Do You Actually Need Direct Indexing?

Every quarter a new pitch lands in the inbox of high earners with seven-figure taxable accounts: direct indexing will unlock tax-loss harvesting that ETFs cannot match, lower your bill to the IRS, and pay for itself. The pitch is real. The loss-harvesting mechanics work. The strategies have academic backing from AQR and Parametric. But for most of the do-it-yourself investors who get the pitch, the underlying question is whether the incremental after-tax value clears the SMA fee, complexity, and lock-in.

There are three options on the table. The first is plain ETF tax-loss harvesting that you do yourself a few times a decade during drawdowns. The second is a long-only direct indexing SMA that owns hundreds of individual stocks to approximate an index and harvests losses on the losers. The third is a long/short direct indexing or tax-aware long-short (TALS) SMA that adds leverage and short positions to extend loss harvesting in both rising and falling markets. These options sit on a spectrum of cost, complexity, commitment, and the kind of investor they fit.

For most DIY investors, broad low-cost ETFs plus occasional tax-loss harvesting during downturns are hard to beat. Long-only direct indexing is often oversold because its biggest weakness is decay: after a few years in a rising market, harvestable losses fade while you continue paying the SMA fee. Long/short direct indexing has a more compelling theoretical case for high-net-worth investors with concentrated stock or a pending liquidity event, but it operates as a tax-aware active long-short factor SMA, with the leverage, shorting, borrow costs, and tracking error that come with active management.

What Direct Indexing Actually Is

Direct indexing is a separately managed account (SMA) that owns the individual stocks of a benchmark such as the S&P 500 or the US total market, rather than owning a pooled ETF that holds them for you. Direct ownership of the underlying securities enables two things: security-level tax-loss harvesting on individual names that fall below cost basis, and customization such as excluding employer stock, ESG screens, or factor tilts.

ETFs are already structurally tax-efficient. Vanguard describes the in-kind creation/redemption mechanism by which “ETF managers only need to create or redeem creation units. This means investors may have less exposure to capital gains.” Vanguard: What are ETFs? Capital-gains distributions are less common in ETFs than in mutual funds, though they still occur in some categories. The relevant question for a direct-indexing decision is whether the incremental tax savings from owning individual stocks justifies the added fees, tracking error, lot complexity, and exit cost compared to that already-efficient ETF baseline.

Three Ways to Harvest Losses, Compared

The choice is rarely “direct indexing yes or no.” It is which of three structures fits your situation.

This guide is the primer that compares the three. For the deep dive on TALS economics, the AQR research on the 30%-vs-100% cumulative loss-capacity gap, IRC §1259 constraints on shorting concentrated stock, and a hurdle calculator, see the companion guide on Tax-Aware Long-Short Investing.

The Decay Problem with Long-Only DI

Long-only direct indexing harvests the most losses early. When a portfolio is fresh, most tax lots are close to cost basis and even a modest market drawdown drops a fraction of names below their purchase price. As the portfolio appreciates, that fraction shrinks. Liberman, Krasner, Sosner, and Freitas (AQR, 2023) found that long-only direct indexing tax benefits taper: “On average, net losses realized by direct indexing loss-harvesting strategies taper off within the first few years after their inception” and reach “a maximum average cumulative level of about 30% of the initially invested capital.” AQR: Beyond Direct Indexing. After that ceiling, you keep paying the SMA fee but the portfolio mostly produces embedded gains, not harvestable losses.

New contributions, market volatility, and charitable giving of appreciated shares all refresh the harvesting engine. A long-only SMA with no contributions in a steadily rising market is the worst case for sustained tax benefits. AQR’s related work shows that “tax-aware 130/30 and 150/50 strategies seek to outperform an equity benchmark index on an after-tax basis, but they also have the potential to benefit both from successful factor tilts and from greater tax loss harvesting due to their long and short ‘extensions.’” AQR: Improving Direct Indexing. The decay visualizer further down lets you see this curve for your own portfolio and fee assumptions.

