Escalent’s latest Advisor Use of Model Portfolios and SMAs study found that 42% of advisors who use model portfolios increased that usage over the past two years, up from 29% in 2023, while the share of advisors who call themselves “technical” fell from 43% to 36%. Read together, those two numbers describe a single move. Advisors are handing the index off to a model and reclaiming the hours they used to spend picking funds. The survey, fielded among roughly 400 registered advisors in September and October 2025, frames it as a role shift from stock picker to relationship manager.
The more interesting half of the story is what advisors are doing with the part of the portfolio they refuse to commoditize. Direct indexing and SMAs let them outsource the beta and keep the tax management in-house, where it produces a number a packaged fund cannot. That is the pivot: cheap, delegated market exposure on one side, and customized, tax-aware ownership of individual securities on the other.
Key Takeaways
- Escalent found 42% of model-portfolio users increased usage over two years, up from 29% in 2023, as “technical” advisors fell from 43% to 36%.
- Average SMA allocations among advisors serving high-net-worth clients are projected to climb from 23% in 2023 to 31% in 2025.
- Cerulli projects direct indexing assets reach roughly $825 billion by the end of 2026, a 12.3% five-year growth rate that outpaces ETFs, mutual funds, and traditional SMAs.
- Tax-managed SMA assets passed $500 billion in 2024, up 67% from 2022, with entry minimums now as low as $5,000 and fees of 0.20% to 0.40%.
- A direct-indexed SMA and an all-ETF model can hold the same exposure, but only the SMA harvests losses at the single-stock level, the source of tax alpha a packaged fund structurally cannot deliver.
What the Escalent Data Actually Shows
The headline reads like a story about model portfolios winning, and partly it is. Escalent’s annual study tracks how advisors split their books across models, separately managed accounts, and direct indexing, and model usage keeps climbing. But the raw growth in model allocations is maturing, even as adoption broadens.
The sharper signal sits in the advisor’s job description. Escalent defines a “technical” advisor as one who spends at least 40% of working time on investment selection and portfolio construction. That group shrank from 43% of the field in 2023 to 36% in 2025. Meredith Lloyd Rice, a vice president at Escalent, put the tension plainly: advisors are reevaluating whether model portfolios offer the sophistication their clients now demand.
So the average advisor is doing less hand construction of core portfolios and more of everything else: planning, tax, client acquisition, and behavioral coaching. A model handles the strategic allocation. The advisor keeps the relationship. That trade has been building for years, and the 2025 reading shows it accelerating rather than leveling off.
Why Are Advisors Handing Off the Index?
Three forces push in the same direction. The first is economic. An advisor who delegates core allocation to a model from BlackRock, Vanguard, Fidelity, or a strategist partner frees up time that converts directly into new client capacity. LPL Research’s 2026 strategic asset allocation framework, which rightsizes equity risk and anchors in high-quality fixed income, is the kind of off-the-shelf input an advisor can adopt without rebuilding a portfolio from scratch.
The second is competitive. Clients no longer pay a premium for beta they can buy in an index fund for a handful of basis points. The fee a client will defend is the one attached to advice, tax management, and coordination, not security selection. Outsourcing the index protects the advisor’s pricing power by moving it to where the value is visible.
The third is operational. Models scale. An advisor running 200 households on three or four model sleeves can rebalance, document, and supervise far more cleanly than one running 200 bespoke portfolios. The compliance and consistency benefits alone explain a meaningful share of the adoption curve Escalent measures.
Where Does Direct Indexing Fit?
Here is where the pivot turns from delegation into customization. Cerulli Associates projects direct indexing assets reach roughly $825 billion by the end of 2026, growing at a 12.3% annual rate that outpaces ETFs, mutual funds, and traditional separate accounts. Direct indexing started from about $462 billion at the end of 2022, so the category is on track to nearly double inside four years. Cerulli also expects direct indexing to make up about a third of the retail separate-account market.
The broader tax-managed SMA market tells the same story. Morningstar data shows assets in tax-managed SMAs passed $500 billion in 2024, up 67% from 2022. What changed is access. Schwab, Fidelity, BlackRock, and Vanguard have pushed entry minimums down to a range of $5,000 to $100,000, with annual fees of 0.20% to 0.40%. A strategy that once belonged to ultra-high-net-worth households now reaches the mass-affluent client an advisor sees every week.
