A Family Office Chief Explains the 70% Failure Rate. Ask for the Footnote.

Reading Schwab’s family office head against his own arithmetic: three figures that do not reconcile, and one fee shift that explains the pitch.

AsiaPresswire – Sam Kang runs Schwab Family Office, an invitation-only unit of Charles Schwab that serves single family offices and multi family offices. In January 2026 he sat down with Steve Sanduski for the Barron’s Advisor podcast and explained, at length and in good faith, what ultra-high-net-worth families want from their advisers. The episode is worth an hour because he is unusually specific for someone in his position. He gives numbers, names the surveys behind some of them, and says plainly which parts of the trade are harder than they look.

It is also worth an hour for a second reason. Three of the numbers he uses in the first fifteen minutes do not reconcile with each other, and nobody on the call notices.

That is not a gotcha. It is the most useful thing in the transcript. When a market’s size, its failure rate and its growth rate are all quoted from memory, the shape of the pitch becomes visible: the industry sells a service for a problem whose dimensions it has never had to defend.

Why this transcript is worth the time

Most wealth-management interviews are one long advertisement with a question mark somewhere in the middle. This one is better than that, for three reasons.

First, the guest is a vendor selling to vendors. Schwab Family Office does not chase families directly; it supports the offices and advisers that do, which means Kang has to describe the industry’s economics accurately enough for professional buyers to nod along. Sales talk to professionals is more informative than sales talk to consumers, because the buyer can check.

Second, he quantifies. He puts a number on the failure rate of wealth transfer, on the growth of the ultra-high-net-worth household count, on the share of the segment held by independent advisers, and on how multi family offices now price. Numbers invite arithmetic, and arithmetic is where a pitch either holds or does not.

Third, he admits a limit. Asked about artificial intelligence, he says adoption is at a very early stage. That single sentence is worth more than the rest of the interview, because it contradicts the conference circuit without sounding like a contradiction.

The argument he is actually making

Strip the framing and the thesis is this: rich families do not fail because their portfolios underperform. They fail because the family has no governance, no shared purpose and no relationship with the adviser who will outlive the founder. The adviser’s job is therefore to become the hub, not the stock picker. Everything else, the tax work, the philanthropy, the bill paying, the family constitution, is a spoke that hangs off that hub.

It is a good argument. It is also, conveniently, an argument for buying more services from a larger platform.

Table 1: The figures Kang used, what they claim, and whether the transcript names a source

Source: Barron’s Advisor , transcript of the episode with Sam Kang, 20 January 2026

The table is not an accusation. An executive on a podcast is entitled to speak from memory. But it is a map of what the industry treats as settled and what it treats as decoration, and the pattern is clear: the only figure with a named source is the one about pricing. Everything about the customer is folklore.

Where each speaker stands

An interview is evidence about the speaker as much as about the subject, so it helps to label the positions before quoting them.

• Kang speaks as a platform executive whose customers are family offices and advisers, not families. His revenue rises when more advisers serve more rich households through Schwab’s custody and technology.
• Sanduski speaks as a coach to advisers. His audience is the people Kang sells to, which makes him a useful proxy for what the trade believes about itself.
• The Family Wealth Alliance survey, the only named source, was conducted with partners, which is normal for industry surveys and worth remembering when the number is a round 70%.
• The families themselves are absent from the transcript. Not one is quoted, named or described in detail.

That last point is the one to hold onto. A forty-minute conversation about what ultra-high-net-worth families want contains no family.

Three figures that do not reconcile

Start with the money, because the arithmetic is simple and the conclusion is not.

In the January 2026 interview, Kang says USD 124 trillion will change hands over the next two to three decades, and in the same answer that USD 2 to 3 trillion is transferring every single year. Divide the first by the second in 2026 and the horizon comes out at 41 to 62 years, not 20 to 30. For both statements to be true, the annual figure would have to be about USD 4.1 trillion at a 30-year horizon or USD 6.2 trillion at 20 years.

Both numbers are standard in the industry. The 124 trillion figure is the one that appears in every wealth transfer deck published since 2020; the 2 to 3 trillion figure is what the same decks use for annual flow. They cannot both describe the same phenomenon over the same period. One of them is a stock measured at a point in time, the other a flow, and somewhere between the two the units were lost.

Now the segment size. Kang puts the ultra-high-net-worth space at USD 17 trillion, with about USD 6 trillion already with multi family offices. Later in the same January 2026 conversation he says independent registered investment advisers hold 32% to 34% of wealth management overall but only 6.6% of the ultra-high-net-worth segment.

