Essay · Agentic Wealth
Agentic wealth — the complexity tax on financial advice
The wealth management industry is short 100,000 advisors and adding them at a crawl. That shortage is a complexity tax on the advisor's time — and it sits next to an older, quieter tax paid by clients who can't always tell whether the person across the table is a fiduciary or a salesperson wearing the same title.
A shortage that looks like arithmetic, not sentiment
McKinsey estimates the U.S. wealth management industry will be short 90,000 to 110,000 advisors by 2034 — 30 to 37 percent of current headcount. Cerulli Associates puts the number of advisors planning to retire in the next decade at 105,887, and finds that 26 percent of them have no defined succession plan. The CFP Board reports that 51 percent of Certified Financial Planners are already over 50. Meanwhile the advisor workforce has grown at roughly 0.3 percent a year over the last decade — essentially flat, against a client base whose wealth keeps compounding.
This is not a story about a profession falling out of favor. It is a story about a profession that has not figured out how to produce enough of itself, fast enough, while the assets it is responsible for keep growing. Arithmetic like that does not resolve by wishing for more career-changers to sit the CFP exam. It resolves by changing what one advisor is capable of doing.
Where the tax is actually paid
Research from Kitces.com on how financial advisors actually spend their time found that the typical advisor spends about 20 percent of a working week meeting with clients — the part of the job clients are paying for — while 45 percent goes to preparing for those meetings, running planning analyses, and managing investments behind the scenes, and the remaining 35 percent splits between business development and practice administration. The same research found that the most productive advisors spend only about 10 percent more time in client meetings than the least productive ones, but claw back roughly 200 hours a year by cutting the back-office work underneath.
Two hundred hours is five work-weeks. Multiply that across a shrinking workforce carrying a growing client base, and the shortage numbers above stop looking like a talent-acquisition problem and start looking like a design problem. The tax is not exotic. It is the reconciliation between the planning software and the custodian portal that does not talk to it. The proposal rebuilt from scratch because the last one lived in a different template. The client update that has to be manually cross-checked against three systems before it can go out. Every one of those is small. All of them, together, are the 45 percent.
A second tax, paid by the client
The advisor's 45 percent is not the only complexity tax at work in this business. There is a second one, and it falls on the client: most people cannot tell, from a title or a business card, whether the person recommending a product is bound by a fiduciary duty to act in their interest, or is held to something narrower. An investment adviser representative owes a fiduciary duty under the Investment Advisers Act of 1940. A broker-dealer representative, even under the SEC's Regulation Best Interest — adopted in 2019, binding since June 2020 — owes a standard of care that is real and enforceable, but distinct: a recommendation has to clear a best-interest bar, not merely a suitability one, yet the underlying compensation can still run through commissions and proprietary products in a way a fee-based fiduciary's cannot. Two people can wear the same title, sit across the same kind of desk, and be paid in fundamentally different ways — one on a fee tied to the client's outcome, one on a commission tied to what was sold. "Advisor" describes both. The client is left to resolve that ambiguity alone, usually without knowing there is anything to resolve.
The counter-argument deserves the same weight every position on this site gets. Reg BI tightened the standard, not just the paperwork — broker-dealers now have to document that a recommendation clears a best-interest bar, and the SEC has brought enforcement actions under it since. Plenty of commission-compensated representatives do right by their clients every day, and plenty of fee-based advisors have found ways to drift from the spirit of their duty without ever violating its letter. A legal standard is not a perfect proxy for the character of the person applying it. But a standard that is narrower on paper, before any individual's conduct enters the picture, is still worth naming — because the client cannot see character in advance. They can only see the title, and the title does not disclose the standard behind it.
This is the same principle argued elsewhere on this site about who gets access to markets — protection by transparency, not exclusion — extended to who gets to know the rules the person across the table is actually playing by. Agentic rails do not resolve the incentive question; no software resolves an incentive. What they can do is make the disclosure automatic and legible at the moment it matters, rather than a document handed over once at account opening and rarely opened again: fee structure, standard of care, and compensation source, surfaced plainly, every time, without a compliance team producing it by hand.
What the paperwork buys
Not all of the advisor's 45 percent is waste, and the case for agentic wealth has to say so plainly. Some of it overlaps directly with the second tax just described: suitability documentation and know-your-customer records are how a broker demonstrates Reg BI compliance, and investment policy statements are how a fiduciary demonstrates the analogous, stricter duty. Compliance recordkeeping exists because advisors owe their clients a duty of some kind, and because the paper trail is what makes that duty enforceable rather than aspirational. An automation that speeds up documentation without preserving its evidentiary integrity has not solved the advisor's problem — it has traded a time problem for a liability problem, which is worse. Anyone selling wealth management technology who treats compliance as friction to be engineered away is not building for advisors; they are building a future audit finding.
This essay has to hold all of this at once: the complexity tax is real, on both sides of the table, and so is the reason some of that complexity exists.
What agentic AI actually changes
The distinction that matters is between the advisor's judgment and the layer underneath it. Judgment is irreducibly human: reading a client's fear correctly during a drawdown, weighing a family's actual values in a wealth transfer, knowing when a founder's instinct to hold or sell is conviction or panic. No agent should touch that, and none of what follows proposes to.
What agentic AI can collapse is the layer underneath — reconciling positions across custodians into one accurate picture instead of three inconsistent ones, turning a client conversation into structured notes and a compliant record instead of a memory and a manual write-up, drafting the first pass of a plan update against the household's actual data instead of a blank template, generating a rebalancing trade list that a human still reviews and approves rather than assembling one by hand against a spreadsheet. Rebalancing by rule, not by mood, has always been the discipline evidence-based investing asks for. Agentic rails are what make that discipline cheap enough to run consistently across every household an advisor serves, not just the largest ones.
This is the build WilliamFrank has been documenting in the open, on the same premise as the founding essay on this site: the tools that used to take a team and a decade to assemble are now available on demand, and the advisors who adopt them first won't be the ones who cut corners — they'll be the ones who get their own time back.
The floor rises twice
The same pattern showed up in an earlier piece on this site about wealth creation more broadly: the floor rises once from what a few builders create, and again as the tools that used to be scarce become available to everyone else. It applies here in a narrower, more concrete way. It rises once because an advisor freed from reconciliation and re-typing can serve more households at the same standard of care — the 200 hours Kitces measured, scaled across a career. It rises again because the advisors who benefit most from that leverage are not only the largest teams with the biggest operations staffs. A well-built agentic rail narrows the gap between a two-person practice and a hundred-person enterprise faster than any prior wave of advisor technology has.
A less taxed advisor, not a smarter one
The shortage does not close by producing advisors faster than the industry ever has — the workforce grew 0.3 percent a year for a decade, and nothing about the incoming class of entrants suggests that reverses on its own. It closes, if it closes, by making the advisors already in the field capable of serving more households well, with the same fiduciary rigor and none of the corners cut. That is the case for agentic wealth. Not a smarter advisor. A less taxed one — with the hours back to spend on the twenty percent of the job that was always the actual point.