Tax-Loss Harvesting in Direct Indexing Accounts

If you're already using tax-loss harvesting inside a standard index ETF and wondering why the results feel modest, the answer is structural. One fund equals one lot. Direct indexing gives you hundreds.

Why Individual Stock Ownership Changes the Harvest Math

An S&P 500 ETF moves as a single unit. On any given day the fund is either up or down, and you either have a harvestable loss or you don't. Own the 500 underlying stocks individually and the picture looks completely different. Even on a flat or positive day for the index, a meaningful number of individual names are down. Apple might be up 1.2% while a handful of energy or consumer discretionary names are off 3 or 4%. Each one of those is a harvest candidate.

This is the core proposition of direct indexing: disaggregating an index into its components creates a near-constant stream of harvesting opportunities that a pooled fund simply cannot generate. Advisors sometimes call this "dispersion alpha," which is just a way of saying that volatility in individual names, even inside a broadly rising index, is your raw material.

The Rules and Tradeoffs You Need to Understand

The fee spread is real and you have to earn it back. A plain Vanguard or iShares S&P 500 ETF runs around 0.03% annually. Direct indexing platforms, depending on the provider and account size, typically charge 0.20% to 0.40%. On a $1 million account that gap is $1,700 to $3,700 per year. The harvesting has to generate enough tax savings, net of that fee drag, to come out ahead. At smaller account sizes it often doesn't.

The break-even threshold most practitioners use is somewhere around $500,000 in taxable assets. Below that, the fee cost usually exceeds the realistic harvest yield. Above $500,000, and especially in the $1 million to $5 million range where high earners are sitting in the 20% long-term capital gains bracket plus the 3.8% net investment income tax, the math can improve meaningfully.

The wash-sale rule applies at the individual security level, and that's actually the advantage. When you sell Apple at a loss, you can't buy Apple back within 30 days. But you can immediately buy a highly correlated substitute, say, Microsoft or a technology sector ETF, and maintain essentially the same market exposure while the loss books. This is the technique that makes direct indexing work. You stay invested, you preserve your factor exposure, and you capture the loss.

IRS Notice 89-19 and factor exposure risk. The IRS has long been attentive to arrangements that purport to take losses while retaining economic equivalence to the sold position. Notice 89-19 is the foundational guidance, and while it was originally aimed at straddle-like structures, practitioners treat it as the baseline for thinking about how "substantially identical" gets interpreted. In practice, swapping one large-cap growth stock for another in the same sector clears the bar comfortably. What creates risk is trying to maintain hyper-concentrated factor tilts with only two or three substitute names. A thoughtfully constructed direct indexing strategy keeps substitutions diversified enough that no single swap looks economically identical to what was sold.

Tracking error is a real cost. The more aggressively you harvest, the more your portfolio diverges from the index. Over time, substitutes accumulate and the portfolio can start to look less like the S&P 500 and more like a custom basket with embedded embedded gains, low-basis positions, and factor tilts you didn't specifically choose. Managing this drift is part of what a good direct indexing custodian, or advisor, has to do actively.

When Direct Indexing Pays Off vs. When It Doesn't

I had a client, a software executive in his late 30s, who came in with roughly $1.4 million in taxable savings and a marginal ordinary income rate above 37%. He was already holding broad index ETFs. In the first full calendar year after transitioning to a direct indexing account, the platform harvested losses that offset a substantial portion of his annual compensation income, not just capital gains. The fee premium paid for itself several times over. (This is an illustrative example, not a guarantee of any specific result.)

Direct indexing is worth evaluating when you have:

• A taxable account above $500,000 that you intend to keep invested for at least five to ten years
• A meaningful ordinary income or capital gains tax rate (the higher your rate, the more a harvested loss is worth)
• Expected capital gains events in the near term, whether from a business sale, concentrated stock position, or real estate transaction, that you'd like to offset
• A long runway before you need to liquidate, because the low-basis positions that accumulate need time or a step-up-at-death strategy to resolve cleanly

It's less compelling if your taxable account is below $300,000, if you're in a low bracket, if you anticipate moving to a no-income-tax state where your rate picture shifts, or if you expect to need liquidity within a few years and will have to sell regardless of basis.

How This Connects to the Tax-Loss Harvesting Pillar

Direct indexing is a tool for maximizing the raw quantity and dollar amount of losses you can harvest in a given year. How you deploy those losses, whether to offset capital gains, reduce ordinary income up to the $3,000 annual limit, or pair with a Roth conversion to neutralize the income it generates, is the broader strategy question. That broader strategy is what the parent guide covers. If you haven't read "Tax-Loss Harvesting: How to Turn a Down Market Into Real Tax Savings", start there for the foundational mechanics before thinking about whether direct indexing makes sense as an implementation vehicle.

The way I think about it: direct indexing sits in the Soil layer of a plan, meaning it's structural tax architecture, not a one-time move. It affects every year's return, every year's tax bill, and ultimately the long-term compounding rate on after-tax wealth.

Frequently Asked Questions

What account size actually makes direct indexing worthwhile?

The practical floor is around $500,000 in a single taxable account. Below that, the fee premium of 0.20% to 0.40% annually tends to exceed the realistic harvest yield. Above $1 million the math improves considerably, especially for high earners in the top brackets.

Does direct indexing work inside an IRA or 401(k)?

No. Tax-loss harvesting only creates value in taxable accounts. Losses inside tax-deferred or tax-free accounts have no tax consequence, so the entire premise of direct indexing doesn't apply there.

Will my portfolio look completely different from the S&P 500?

Not significantly in the early years. Over time, substitutes accumulate and tracking error grows, but a well-managed account stays correlated enough to the index for most practical purposes. The bigger issue is managing the low-basis positions that build up as years pass.

How does direct indexing interact with a Roth conversion strategy?

Well, when the two are coordinated. A Roth conversion adds to your ordinary income in the year it's done. Harvested losses can offset up to $3,000 of that against ordinary income directly, and losses that offset capital gains free up bracket room. This is one of the strongest reasons to consider direct indexing alongside a multi-year conversion plan.

Is there a risk the IRS disallows my harvested losses?

Not if the substitutions are handled properly. The wash-sale rule applies to substantially identical securities. Swapping individual stocks within the same sector for reasonably diversified substitutes, and following the 30-day window discipline, is well-established practice. Aggressive single-stock swaps or arrangements that look economically identical raise the flag; a broadly diversified approach does not.

Do I need a special custodian or can I do this at a major brokerage?

Most major custodians don't offer automated direct indexing natively. The dedicated platforms (Parametric, Aperio, Canvas, and others offered through major asset managers) manage the harvest automation, rebalancing, and substitute selection. Your advisor needs to either use one of those platforms or build and manage the individual stock portfolio manually, which is operationally complex at scale.

What to Do Next

1. Pull your last two years of Schedule D and tally the capital gains you actually paid tax on. That number tells you how much harvested loss would have been worth in real dollars.
2. Confirm your taxable account balance. If it's below $300,000, direct indexing likely doesn't clear the fee hurdle yet; revisit when it does.
3. If you're above $500,000, ask a fee-only advisor to model the after-fee, after-tax return improvement estimate for your specific bracket, account size, and time horizon before committing to any platform.
4. Consider whether you have any anticipated gain events, business sale, RSU vesting, real estate, in the next two to three years that would benefit from a pre-built loss bank. Timing the transition to direct indexing before those events matters.

The information provided is for educational purposes only and does not constitute investment, legal, or tax advice. Tax law changes frequently — verify current rules before acting. Consult with qualified professionals for guidance specific to your situation.