit funds marked assets down, and adding listed stocks. That is the second move after the 2025 rotation, not a reversal of the whole bet.

Most 2025 Venture Dollars Went Into AI Megadeals

Early-stage venture did lose the family-office ranking. It did not vanish as an industry. The National Venture Capital Association, using PitchBook data in its 2026 Yearbook, said U.S. firms closed 15,352 deals worth $320 billion in 2025, a 51% jump in value and the second-highest year on record. Artificial intelligence took 65.4% of that value.

The money clustered. Traditional venture fundraising was $67 billion across 585 funds. The top ten funds took $22 billion, or 32.9%, leaving $44.9 billion for the other 575. First-time funds fell to 101, the lowest count since 2011 and 77.9% below the 457 launched in 2021. Exits reached $217.1 billion across 1,463 deals, more than double the prior year, and still not enough to clear the backlog of private companies.

WHERE VENTURE CAPITAL ACTUALLY WENT

• Headline 2025 book: $320 billion of deal value, with AI at 65.4%.
• Fund formation: $67 billion into 585 funds, a third of it in ten vehicles.
• New managers: 101 first-time funds, the thinnest crop in 15 years.
• H1 2026 sequel: $412.7 billion of deal value, with rounds of $100 million or more taking 87.5%; deals under $100 million took $51.4 billion, or 12.5%, down from 43.8% of value in 2024 and 33.1% in 2025.

Bobby Franklin, president and chief executive of the NVCA, called 2025 “an industry at an inflection point, strong investment on one hand, constrained liquidity on the other.” For a family office that used to write seed checks, the relevant line is the last one. The dollars went to giant AI rounds. The early book shrank as a share of the market, and family-office patience went with it.

A California single-family-office chief executive told Campden that early-stage deals can still deliver windfalls, “this only happens 20 to 30% of the time. The trick is to minimize the number of failures.” That is a credit officer’s sentence, not a seed investor’s.

The 20% Property Sleeve in U.S. Offices

Real estate never left these portfolios. About 75% of the Campden offices held some property. Enthusiasm clustered in industrial and logistics at 30% and in residential housing at 23%. The report said affordability pressure had “made home ownership increasingly difficult and channeled demand into the rental sector.” Offices in Sun Belt states, where population and jobs have grown faster, were beating the national average, without a published spread.

Campden’s average North American weight was 18%, with a +13 net score for more. UBS’s 2026 Global Family Office Report, built on 307 offices in more than 30 markets with average net worth of $2.7 billion, split the map. Globally, real estate was 11% of 2025 allocations, and offices planning changes were aiming at 8%. In the United States the same survey put the average real estate sleeve at 20%.

So the “pour money into property” line is an American fact more than a global one. U.S. offices kept a landlord’s share of the book, with warehouses and rentals in front, while peers elsewhere started to pare the sleeve. The rental bid is the household side of the same yield hunt that sent them into private credit: they collect rent from people who cannot buy, and interest from companies that cannot borrow from a bank.

Improving Liquidity Was the Stated Mandate

Ask the same offices what they wanted in 2025, and the answers sit badly next to a warehouse and a five-year direct loan. The Campden ranking of primary objectives was a risk-off list.

PRIMARY OBJECTIVES FOR 2025

• Improve liquidity: 48% put this first, citing dry powder for a drawdown and the trouble of exiting private equity.
• De-risk the portfolio: 33%, aimed more at public holdings that could be hit by tariffs and geopolitics.
• New asset classes: 26% wanted fresh diversifiers.
• Geographic diversification: 25% wanted less political concentration.
• Cut private markets: only 12% named that as a primary goal.

Average expected return for 2025 was 5%, down from 11% in 2024. Only 13% of offices expected more than 10%, against 40% a year earlier. And 15% expected a negative year, against 1% in 2024. Cash was the 12-month favorite. On a two-to-five-year view they still ranked private equity and venture as the best risk-adjusted assets, which is how a 5% venture sleeve survives a -33 surprise.