The “Stuck With It Forever” Problem

Starting direct indexing is easy. Unwinding it is not. Replacing a single S&P 500 ETF with a long-only SMA means owning the underlying constituents directly, often several hundred individual positions, each with its own cost basis, holding period, and wash-sale state. The customization that makes direct indexing attractive (factor tilts, employer-stock exclusion, ESG screens) compounds the lot-level complexity. A few years of active harvesting can produce thousands of tax lots from rotating in and out of replacement securities.

Exits from an SMA full of appreciated lots are limited. Selling out to move back to an ETF realizes the embedded gains, often clawing back much of the tax benefit harvested earlier. In-kind transfer to another advisor or custodian is usually possible, but cost basis travels with the shares and the new manager inherits the same lot inventory. The clean exits are donating appreciated shares to a donor-advised fund or charity, holding until death and getting the step-up in basis, or moving to a lower-tax state before liquidating. Without one of those exits, direct indexing functions as a long-term commitment, not a year-to-year product choice.

When Long-Only DI Actually Helps

Long-only direct indexing earns its fee when the harvested losses have somewhere productive to go. The fact patterns where the math holds up:

• Recurring realized capital gains from other sources. K-1s from private equity or hedge funds, distributions from a business, rebalancing across a multi-asset portfolio, or pass-through gains all create demand for losses to offset. Without gains to offset, harvested losses sit as carryforwards and pay you $3,000 against ordinary income per year per IRS Topic 409, which is a slow way to recover an SMA fee.
• Large ongoing taxable contributions. New cash creates fresh high-basis lots and refreshes the harvesting engine. A high-saver building taxable wealth at $50k-$200k per year gets more out of direct indexing than a buy-and-hold investor whose taxable account has long since been deployed.
• A regular charitable-giving program. Donating appreciated shares from the SMA to a donor-advised fund removes embedded gains from the portfolio and restores loss-harvesting capacity. Investors who give 5-figure amounts each year see direct indexing pay off more consistently because the cycle resets.
• Customization that an ETF cannot deliver. Investors who genuinely need to exclude employer stock, run an ESG overlay, or build a factor tilt may find that the SMA structure is the way to get those exposures cleanly, independent of the loss-harvesting case.
• A known liquidity event in 1-5 years. A business sale, IPO, RSU cliff, or property sale on the horizon creates future demand for losses. Building a loss bank ahead of the event can absorb a meaningful share of the eventual gain.

Outside these fact patterns, long-only direct indexing is probably oversold. A high salary alone is not one of the fact patterns.

Why High Income Alone Isn’t Enough

High W-2 income produces ordinary income, not capital gains. Harvested capital losses can offset capital gains without limit, but only $3,000 per year ($1,500 married filing separately) of net losses can be applied against ordinary income per IRS Topic 409. Unused losses carry forward indefinitely, but a carryforward is only valuable when realized gains eventually appear to consume it.

The high-earner archetype most exposed to the DI pitch (six figures of W-2 income, max retirement contributions, a seven-figure taxable account held in a few broad ETFs, no recurring private-investment gains) often already runs a very tax-efficient portfolio. ETFs absorb most of the tax drag through in-kind redemptions. The remaining opportunity is the occasional ETF tax-loss harvest during a market drawdown, which a DIY investor can do with two trades.

Author note. The pitch most direct-indexing salespeople bring to high earners undersells how much ETF tax-loss harvesting an attentive investor can do alone. A few wash-sale-safe ETF swaps during routine market drawdowns can capture meaningful losses with no SMA fee and no lot-level complexity. That baseline is what an SMA has to clear to add incremental value. The calculus shifts for an investor sitting on a low-basis concentrated position, planning a real estate exit, or expecting a business sale, where the demand for losses is larger and the SMA fee starts to look reasonable. Outside those fact patterns, the math tends to disappoint.

Long/Short DI Is a Different Animal

Long-only direct indexing harvests losers inside a long equity book. Long/short direct indexing, often marketed as tax-aware long-short or TALS, adds short positions to extend loss harvesting in both rising and falling markets. Parametric makes the mechanical case directly: these strategies “are designed to harvest losses in most environments, whether in a bull market where stocks are rising or a bear market where they are falling. Short positions potentially continue to generate losses even as the long portfolio appreciates.” Parametric: Long-Short Equity with Tax-Managed Portfolios. AQR’s research shows that with sufficient leverage, long-short strategies can realize cumulative net capital losses of up to 100% of invested capital within a few years, compared to the long-only ~30% ceiling.