Here is the problem. A multi family office is, by his own definition, a registered investment adviser. By his own 2026 numbers, if USD 6 trillion of a USD 17 trillion segment sits with multi family offices, then independent advisers hold about 35% of that segment, not 6.6%. The two statements can be reconciled in exactly two ways: either the 6.6% is measured on a different denominator, such as households rather than assets, or the multi family office total includes bank-owned platforms that are not independent. The transcript does neither, and the difference between 6.6% and 35% is the difference between a market with room to grow and a market that is already two-thirds intermediated.

The third is the range he gave in January 2026: “anywhere from five to 6,000 family offices globally.” A range that spans a thousand institutions, delivered without hesitation, in answer to a definitional question. If the count is uncertain by 20%, every market-share claim built on it inherits the uncertainty.

The 70% number and its long tail

The failure rate is the load-bearing statistic of the entire family office industry, and it arrives without a citation.

“By the end of the second generation, 70% of families, the wealth transfer does not succeed past that second generation. By the third generation, only 13% survive. By the fourth generation, it’s only 3%.”

The figure is one of the most recycled numbers in wealth management. It is usually traced to a study of wealthy families published more than a decade ago, and it has been repeated so often that its original sample, definition of failure and treatment of inflation have all been sanded off. Kang does not name it, which is normal. Sanduski does not ask, which is also normal.

What makes the number useful is not whether it is true. It is what it justifies. In 2026, a 70% failure rate converts a discretionary service into a necessity, and it does so without requiring the adviser to demonstrate that their own clients beat the base rate. The statistic is unfalsifiable in practice, because the families that fail are by definition not around to be surveyed.

There is a version of this argument that survives scrutiny: family wealth is fragile, governance helps, and the failure mode is usually human rather than financial. That version needs no percentage. The percentage is there to end the conversation, not to start it.

What the fee shift reveals

The one number in the interview with a named source is also the one that explains the business model, which is probably not a coincidence.

“We’ve seen a big shift in trend there also where these multifamily offices typically used to charge in a very similar way as independent RIAs in a AUM-based fee model. What we’re seeing more and more now over the last 10 years, in the latest survey that our partners at Family Wealth Alliance has done, 70% of multifamily offices now have some type of hybrid approach, meaning they offer a lot of these additional capabilities in more of a retainer fee model.”

He then gives the arithmetic: “you might be charging 40 basis points for the investments and financial planning, but there might be the family dynamics piece or bill pay and you can charge retainer fees. Overall, then you get into more like that 70, 80 basis points charged for the family.”

Read that as a business model and it is elegant. Asset-based fees compress as index funds get cheaper and as clients negotiate. Retainers do not compress, because they are priced against complexity rather than against performance. A firm that moved from 80 basis points all-in on assets in 2015 to 40 basis points plus a retainer in 2026 lands at the same 70 to 80 basis points, but with a revenue line that does not fall when the market does.

The retainer also changes what the firm is selling. An asset-based adviser needs assets to grow. A retainer adviser needs the family’s affairs to be complicated enough to justify a fixed fee, which is why governance, family constitutions and multi-generational education appear in the same conversation as pricing. Complexity is not a side effect of serving rich families. It is the billable unit.

None of this is sinister. It is a mature industry finding a revenue base that survives fee compression, and the families who need the service get it. What it does mean is that the retainer is priced against the family’s admin load rather than against the adviser’s skill, and admin loads grow with every entity, trust and jurisdiction a family adds. But it does explain why the 2026 pitch leans so heavily on the 70% failure rate. The product is not returns. The product is continuity, and continuity is sold against a risk that nobody can measure.

The AI anecdote is doing a job

Asked about artificial intelligence, Kang is careful, and then not careful.

“I think, one, many of the family offices are trying to adopt artificial intelligence. I would still say it’s at the very early stages. Much of it is focused on building scale, creating better knowledge components.”

That is an honest answer, and it is the most credible sentence in the interview. A vendor with an AI product to sell would not say adoption is early.

Two sentences later: “very high-end multifamily offices in the Silicon Valley, they’ve actually brought in a AI agent as part of the senior management team as one example.”

An AI agent on the senior management team is a wonderful phrase and an empty one. It could mean a model that drafts investment memos for the investment committee. It could mean a chatbot with a title. It could mean a language model configured to answer partners’ questions about the family’s trust documents. The example is doing the work that anecdotes do in every sales conversation: it makes an early-stage technology feel inevitable without committing to a definition.