A Florida single-family-office president put the duration mismatch in one breath: the family’s horizon is longer than a private equity fund, and “if we own good companies, we don’t necessarily want GPs to exit and send the money back.” That is patience. The 48% liquidity score is the opposite impulse, from offices that could not get their money out of older funds and did not want another decade in seed.

Compared to 2024, family offices have become notably more cautious stewards of capital. They have moderated their investment expectations, developed stronger governing guidelines and are actively looking for ways to be more efficient with their human capital.

Adam Ratner, Director of Research, Campden Wealth

Manju Jessa, who heads family office and strategic clients inside RBC’s enterprise group, still called family offices “well-suited to the private markets” because they can be patient. The allocation table says the patience is being spent on credit and buildings, not on the next unproven company.

Why Seed Managers Lost the Family-Office Meeting

Returns are the clean explanation. Process is the one that lasts. A lean family office that also has to cover private credit, buyouts, buildings and the public book does not have a calendar for 200 seed managers. Charles Hudson, managing partner at Precursor Ventures, has been saying the quiet part to anyone who will listen: an allocator with a low hurdle rate and a short staff will write two large checks to platform firms that span pre-seed to growth rather than ten small ones, because ten funds means ten meetings, ten re-ups and ten annual sessions.

That is how early-stage venture loses even when the long-run math still looks decent. Campden’s own 25-year yardsticks put private equity at about 12% a year and venture at about 8%, against 9% for the S&P 500. The offices still rank those sleeves first for risk-adjusted return. They are not putting the next dollar there, because a 5% portfolio target makes a ten-year lockup look expensive, and because the staff who would underwrite seed are busy underwriting coupons.

A California chief executive in the Campden study described the generational fork in similar terms. A first-generation founder will live with a concentrated, illiquid book, “because that’s his professional life.” A fourth- or fifth-generation office “will be highly conservative because they need liquidity and duration to support the income needs of family members.” Almost half of the North American offices, 47%, expect control to pass within ten years. The next generation is the one collecting rent and interest.

Sixty Percent Are Rewriting Allocations Again

The 2025 rotation was not the last one. UBS found that 60% of family offices plan changes to strategic allocation over the next 12 months, the highest reading in that survey, against 35% in 2025. A U.S. client told UBS, “The world has changed in the past one-and-a-half years, so it makes sense that people are responding to that.”

THE ALLOCATION CLOCK

1. April to August 2025: Campden and RBC poll 141 North American offices. Private markets sit at 29%. New money is earmarked for direct buyouts, private credit and property. Expected return is 5%.
2. October 16, 2025: The North America Family Office Report is published in Toronto and Minneapolis. Venture has fallen down the new-money ranking after a -33 year-to-date surprise.
3. May 28, 2026: UBS publishes its 2026 global study. Sixty percent plan another shift. Global real estate is marked for a cut from 11% to 8%. U.S. offices still hold 20% in property. Global private debt is 3%.
4. September 10, 2026: Houlihan Lokey reports that 12% of loans to the smallest private-credit borrowers now price below 90, against about 1% in 2023.

H1 2026 venture fundraising, $72.4 billion across 405 funds, almost matched the whole of 2025, and still ran through a handful of names. The early companies that used to court family offices are competing for the 12.5% of venture value that is not a megadeal. The offices themselves are competing for yield in a credit book whose weak tail is now visible on a mark-to-model screen.

The 4% private-credit sleeve and the 18% property sleeve are doing a job banks and public landlords used to do. That was the point of the 2025 move. The bill for it shows up first in the smallest EBITDA bucket, at 12 cents below par, in the asset class they chose because the borrowers were willing to pay.

Disclaimer: This article is news reporting and analysis of family-office surveys and private-market data, and it is for information only. It is not investment advice, a recommendation to buy or sell private credit, real estate, venture capital or any other asset, or a forecast of returns for any family office or fund. Readers who are considering changes to a portfolio should consult a licensed financial adviser or wealth manager who can review their own holdings, taxes and liquidity needs. Figures and allocations here reflect the Campden Wealth / RBC, UBS, NVCA / PitchBook and Houlihan Lokey publications cited, and those readings can change with the next survey or valuation cycle.