Direct Indexing Worth It? The $250K Crossover Point

The decision: Priya’s $250K rollover

Priya is a single software engineer in Austin, Texas, with $250,000 sitting in a taxable brokerage account after she sold a chunk of vested RSUs. Her financial advisor pitched a direct-indexing account at 0.40% all-in, versus the 0.03% Vanguard S&P 500 ETF (VOO) she could buy herself. Her taxable income is about $185,000 — inside the 24% federal bracket, which runs $103,351 to $197,300 for a single filer in 2026 — and Texas has no state income tax. The question she actually needs answered is not “which provider” but “is the higher fee worth it at all, at my size?”

Here is the resolution. The fee premium is 0.40% − 0.03% = 0.37%, or $925/year on her $250K. To justify that, her direct-indexing account has to harvest enough losses that the tax savings exceed $925. At her marginal long-term capital-gains rate of 15% (her income is under the $533,400 single threshold) plus the 3.8% NIIT (her MAGI is under $200K, so NIIT does not apply to her — more on that below), the losses save her 15 cents on the dollar. She needs $925 ÷ 0.15 = roughly $6,170 of incremental harvested losses per year to break even. A fresh $250K account in a normal market typically throws off $5,000–$15,000 of harvestable losses during the first few years. Priya clears the bar — but only because her advisor’s 0.40% fee is on the high side and her account is large enough to harvest meaningfully. At a 0.25% provider the math is far easier; at $75K it falls apart.

What direct indexing actually is

Direct indexing means owning the individual stocks that make up an index — say the 500 names in the S&P 500, or a representative sample of 150–250 of them — inside a separately managed account, instead of owning a single ETF or mutual fund that holds them for you. The point is not better returns. A direct-indexed S&P 500 account and VOO track the same benchmark and earn roughly the same pre-tax return. The point is tax-loss harvesting at the individual-stock level.

In any given year, even when the index is up, some constituent stocks are down. An ETF can only be harvested as a single lot — if VOO is up on the year, there is no loss to take. A direct-indexed account can sell the individual losers, book the capital losses, and immediately buy a not-substantially-identical replacement to stay invested. Those harvested losses offset capital gains elsewhere and up to $3,000 of ordinary income per year under IRC §1211(b), with the excess carried forward indefinitely.

The break-even formula

Strip away the marketing and the entire decision reduces to one inequality:

Direct indexing is worth it when tax value of losses > fee premium. The single most-quoted number from the spec is this: a 0.15%–0.35% fee premium needs roughly $2,700+ of incremental harvested losses to pay for itself once you value those losses at a meaningful rate. Here is where that comes from. Take a mid-range 0.25% premium on $250K = $625/year. At the top federal LTCG-plus-NIIT rate of 23.8%, break-even losses are $625 ÷ 0.238 = $2,626. At a combined 27.8% rate (the 23.8% federal plus a roughly 5% state offsetting short-term or ordinary income), it is $625 ÷ 0.278 = $2,248. Call it $2,700+ as the round-number bar a high-bracket investor needs to clear, and more if you can only value the losses at 15%.

Why the rate you value losses at matters

A harvested loss is only worth your marginal rate on whatever it offsets. The rates that apply, per the 2026 brackets:

• 15% LTCG — the rate for single filers with taxable income between $48,351 and $533,400 (MFJ $96,701–$600,050). Losses offsetting long-term gains here save 15 cents on the dollar.
• 23.8% top federal — the 20% LTCG rate (single income above $533,400) plus the 3.8% NIIT under IRC §1411, which kicks in on MAGI over $200,000 single / $250,000 MFJ. This is the effective top federal rate on investment income.
• 27.8%+ combined — add a state that taxes capital gains as ordinary income (most do). A 5% state on top of the 23.8% federal pushes past 27.8%. Losses are even more valuable when they offset short-term gains or up to $3,000 of ordinary income, which are taxed at ordinary rates — for Priya, the 24% bracket.

