Direct Indexing Against a Concentrated Stock Position

Direct indexing holds the individual components of an index in a separately managed account instead of a fund. The goal is to match index returns within a tight tracking error while realizing losses on individual stocks that diverge from the average. These realized losses offset capital gains elsewhere, typically from selling down a concentrated single-stock position. In a portfolio with $3M of direct-indexed assets alongside a $2M-per-year concentrated sell-down, direct indexing typically generates $60-90k of harvested losses per year, saving the holder roughly $15-25k per year in federal plus state tax.

For a tech employee with a large embedded gain in their employer’s stock, this changes the math on diversification. A straight sell-and-diversify approach taxes the entire gain at 23.8% federal plus state. Paired with direct indexing, a meaningful chunk of the gain is sheltered by harvested losses, improving the after-tax outcome. Over a 5-10 year deconcentration period, the compounded savings can exceed $150-250k.

This guide covers how direct indexing works mechanically, the tax-loss-harvesting math, the wash-sale rule implications, the typical fee structure compared to ETFs, and when the approach makes sense versus when it does not.

What direct indexing actually does

A direct-indexed portfolio replicates an index (commonly S&P 500, Russell 3000, or a custom factor-tilted index) by holding most of the underlying stocks directly. A $3M S&P 500 direct-index account might hold 300-400 of the index’s 500 constituents, weighted to track the index within roughly 0.5% tracking error.

The account is held in an SMA (separately managed account) at a custodian like Fidelity, Schwab, or Pershing. A direct-indexing provider (Parametric, Aperio, Frec, Vise, and others) manages the ongoing trading, rebalancing, and harvesting.

On a typical day, some stocks in the index are up, some are down. If a stock you hold is down, say, 10% or more from your basis, the algorithm sells it (harvesting the loss) and immediately buys a similar stock (a competitor in the same sector, a stock with similar factor exposure) to maintain the portfolio’s index-tracking characteristics. The loss is realized for tax purposes; the economic exposure to the index is preserved.

Over a year, the algorithm harvests losses continuously. In up markets, the number of loss-candidate stocks is smaller but still meaningful because of natural dispersion. In down markets, the harvest potential is larger.

Typical loss-harvesting yields

Historical backtesting from Parametric, Aperio, and other direct-indexing providers shows average annual harvested losses of 1-4% of portfolio value, varying by market conditions and portfolio age.

Harvesting yields decline as the portfolio ages because the unharvested gains accumulate (every stock that has appreciated since purchase has less harvesting potential). After 5-7 years, most portfolios have harvested through the majority of their early loss potential.

For a $3M portfolio funded with cash (not transferred stock), typical first-year harvested losses are $60-100k. These losses offset capital gains from any source on the tax return. For a holder selling $2M of concentrated stock with $1.8M of embedded gain, $80k of harvested losses reduces taxable gain to $1.72M, saving $19k at 23.8% federal plus roughly $10k at California rates.

The wash-sale rule

IRC §1091 disallows losses on the sale of securities if substantially identical securities are acquired within 30 days before or after. The rule is central to direct indexing, because the whole point is to sell a losing stock and replace it with a similar one without triggering wash-sale disallowance.

“Substantially identical” is narrow. A stock and an ETF holding that stock are not substantially identical for §1091 purposes. Two stocks in the same sector are not substantially identical. ETF X and ETF Y tracking similar indexes are usually not substantially identical (though tracking the exact same index could be challenged).

Direct indexing providers avoid wash sales by maintaining sector and factor exposure through a different stock. Selling a losing biotech and buying a different biotech in the same sector keeps the portfolio roughly flat in that exposure without triggering §1091.

Two common wash-sale traps:

1. Buying the concentrated-stock’s replacement elsewhere. If you sell employer stock, you cannot buy employer stock within 30 days before or after in any taxable or retirement account (wash-sale rules apply across accounts).

2. Direct-indexed account holding the same stock you sold. If the index includes your employer’s stock, the direct-indexing provider must exclude it from the account to avoid wash-sale conflicts with your concentrated sales. Most providers offer “restriction lists” at account setup.

The restriction list is essential for employees running direct indexing alongside a concentrated sell-down. Block your employer from the account entirely.

Fees and the comparison to ETFs

Direct indexing fees are typically 0.15-0.40% of AUM at the larger providers, sometimes lower for newer platforms targeting retail. SMA custodial fees are usually absorbed in the management fee. Underlying trading costs are minimal because most brokerages offer commission-free individual stock trades.