That extension is the value proposition. The added cost is complexity and active risk. Long-short DI typically runs at 130/30, 150/50, or higher leverage, uses multifactor signals (quality, momentum, value, low-volatility) to decide what to overweight, underweight, and short, and operates as a tax-aware active long-short factor SMA. Parametric explicitly notes the added cost: “Long-short construction introduces higher explicit costs, greater operational complexity and increased turnover, along with another dimension of risk: active factor exposures designed to produce pre-tax alpha.”

Elm Wealth’s skeptical analysis runs the math on a representative investor and concludes that the manager fees often consume a meaningful share of the tax benefit, with favorable economics requiring “relocating from a high-tax state like California to Florida, expecting step-up in basis within ten years, or having a manager capable of generating ‘significant alpha’ exceeding 0.5% annually” Elm Wealth: Robbing Peter to Pay Paul. That is a narrow set of fact patterns. The companion guide on Tax-Aware Long-Short Investing covers the IRC §1259 constraint on shorting concentrated stock, the Swedroe pre-tax test, and a hurdle calculator for evaluating specific strategies.

Risks Across All DI Variants

• Fees can consume the tax benefit. The break-even harvest rate the SMA must clear annually scales inversely with the marginal tax rate on usable losses. At a 35% blended rate and a 40 bps fee, the strategy needs to harvest losses worth about 1.14% of portfolio each year just to break even. The decay visualizer below shows when the harvest curve crosses below that line.
• Tax deferral is not tax alpha. Harvested losses lower the basis of the replacement securities, so today’s loss usually becomes tomorrow’s larger gain unless you donate, get a step-up at death, or move to a lower-tax state. Permanent tax savings require an exit plan; otherwise the strategy delivers timing, not elimination.
• Wash-sale rules apply to short sales. IRS Publication 550 specifically addresses wash-sale treatment on losses from short sales IRS Pub 550. Long/short DI strategies must manage wash sales on both books.
• Tracking error against the benchmark. Replacement securities and active factor extensions diverge from the stated index. Long-only DI tracking error is usually modest; long/short DI tracking error is material.
• Manager and platform risk. You move from owning fund shares to holding an SMA. Performance, execution, tax reporting, and service quality become manager-specific. Switching providers means moving a large inventory of individual tax lots.
• Long/short complexity adds borrow costs, leverage, and §1259 limits. The companion TALS guide covers these in detail. The short version is that adding shorts adds risks and costs that long-only direct indexing does not face.

Who Should Use Each, and Who Shouldn’t

Try It: The DI Decay Visualizer

The calculator below models a long-only direct indexing portfolio over the horizon you specify. The blue line shows the expected harvestable losses each year as a percentage of the portfolio. The dashed line is the break-even harvest rate the SMA must clear to cover its fee at your marginal tax rate. The decay curve starts above the break-even line and ends below it; the verdict tile reports whether cumulative tax savings clear cumulative fees over the full horizon. Move the contributions, volatility, fee, and tax rate sliders to see how sensitive the conclusion is to your inputs.

Frequently Asked Questions

Is direct indexing worth it if I have a $1M+ taxable account?

Having a large taxable account is necessary, not sufficient. The strategy pays off when harvested losses are absorbed by usable realized gains and the resulting tax savings clear the SMA fee. A $1M taxable account held in broad ETFs with no recurring realized gains and no charitable giving is already very tax-efficient. Run the decay visualizer with your tax rate, contribution rate, and SMA fee to see whether the math works for your specific situation.

How is direct indexing different from a regular ETF?

A direct-indexing SMA owns the individual stocks of a benchmark directly. An ETF owns those same stocks but you own the fund shares, not the underlying. The SMA structure lets the manager realize losses on individual names that fall below cost basis even when the broader index is up. ETFs achieve tax efficiency through in-kind creation and redemption that minimizes capital-gains distributions at the fund level, but you cannot harvest losses on individual constituents.

