Direct indexing's trillion-dollar forecast is a platform stat, not demand

The direct indexing numbers are familiar by now: assets doubling, adoption curves bending upward, analysts projecting a trillion-dollar category inside a decade. A commentary WealthManagement.com published September 15, from an author who says the work of building this business has occupied nearly three decades, accepts every one of those numbers, and to that author the growth figure is the least interesting fact about the category. The column points instead at the gap the statistic hides — between what the technology and platforms can now do and what advisors understand well enough to sell.

The author is specific about where the gap does not sit: custodial connections, account minimums, and platform integration, the reasons most often offered, have all been built out, the column says. That leaves the telling, and the advisors who have not worked out how to describe the product in a client meeting.

The column traces the capability — holding a benchmark's constituent securities in a separately managed account and harvesting losses at the security level — back more than two decades. For most of that span, the tax work happened one account at a time and one manager at a time, so a client who needed large-cap exposure alongside small-cap and international sleeves often ended up with several custodial accounts, each optimized in isolation, and nothing in the stack looking across the household.

The change was where the machinery sat: direct indexing moved into the unified managed account, where an overlay team applies tax management across every sleeve and every manager inside a single custodial account, while the harvesting itself stayed as it was and the number of accounts one operation could reach from one place grew. That relocation took the category from a boutique service to something a platform can run across tens of thousands of accounts, and it is why the column treats the ceiling as effectively gone.

Once tax management is applied across every sleeve and every manager in the account, loss harvesting stops being a reason to choose one manager over another and becomes a property of the account structure. Standalone tax-managed SMA providers built their pitch on precisely that harvesting, which suggests the overlay squeezes the differentiator rather than the demand; the beneficiary of the move is the custodian and the platform running the overlay.

Put yourself in the platform's seat and the appeal of centralization is plain: tax work that no longer scales with the number of custodial accounts makes each incremental account cheaper to serve. That is a read on platform economics rather than anything the column discloses, but it explains why the build-out continues regardless of adoption, while the advisor's cost of a conversation about a statement holding 200-plus positions does not fall. The growth statistic blends the two curves and prints the friendlier one.

Read against that history, the trillion-dollar projection looks like a capacity forecast wearing demand's clothes — a statement about how many accounts a platform can administer once the overlay is centralized. Capacity forecasts come true for their own reasons, and they can come true with barely a client understanding what sits in the account. This publication has argued that in private markets the distribution network became the product and the platform itself became the acquisition target; direct indexing is the same trade with the tax work attached, and the value went to the wrapper. One recent example from these pages: Rockefeller, BNY and Focus took stakes in Luminary's $22 million round in September, backing software that decides where a client's assets go at death.

Ask advisors why they have not bought in, the column says, and three answers recur: they want an actively managed strategy whose benchmark-beating justifies the fee; they are uneasy handing a client a statement running 200-plus small positions; the account minimum is out of reach for the client in front of them. The column's position is that all three have been resolved, the technology being the finished part of the problem and the advisor's narrative the open one.

The list is worth less in three parts than in one, because the statement and the fee are a single conversation: a long list of small positions forces an advisor to explain what the client is paying for, and tax alpha does not render as a figure that can be laid beside an index return. A statement is not frightening because it is long; it is frightening because it invites a question the advisor cannot answer with a number. The minimum is the outlier, the one objection a platform can retire without an advisor changing anything, and the column counts it as already retired.

Pushing the column's logic a step further: whoever puts a household-level tax number in front of an advisor — dollars harvested, measured against that client's plan — before the client asks for one will take the adoption argument out of the meeting. That is a reporting problem as much as an investment one. The AI fight has already been through this: the model is becoming the commodity while the advisor relationship and the integration plumbing are what get paid for, and direct indexing's version of the rule is a number the advisor can point at.

The price structure gives it away. While the overlay is billed inside a sleeve, it competes for the same basis points as every active manager, on a return chart where tax alpha does not appear — which is precisely the competition the column says advisors are judging it by. Move the overlay into the household's advisory fee, where it becomes one more thing the advisor does for the relationship, and the objection at the top of the list loses its footing. Whichever platform bills it that way and prints harvested losses on the same page as the allocation will learn whether those three objections were real or simply unanswered. That test costs less than the next trillion-dollar projection.