The personalization wave: why direct indexing is eating the SMA

If you want to understand where wealth management is going, stop looking at the funds and start counting the accounts. A pooled index fund is a single portfolio shared by millions of investors, and for fifty years that shared-portfolio model was the cheapest, most scalable way to deliver diversified market exposure. It worked because the unit of production was the fund, not the client. You built one portfolio and sold it a million times. The economics of that arrangement were beautiful precisely because nothing about the portfolio had to change when a new investor showed up.

Direct indexing breaks that arrangement on purpose. Instead of buying a slice of a shared fund, the investor owns the underlying securities directly, in a separately managed account that can be shaped around their specific situation — their tax lots, their concentrated stock position, their values screens, their state of residence, the gain they are sitting on and cannot afford to realize. The investor gets a portfolio that tracks an index closely while doing things a fund legally cannot do for them, the most important of which is harvesting losses at the individual-security level to generate a real, spendable after-tax benefit. That is a genuinely better product for a large and growing class of investors. It is also a fundamentally harder thing to manufacture, and that difficulty is the whole story.

The shape of the market

The headline numbers are easy to find and easy to misread, so it is worth being careful. Industry researchers who track this category — Cerulli Associates being the most frequently cited — have for several years projected that assets in direct-indexing and personalized SMA strategies would grow at a double-digit annual rate, materially faster than the broader managed-account universe and far faster than traditional pooled funds. The exact figures depend on how you draw the boundary around "direct indexing," and any single number you see quoted should be treated as an industry estimate rather than a measured fact. But the direction and the slope are not in dispute, and the slope is what matters.

Notice what the chart is actually measuring on its steep axis. It is not assets — it is accounts. This distinction is the one that operations teams feel in their bones and that strategy decks routinely miss. Assets can grow by a market rally without adding a single new problem to solve overnight. Account count growing is something else entirely: every new personalized account is another optimization problem that has to be constructed, rebalanced, harvested, and reconciled on its own schedule, with its own constraints, against its own benchmark. The work scales with the number of distinct problems, and the number of distinct problems is growing faster than the assets.

Why now, and not five years ago

Several things had to become true at once for personalization to move from a boutique offering to a default expectation, and they did.

The first was the collapse of trading frictions. Once commissions went to zero and fractional shares became routine, the mechanical objection to holding hundreds of individual positions in a modest account simply evaporated. You can now build a faithful replica of a broad index in an account of a few hundred thousand dollars without the transaction costs eating the benefit, which was not practical in a world of per-ticket commissions and round-lot constraints.

The second was the rise of tax awareness as a competitive battleground. As market beta became a commodity available for a basis point or two, the way a manager could still add measurable, defensible value was after tax. Loss harvesting, gain deferral, and lot-level discipline are among the few levers that reliably move an investor's spendable outcome without taking more market risk, and they only work if you hold the securities directly. Tax alpha became the differentiator precisely because everything else had been competed down to nothing.

The third was demographic and advisory. Advisors serving high-net-worth households found themselves competing on customization — values screens, concentrated-position management, legacy-holding transitions — and personalization is the natural answer to all of it. The client who walks in with a large low-basis position in a single stock cannot be served by a fund. They can be served by an account that is optimized around that exact problem.

The operational consequence nobody priced in

Here is where the strategy story collides with reality. When you sell personalization, you are not selling a portfolio — you are selling a promise to re-solve a portfolio, correctly and on time, for as long as the client stays. Every night, the universe moves, prices change, dividends land, cash arrives, restrictions get added, and lots age past their wash-sale windows. Each of those events can change the right answer for an account, which means the book has to be re-evaluated continuously, not occasionally.

A desk that was comfortable running a few hundred model-driven accounts discovers, somewhere between five thousand and fifty thousand accounts, that the thing which used to finish comfortably overnight no longer does. The batch that cleared by 4 a.m. starts spilling past the open. The optimizer that returned in two seconds on a thousand-name universe starts timing out, or worse, silently returning a lower-quality answer to hit the deadline. The cost of compute, which was a rounding error, becomes a line item that grows linearly with the book while fees per account are flat or falling. This is the scaling wall, and it does not announce itself politely. It arrives as a missed rebalance, a compliance question about why two identical accounts traded differently, and a quarter where operations headcount grew faster than assets.

The firms that treat this as an investment problem will keep hiring portfolio managers and analysts and wonder why margins keep compressing. The firms that recognize it as a manufacturing problem will invest in the one thing that actually relieves the constraint: throughput. The ability to price the entire book — every personalized, tax-aware account — before the market opens, deterministically, with an audit trail, and without a per-seat solver license metastasizing across the cluster, is not a nice-to-have at this scale. It is the difference between a business that can say yes to the next ten thousand accounts and one that has to say no.

Who actually wins

It is tempting to conclude that the winners will be whoever has the best investment models, and investment quality certainly matters. But in a market where the product is personalization at scale, the durable advantage accrues to whoever can industrialize it — who can take the same constrained, tax-aware optimization and run it across the whole book, every day, with results that are reproducible enough to satisfy a compliance reviewer and fast enough to never threaten the deadline. That is an infrastructure advantage, not a stock-picking one, and infrastructure advantages compound. The desk that solves throughput once can absorb growth that buries its competitors in operational cost.

None of this means the small, artisanal SMA desk disappears. It means the center of gravity moves toward firms that have made personalization a production line rather than a craft. The wave is real, the growth is real, and the operational bill is coming due. The honest question for any firm in this market is not whether personalization is the future — it plainly is — but whether the plumbing can carry the volume when the future arrives, which on current trajectories is sooner than the planning horizon most teams are using.