Will I save more in taxes than I pay in fees?

Often no, especially for long-only direct indexing in a rising market with no contributions or charitable giving. AQR’s research finds the cumulative loss capacity of long-only direct indexing caps around 30% of invested capital and the harvest engine tapers within the first few years. Plug your numbers into the decay visualizer to see whether the cumulative tax savings clear the cumulative SMA fees over your horizon.

What happens when I want to leave a direct-indexing SMA?

You inherit a portfolio of hundreds (or thousands) of individual tax lots. Selling out to move back to an ETF realizes the embedded gains, often clawing back much of the tax benefit harvested earlier. In-kind transfer to another custodian is possible but the lot complexity travels with the shares. Clean exits are donating appreciated shares, holding until death for the step-up in basis, or moving to a lower-tax state before liquidating.

Should I do long-only direct indexing or long/short (TALS)?

Long-only DI is the simpler answer for HNW investors with recurring realized gains or ongoing contributions. Long/short DI has a more compelling theoretical case for investors with concentrated stock or a pending liquidity event, because the short book extends loss generation. But long/short DI adds leverage, shorting, borrow costs, and active factor risk, and is properly evaluated as an active SMA. See the TALS guide for the deep dive and a hurdle calculator.

Does direct indexing make sense in a Roth IRA, 401(k), or HSA?

No. Tax-advantaged accounts already shelter capital gains from tax, so security-level loss harvesting produces no benefit. Direct indexing only makes sense in taxable accounts. Use simple low-cost index funds inside tax-advantaged wrappers.

Related Guides

• Tax-Aware Long-Short Investing is the deep dive on the long/short variant: AQR’s 30%-vs-100% loss-capacity gap, gain deferral, IRC §1259 constraints, the Swedroe pre-tax test, and a hurdle calculator.
• Tax-Loss Harvesting covers the long-only mechanics most DIY investors actually need: wash-sale rule, the $3,000 ordinary-income offset, replacement-fund mechanics, and a savings calculator.
• Concentration Risk covers the math of holding a single position. Often the reason investors look at direct indexing in the first place.
• Equity Compensation and Sell Your RSUs at Vest handle the simpler upstream question for employees with large RSU positions.

Key Takeaways

• Most DIY investors don’t need direct indexing. Broad low-cost ETFs are already tax-efficient. A few ETF tax-loss harvesting trades during drawdowns capture most of the available tax value without SMA fees or lot-level complexity.
• Long-only direct indexing decays. AQR finds cumulative loss capacity caps near 30% of invested capital and tapers within the first few years. Without fresh contributions, charitable giving, or recurring realized gains, the harvest engine stops paying for the fee.
• The strongest standalone case is concentrated stock or a pending liquidity event. A business sale, RSU cliff, low-basis concentrated position, or property sale creates real demand for losses. Outside those fact patterns, direct indexing is often oversold.
• Long/short direct indexing operates as a tax-aware active SMA. The short book extends loss generation, but the strategy adds leverage, factor bets, borrow costs, and tracking error. Evaluate it as you would any active long-short factor product.
• Permanent tax savings require a non-sale exit. Harvested losses lower the basis of replacement securities, so the strategy delivers tax deferral, not elimination, unless you donate appreciated shares, hold until death for the step-up, or move to a lower-tax state before liquidating.

Sources

1. AQR (Liberman, Krasner, Sosner, Freitas), “Beyond Direct Indexing: Dynamic Direct Long-Short Investing” (2023). aqr.com
2. AQR, “Improving Direct Indexing: 130/30 and 150/50 Strategies” (Jul 2021). aqr.com
3. Parametric, “Long-Short Equity Strategy: Tax-Managed Portfolios for the Right Investor.” parametricportfolio.com
4. Elm Wealth, “Robbing Peter to Pay Paul: A(nother) Look at Long/Short Direct Index Tax-Loss Harvesting” (Apr 2026). elmwealth.com
5. Vanguard, “What are exchange traded funds (ETFs)?” investor.vanguard.com
6. IRS, Topic 409 “Capital Gains and Losses.” irs.gov
7. IRS, Publication 550 “Investment Income and Expenses.” irs.gov