A direct-indexed account holds the individual stocks of an index inside an SMA wrapper. The client owns the names, not a fund. That ownership is the whole point, because it unlocks tax-loss harvesting at the single-security level, plus customization for factor tilts, values screens, or a concentrated legacy position the client wants to diversify around without a tax event. The same machinery that builds private-market model portfolios for advisors is now being pointed at the taxable core of ordinary portfolios.
The concentrated-position case deserves its own mention, because it is where direct indexing earns its keep with the wealthiest clients. A founder sitting on a single low-basis stock can build a direct-indexed completion portfolio around it, harvesting losses elsewhere in the account to fund a gradual, tax-aware unwind of the concentrated holding over several years. That is a problem a packaged fund cannot touch, and it sits in the same tax-efficiency conversation as the ETF share-class conversions reshaping fund wrappers this year. The wrapper is doing the work that security selection used to.
The Tax Alpha a Model Portfolio Cannot Capture
This is the part worth modeling, because it is the reason the SMA side of the barbell exists. Take a client with a $2 million taxable account. Run that money through an all-ETF model portfolio and the only losses available to harvest sit at the fund level. In a year where the index finishes up, the ETF shows a gain, and there is nothing to harvest.
Now hold the same exposure as a direct-indexed SMA of 250 individual stocks. Even in an up year, a meaningful slice of those names trades below cost at some point. Research from Vanguard and Parametric puts the after-tax benefit of systematic single-stock harvesting at roughly 1% to 2% a year in the early years for high-tax investors, decaying as the portfolio’s embedded gains build. Apply a conservative 1.5% to the $2 million account and that is about $30,000 of after-tax value a year, realized as harvested losses that offset gains elsewhere in the client’s return.
The model portfolio and the SMA can track the same benchmark to within a few tenths of a percent. The difference is not the exposure. It is that one wrapper lets the client bank losses at the name level and the other does not. For a household selling a business, exercising options, or running a parallel hedge-fund sleeve that throws off short-term gains, that harvested-loss inventory is the asset. This is the calculation, current as of June 2026, that explains why advisors will delegate the index and refuse to delegate the tax management sitting on top of it.
What Could Go Wrong With the Barbell?
The structure has real failure modes, and an advisor who sells it as free money is setting up a disappointment. Tax alpha is front-loaded and finite. Once a direct-indexed account has harvested its early losses and built large embedded gains, the harvesting engine slows and the account becomes harder, not easier, to adjust. The benefit was never permanent.
Wash-sale rules add friction. Harvesting a loss and buying a near-identical security inside 30 days disallows the loss, so the manager has to swap into a correlated but distinct name, which introduces tracking error against the benchmark. Stack too many customizations, a values screen here, a factor tilt there, a legacy position carved out, and the SMA can drift far enough from its index that the client is taking active risk no one underwrote.
Fee stacking is the quiet one. A model-portfolio fee, an SMA management fee, the advisor’s own fee, and an overlay platform charge can compound into a number that erodes the very tax alpha the structure was built to capture. The discipline is to make sure the harvested value clears the all-in cost. When advisors layer defined-maturity bond ETFs and other building blocks into the same accounts, or reach for interval funds and other less-liquid sleeves, the cost and complexity math gets harder to track, not easier.
Three Questions to Bring to the Next Investment Committee
The pivot rewards advisors who can say precisely where they add value and price it accordingly. Three questions sharpen the decision before the next reallocation.
1. Which clients actually generate enough tax alpha to justify a direct-indexed SMA over a cheaper model? The benefit scales with the client’s tax rate, the size of the taxable account, and the presence of outside gains to offset. A client in a low bracket with a tax-deferred-heavy balance sheet may get more from a low-cost model than from a harvesting engine. Segment the book before defaulting everyone into the same wrapper.
2. What is the all-in cost of the barbell, and does the measured tax alpha clear it? Add the model fee, the SMA fee, the platform overlay, and the advisory fee, then compare the total against the realized harvested value, not the marketed estimate. If the stack eats the alpha, the structure is complexity for its own sake.
3. How much tracking error has customization introduced, and did the client agree to it? Every screen, tilt, and carve-out moves the account away from its benchmark. Quantify the drift, document that the client signed up for it, and revisit it as embedded gains lock the portfolio in place over time.
About Me
Tradingmarketsignals serves as the definitive digital ecosystem for the modern wealth management community. As a premier source of intelligence, it delivers high-level analysis, regulatory updates, and technological insights tailored specifically for independent financial advisors, RIA leaders, and investment professionals.