For comparison, holding the S&P 500 through a low-cost ETF (VOO, IVV) costs 0.03% per year. The incremental cost of direct indexing over the ETF is 0.12-0.37%.

On a $3M portfolio, the incremental fee is $3,600-11,100 per year. The incremental tax savings from harvested losses, conservatively $60-80k per year of realized losses offsetting gains at 23.8% federal plus state, is $18-25k. Net benefit of direct indexing: $7-20k per year on a $3M portfolio.

The benefit scales with harvestable losses and with the presence of offsetting gains. For a holder with no concentrated position to sell, harvested losses can still offset $3k of ordinary income per year (the §1211 limit) and carry forward against future gains, but the current-year benefit is much smaller.

When direct indexing makes sense

Best fit scenarios:

• Concentrated-stock holder executing a multi-year deconcentration plan. Harvested losses offset the gains from sales.
• High-income household with significant realized gains elsewhere (private-market exits, real estate sales). Losses shelter those gains.
• Long-term holder who values tax efficiency and is comfortable with complexity. Harvested losses compound into after-tax return advantages over time.
• Charitably minded holders. Concentrated individual stock positions can eventually be donated to a DAF, which avoids triggering the embedded gain altogether. Direct indexing’s stock-by-stock structure enables this.

Worse fit scenarios:

• Small portfolios under $250k. Fixed costs and minimum account sizes make direct indexing uneconomical.
• Tax-advantaged accounts (IRA, 401(k)). Loss harvesting has no tax benefit inside these accounts.
• Holders with no realized gains to offset. Losses carry forward but provide no current-year benefit beyond $3k.
• Holders who will never sell. If the position is held until death (basis step-up under §1014), the harvested-loss advantage is partially wasted.

Alpha from direct indexing beyond tax

Tax efficiency is the primary case. Secondary benefits include:

• Customization. Excluded stocks (ESG screens, employer stock, industries the holder wants to avoid) can be omitted without breaking index tracking.
• Incremental alpha through factor tilts (value, momentum, quality) at low cost, if the provider offers such overlays.
• Donation optimization. Individual lots with the highest embedded gains can be identified and donated directly to charity, maximizing the deduction and minimizing the retained position’s basis erosion.

For most retail users, the tax-loss harvesting is the dominant benefit and the others are secondary.

Frequently asked

Is this just “tax loss harvesting” that robo-advisors already do? Partially. Wealthfront, Betterment, and Schwab Intelligent Portfolios all do basic tax-loss harvesting within a 3-5 ETF portfolio. Direct indexing takes it further by holding individual stocks rather than ETFs, which increases the surface area for harvesting dramatically. A 500-stock S&P 500 direct-index portfolio has far more dispersion to work with than a 4-ETF robo-advisor portfolio.

Will the IRS challenge direct indexing? The practice is well-established and widely used. As long as the replacement stocks are not substantially identical to the sold stocks, the wash-sale rule is respected. The IRS has not issued guidance challenging direct indexing, and tax court rulings have generally accepted the approach.

Can I do this myself without a provider? Technically yes, but holding 300+ individual stocks and running the harvesting algorithm is beyond most retail capacity. Even at the DIY level, the operational cost (time, spreadsheet management, rebalancing) is meaningful. Providers exist because they execute the math daily at scale.

What happens to the tax benefit when I eventually sell? Harvested losses reduce your current-year taxable gain. When you eventually sell the direct-indexed portfolio, you owe capital gains on the appreciation relative to the post-harvested basis. This is a deferral, not an elimination; the benefit is the time-value of tax deferred plus any rate arbitrage between current and future brackets.

How long before direct indexing is “tapped out”? Usually 5-7 years for a portfolio funded with cash. After that, most stocks have appreciated and the harvestable loss pool is smaller. Adding new cash or transferring in new positions extends the runway. Some providers offer “re-set” options that restart the portfolio at current prices to reopen the harvest window, but this triggers gain recognition on the positions being sold out.

Next step

Pair direct indexing with a deliberate deconcentration plan to maximize the benefit. The direct indexing calculator models expected loss harvests and after-tax savings across portfolio sizes and time horizons. For portfolios above $500k of investable cash alongside a concentrated position above $1M, direct indexing usually pencils out; below those sizes, a low-cost ETF plus manual TLH is typically sufficient.