If your wealth manager has pitched "co-invest opportunities" or "specialist healthcare access" lately, two stories from this week change what those pitches actually mean. They run in opposite directions.
The first: family office direct investing more than doubled last year. Healthcare sits second-deepest in major-FO portfolios, right behind AI. And federal money for biomedical research is on track for a $5bn cut. Private capital is rushing into a gap public capital is backing out of.
The second: on Sunday, Citi named — in its own corporate language — which part of the adviser job AI now does. Not a vague productivity story. A specific function, called by name. "Coordinator."
More sectors for capital to flow into. Less work for advisers to be paid for. Both shifts show up in the same conversation any founder is having with a wealth manager right now.
This Week in 30 Seconds
- Family office direct investing doubled in 2025. $12.9bn across 158 transactions. Healthcare sits second only to AI as the deepest investment theme.
- Federal funding is pulling out as private capital pours in. The Trump FY2027 budget cuts NIH by $5bn. Specialist family-office syndicates are filling the gap.
- Citi named the adviser function AI replaces. The "coordinator" role: meeting prep, portfolio data, scenario modelling. Judgment, access, and relationship are what stays.
- Dispersion is the issue. Major-FO concentration shows where institutional gravity sits, not where average outcomes land. Biotech failure rates still sit close to 90%.
Why Family Office Capital Is Concentrating in Healthcare
In April, family offices did 55 direct deals, up from 39 in March. Almost a third went into healthcare or life sciences. Fintrx gave the data to CNBC and called it a rebound, with March slumping because the Iran war broke out. April was always going to look better. The bounce isn't the story.
Family-office direct investing hit $12.9bn across 158 transactions in 2025. That's the highest annual total since at least 2021, and more than double 2024. JPM's latest survey puts direct investing third in family-office priorities, behind only cash and estate planning. Private markets take 30.8% of the average book. AI is the top investment theme. Healthcare comes second.
AI is consensus. Every wealth-management deck talks about it. Healthcare is quieter, deeper, slower-cycle, and that's where the second-largest pool of major-FO attention is going.
Federal money is moving the other way. The Trump FY2027 budget cuts NIH funding by $5bn, axing three of 27 institutes and cutting ARPA-H by 37%. Last year's proposed $19bn cut was rejected by Congress, which actually raised NIH funding by $400m, so this isn't certain. But while the politics play out, private capital is filling the gap public money has historically backstopped.
April alone gives a clean snapshot of what that looks like at the deal level:
- Stipple Bio. $100m oncology Series A, a16z-led, with Emerson Collective participating through Yosemite, Reed Jobs's oncology fund that spun out of Emerson in 2023.
- Ultralight. $9.3m seed for AI-personalised healthcare; Emerson again on the cap table.
- Exciva. $62m Series B for Alzheimer's agitation treatments; Dolby Family Ventures joined.
Different deals, different stages, but same pattern. Major-FO money pouring into specialist healthcare bets at exactly the moment public funding is contracting.
Yosemite is the structure worth noticing. Emerson didn't run a direct biotech program from inside the family office. They incubated a specialist vehicle, spun it out with $200m+, and now invest as an LP. Hillspire ($28bn AUM, Eric Schmidt's family office) does similar work in AI. That's how the big offices do it now: fund a team that hunts deals full-time, keep the LP-level discretion.
Most $5M–$100M founders can't access these deals directly. The cheque sizes are too big and the relationships too established. The spillover is the access point: co-invest SPVs, sector-specialist syndicates, and the secondary deal flow that forms wherever institutional capital concentrates.
Why I'd Be Cautious Anyway
The two-decade history of family-office direct investing is full of programs that stalled. Three years ago, the trade press was running pieces about family offices postponing direct deals after the returns came in. At a Bloomberg family-office summit in March, someone called it "truffle hunting": the slow business of separating real opportunities from a flood of mediocre deal flow. Two-thirds of family offices say due diligence is the hardest part of running this themselves.
Biotech is harder still. Seed and Series A biotech funding is running at roughly 62% of Q1 last year, even as major-FO healthcare interest hits a record. Pharma M&A surged at the same time: $40.9bn upfront in Q1 versus $28.7bn the year before. Capital is concentrating at late-stage and exit, not seed. The middle is hollowing.
Drug development takes 10 to 15 years. Failure rates near 90%. The average across 158 family-office deals hides a return distribution wider than most balanced portfolios are built to absorb. None of that says avoid healthcare exposure. It says know the difference between getting access to institutional deal flow and getting whatever's left after the institutions have picked.
The Founder's Guide to Building a Private Investment Office covers the structural side: how a sub-$100m pool can build sourcing infrastructure without trying to be Hillspire. The real question for the next 12 months isn't whether to overweight healthcare. It's whether your current setup would even spot a serious specialist syndicate if one got offered.