The honest reading of both sentences together is that the industry has a slide about AI, a handful of experiments, and no operating model. That is fine. It is also the opposite of what the conference circuit says, which is why the first sentence deserves more attention than the second.

Questions the transcript does not answer

Five questions would have changed the value of the hour, and none of them was asked.

• What does Schwab earn from a family office that custody assets with it, and how does that compare with what the family pays the office? The conflict is structural and disclosing it would have made the rest of the interview more useful, not less.
• How many of the families that “fail” do so because of governance, and how many because the business they owned went into decline? The two failure modes have opposite remedies.
• What is the retention rate of advisers’ clients across a generational transfer, measured rather than asserted? The claim that 70% of adult children would fire their parents’ adviser implies a 30% retention rate, which would be a catastrophic number for any other subscription business.
• If 45% of new business comes from referrals, why does the industry spend so heavily on websites and branding? Sanduski raises this himself and the answer is a shrug about “yes and”.
• How large is the retainer book, in dollars, across the multi family offices that report hybrid pricing? A 70% adoption rate for a pricing model says nothing about how much of the revenue it produces.

The absence of these questions is not a failure of the interviewer. It is what happens when both parties share an interest in the frame. The frame is that families need more advice, and every question that would test the frame is a question that would slow the conversation down.

What the transcript exposes

Put the pieces together and a coherent picture emerges, and it is not the one on the tin.

The family office industry is real, growing and useful. The number of ultra-high-net-worth households has grown enormously since 2011, by his own figure more than 700%, against 84% for the merely affluent. That growth is the actual engine of the business, and it needs no failure statistic to explain itself. More rich families means more complexity, more advisers and more demand for the administrative layer that a family office provides.

The failure narrative is a separate product. It sells continuity to people who are afraid, and it is priced as a retainer because retainers survive fee compression. The evidence behind it is a decade-old study that nobody in the room can cite, restated as a percentage that ends the conversation.

The most telling moment is not a number. It is the definition of a multi family office as a registered investment adviser that serves several ultra-high-net-worth families at once, followed by the observation that the complexity curve jumps at around USD 30 million of net worth. That is the customer qualification line as stated in 2026: below 30 million, complexity is manageable and fees are competitive; above it, complexity compounds and the family starts paying for coordination rather than for returns.

Everything else in the interview, the family constitution, the hub and spoke, the AI agent, follows from that line. The industry is not in the business of making families richer. It is in the business of making wealth survivable, which is a harder job, a more defensible fee, and a much better story to tell at a conference.

Sourcing note

Data as of October 2026. This piece is built on a single published transcript: the Barron’s Advisor podcast episode with Sam Kang, host Steve Sanduski, published on 20 January 2026 by MarketWatch. All quotations are taken verbatim from that transcript, which was produced by a transcription service and carries the publisher’s caveat that it may be updated. As published on 20 January 2026, the page does not print a runtime; the episode follows the show’s standard full-length interview format, and the transcript runs to roughly 9,000 words.

No family, family office or client is named in the transcript, so none is named here. The arithmetic in the section on unreconciled figures is mine: dividing the stated transfer total by the stated annual flow, and dividing the stated multi family office assets by the stated segment size. Both operations use only numbers the guest supplied. Where a reconciliation is possible, the two candidate explanations are given and neither is chosen, because the transcript does not settle it.

The label system in the section on speaker positions is editorial. It records what each participant’s commercial interest is, not what they intended.

Limitations

• One interview cannot establish what the industry believes; it establishes what one senior vendor says when a friendly interviewer asks.
• The 9,000 word estimate is a word count of the transcript page, not an official runtime.
• The arithmetic on 124 trillion and 2 to 3 trillion assumes both figures refer to the same population and period, which may not be the case.

Sources

• Barron’s Advisor via MarketWatch , Sam Kang on what ultrawealthy families expect from advisers, 20 January 2026
• Schwab Family Office , service description for single and multi family offices (site home)
• Knight Frank , The Wealth Report 2026, ultra-high-net-worth population estimates (site home)

This article was prepared by the expert team at Global Infinity Labs, drawing on insights and data provided through its MCP data services.

About Global Infinity Labs

Global Infinity  Labs is an AI innovation company specializing in MCP data services.

Email: support@globalinfinitylabs.com