This is the lever most people miss: a loss that offsets $3,000 of ordinary income for a 24%-bracket filer is worth $720 of tax, far more than the same loss offsetting a 15% long-term gain ($450). The higher your bracket and the more short-term gains or ordinary income you can soak up, the lower your break-even and the more direct indexing pays.

Three worked tiers: $100K, $250K, $500K

Assume a 0.25% direct-indexing fee versus a 0.03% ETF (a 0.22% premium), a high-bracket investor valuing losses at the 23.8% top rate, and harvestable losses running at roughly 3% of the account in normal early years.

The $75K account the spec contrasts against tells the other side of the story: a 0.22% premium is only $165/year, but the harvestable losses are also small (often under $2,000), per-name minimums force you into far fewer stocks, and a single up year can leave almost nothing to harvest. Below roughly $100K–$150K, a 0.03% S&P 500 ETF plus occasional fund-level harvesting wins.

What most people get wrong about direct indexing

Three myths drive bad decisions here.

1. “Direct indexing beats the index.” It does not. It tracks the same index at roughly the same pre-tax return. The only edge is tax. If your account is in an IRA, 401(k), or Roth, there are no taxable gains to offset — direct indexing is pointless in a tax-advantaged account. It is a taxable-account-only strategy.
2. “Tax-loss harvesting is free money.” It is mostly a deferral, not a permanent saving. Harvesting a loss lowers your cost basis in the replacement security, so you owe more gain later when you sell. The real, permanent benefit is the rate arbitrage (offsetting ordinary income or short-term gains today, paying long-term rates later) plus the time value of deferring the tax. If you plan to liquidate the whole account in two years, most of the “savings” reverses.
3. “Bigger is always better.” The harvesting benefit fades. In year one almost every lot can be harvested. By year 5–10, most lots sit above cost basis — the embedded gains the spec flags — and there is little left to harvest. The fee premium keeps running at full freight while the harvest that justified it dwindles. This is why the honest pitch is “front-loaded tax alpha,” not “permanent outperformance.”

When embedded gains kill the case

The fastest way to destroy the math is to fund the account with already-appreciated stock. If Priya transferred in $250K of company shares with a $40K cost basis, the account starts with $210K of embedded gain and almost no losses to harvest — she is paying the fee premium for harvesting that cannot happen. Direct indexing works best funded with cash or near-basis assets, so the manager can build the position from scratch and harvest the natural dispersion as individual stocks move.

Even a cash-funded account ages. Tax alpha typically runs 0.5%–1.5% of the account in the first few years and decays toward 0.2%–0.5% as gains accumulate. Once your blended position is mostly green, run the break-even again: if the realistic harvest has dropped below your break-even-loss figure, the right move may be to stop paying the premium and hold the position (or transition to a low-fee ETF for new money). Reassess every few years; this is not set-and-forget.

The cheaper alternative below the crossover

If you are under about $150K, you do not need a separately managed account to harvest losses. You can do fund-level harvesting yourself: when your S&P 500 ETF is at a loss, sell it and buy a not substantially identical replacement — a total-market or large-cap-blend fund from a different provider tracking a different index — to avoid the §1091 wash-sale rule, then book the loss. It is less granular than stock-level harvesting (you can only harvest when the whole fund is down), but it costs nothing beyond a few minutes and keeps your expense ratio at 0.03%.

Your decision lever

Compute one number before you sign anything: (direct-indexing fee − 0.03%) × your account value ÷ your marginal rate on offsettable income. That is the dollar amount of losses the account must harvest every year to justify itself. If your taxable account is $250K or more, you are in a high bracket, you have other gains or $3,000 of ordinary income to offset, and you funded with cash — the harvest will almost certainly clear that bar and direct indexing is worth it. If you are under $150K, in a low bracket, holding the account inside an IRA, or funding with appreciated stock, skip it and run a 0.03% ETF with manual fund-level harvesting. The crossover is a number, not a feeling — for most investors it sits at $250K.