What Citi Just Itemised About the Adviser Job
On Sunday, Citi launched Arc, its internal platform for building and governing AI agents across the bank. Arc fits a 4-month sequence:
- February: CitiScribe note-taking for advisers
- April: Client 360 dashboards and AskWealth CIO chat
- 22 April: Citi Sky, a client-facing AI avatar for Citigold accounts at $200k and up
- 3 May: Arc, the agent infrastructure underneath all of it
Morgan Stanley, BNY, and Wealthbox are doing similar work. Within 24 months, AI handling meeting prep and portfolio analysis will be table stakes, not a differentiator.
Most banks talk about AI in vague enabler terms. Citi did something different. Their corporate words: "The role of the banker, therefore, evolves more decisively from coordinator to architect and adviser." Translated: the hours an adviser spends gathering portfolio data, modelling scenarios, prepping for a client meeting. That's the work AI agents do now.
Coordinator work is being commoditised. Architect-and-adviser work is what's left. If coordinator-function work is now agent work, the price moves toward agent cost. The parts that hold their fee carry the weight: judgment, sector access, network introductions, the call you can make at 9pm when a deal is moving.
The same agent tools available to Citi's advisers are increasingly available to founders running their own setups. The adviser's edge narrows to access, judgment, and the relationship. April's Wealth Architecture Is Running Behind the Liquidity covered the same lag: wealth structures running behind the liquidity event that creates them. The fee question is the same one. Of the work the firm is charging for, how much is coordinator work, and how much is judgment AI can't replicate yet?
On the Radar
Fed's Barr names "psychological contagion" risk in private credit. Fed Governor Michael Barr told Bloomberg that stress in the $1.8tn private credit market could spread into corporate bonds. He singled out PIK loans as the opacity risk. Question for the next quarterly review: what share of your private-credit book is paying interest in PIK form? Link.
Wealth Enhancement and Steward Partners book another $768m of acquisitions. Wealth Enhancement bought Lake Tahoe Wealth Management ($318m), then added two more deals the same week. Steward Partners picked up Tampa-based Jazz Wealth ($450m, 3,500 clients). If your adviser sits in a small-to-mid RIA, there's a real chance the firm is talking to a buyer this year. Two questions to ask: who owns the firm in 24 months, and what changes for clients if it sells. Link.
Worth looking at: how Nasdaq turned QQQ into its biggest revenue line. Marc Rubinstein's Bye the Index covers QQQ ($456bn AUM) and how Nasdaq's index-licensing line ($854m TTM) now beats cash equities, data, and listings combined. Passive drives a third of US equity trading. For anyone with concentrated tech equity from a recent listing, index inclusion has become a real capital-flow event worth understanding before any disposal plan. Link.
Worth looking at: INSEAD's annual ETA conference and the Stanford numbers underneath. INSEAD ran its ETA & Search Fund Conference in Fontainebleau on 9 May. Stanford's most recent search-fund study anchors the category at 35.1% mean IRR and 4.5x ROI. Headlines are real; dispersion underneath is wide. If you're thinking about backing a searcher, source 2 or 3 deal memos and stress-test them against your own criteria first. Link.
Worth looking at: a licensed therapist's account of the post-exit identity crash, written from inside it. Annie Wright exited her own multi-million-dollar company and now works clinically with founders going through the same transition. Her recent essay frames the post-exit period as a clinical pattern, not founder folklore. Useful as a self-diagnostic in the months before or after the wire transfer. Pairs with Founder's Identity Crisis After Exit and What Founders Actually Do After Exit. Link.
What These Two Shifts Change About the Adviser Conversation
Both shifts show up in the same conversation any founder is having with a wealth manager right now. Capital is pouring into healthcare at scale, but the pools doing it operate well above the $5M–$100M tier. And some of the work that wealth manager is paid for is being absorbed by software.
Next time someone pitches you healthcare exposure, ask what the access point actually is. A co-invest into a major-FO-sponsored deal is one thing. A fund-of-funds with three layers of fees underneath is another. Both will look identical from the outside.
Next time your wealth firm talks about its fees, ask which work the fee is paying for. Citi has put it in writing. Morgan Stanley, BNY, and Wealthbox are doing the same thing more quietly. Coordinator work is moving to software. Judgment, sector access, and relationship hold their value.
The rest is being repriced whether your firm tells you or not.
New on the Site
Two Thursdays back, the long-form piece on holding structures for global founders covered the jurisdictional architecture across the US, UK, Singapore, and UAE that shapes what access points and tax exposure a portfolio actually has. With family-office direct investing concentrating into specialist syndicates this year, where you hold capital starts to matter as much as which deals get to your desk.
Read it: Holding Structures for Global Founders
This was Capital Signals — weekly briefings on what's reshaping founder strategy on wealth.
Go deeper: Playbooks · Wealth Architecture · Investment Strategy · Business Building · Life Design
Found this useful? Forward it to a founder who's thinking about this stuff. Got a question or disagree with something? Get in touch.
New here? Subscribe for one email a week.
