Key takeaways
- Tax-loss harvesting opportunities fade away. Research found about 0.69% per year of tax benefit in a direct indexing portfolio’s first five years, then considerably less in the next five years.
- A mature direct indexing portfolio holds hundreds of positions selected somewhat arbitrarily, concentrates toward its winners, misses new index additions, and keeps charging fees.
- Individual stock research helps make unwinding it a real investment decision, not a highly-taxed liquidation. The goal is to keep the businesses worth owning and sell the rest on a schedule.
Direct indexing began with an appealing pitch: own the stocks in an index directly instead of through a fund, let sophisticated algorithms harvest tax losses all year, and customize the portfolio to avoid concentrated risk related to equity compensation. For the first few years, harvesting works and delivers the benefit the industry calls tax alpha. Then it dwindles away.
If your account now holds hundreds of stocks, produces no new tax losses, and still charges the fee, you might think your portfolio is broken. But the strategy completed what it was built to do.
This article explains why the losses dry up, what the portfolio becomes afterward, and realistic ways to unwind direct indexing without paying more tax than necessary.
Why the tax losses run out
Tax-loss harvesting sells positions that trade below cost and immediately buys similar stocks to keep you invested. Each round books a loss but resets cost basis lower. When markets rise, on average positions climb above their new basis and remain there. Without fresh deposits, the algorithm eventually finds nothing to sell at a loss.
Research by four AQR authors in the Journal of Beta Investment Strategies measures the fade. Harvesting added about 0.69% per year in tax benefit over a portfolio’s first five years, then about 0.09% per year in years six through ten. Cumulative harvested losses reached about 13% of invested capital in year one, crossed 20% near year three, and leveled off around 30%. Turnover fell from 155% per year to 55% as opportunities dried up.
Providers describe these portfolios as “mature.” A drawdown like 2022 might revive harvesting for a while, and steady new contributions could extend it. But a lump sum investment during a rising market simply exhausts the benefits.
What a mature portfolio looks like
Every harvest swaps one company for a correlated one. The strategy could buy the original back once the wash sale window passes, but the replacement often remains. Repeat that process dozens of times and you end up owning a collection of companies selected somewhat arbitrarily. Weightings drift as well, since trimming winners would realize gains. The portfolio gradually concentrates in whatever ran the most, and companies that joined the index in later years never make it in because buying requires fresh cash. Year by year it tracks the index less faithfully while still charging the fee.
Consider a hypothetical investor who moved $1 million into an S&P 500 direct indexing account in January 2021. The index returned about 96% over the next five calendar years with dividends reinvested, and the account would have grown toward $2 million. Direct indexing strategies likely harvested aggressively through the 2022 downturn, and passed through those tax losses. Since then, the investor’s harvest activity would show next to nothing. They’d hold roughly 350 positions across well over a thousand tax lots, almost all above basis. Large-cap direct indexing fees typically run 0.20% to 0.40% per year, so at 0.40% they’d pay nearly $8,000 a year for harvesting that no longer happens. Contributors to the Journal of Financial Planning noted that the fee difference alone can consume the projected tax benefit well within ten years.
Many investors are a poor fit for direct indexing
Direct indexing suits a specific profile. The ideal user sits in top tax brackets, adds new cash regularly, realizes large capital gains year after year from other activities, and gives appreciated securities to charity. New deposits create fresh lots to harvest, and recurring gains give every loss a job.
Unfortunately, the marketing for direct indexing has reached far beyond that profile. A retiree who direct-indexed home sale proceeds has no future gains to offset, so losses pile up against the $3,000 of ordinary income allowed each year. A salaried professional whose growth sits inside retirement accounts has little taxable gain to shelter. Anyone who needs the money within a decade will realize the embedded gains anyway, right around when the harvesting stops. If those sound familiar, the industry probably oversold the strategy to you, and unwinding it deserves more care than entering it took.
Ways to unwind a direct indexing portfolio
No single exit fits everyone, and sound transition plans usually combine several of the moves below, sized to the embedded gains and your tax situation. The options fall into two groups. Conventional options leave the positions in place and reduce the cost of waiting. The rest restructure the portfolio itself, each through a different mechanism with its own price. The table compares all six at a glance.
| Option | What it does | Typical costs | Best suited for |
|---|---|---|---|
| Turn off the service | Stops fees and new swaps while positions stay put | Your own time and oversight | Anyone paying fees for rare trades |
| Donate appreciated lots | Removes the gain on gifted shares and supports a deduction | None beyond the gift itself | Investors who already give to charity |
| Hold for the step-up | Defers the gain until basis resets for heirs | Decades of drift and any remaining fees | Money you never intend to spend |
| Section 351 exchange | Defers the gain by consolidating into one new ETF | Fund fees of ~0.09% to ~0.54% per year, well above plain index funds | Portfolios that meet minimums, owners content with one fund |
| Long-short overlay | Generates fresh losses with leverage to fund a transition | Fees near 1% on extended positions plus financing costs | Large accounts with recurring gains and real risk tolerance |
| Sell down under a gains budget | Realizes gains deliberately across several tax years | Planned taxes at rates you control | Any investor at any account size |
Options that keep the portfolio in place
None of these require selling a single share.
Turn off the ongoing service
You can usually switch off the strategy or transfer every position in kind to an ordinary brokerage account. The transfer sells nothing and triggers no tax. The fee stops, the wash-sale swapping stops, and dividends can pay out as cash rather than buying more of the problem. The freeze itself remains, and several hundred positions become yours to watch, but halting the direct indexing service saves on fees. For many investors this is the sensible first move no matter what follows.
Give appreciated lots to charity
If you already give to charity, give stock instead of cash. Donating long-held appreciated shares to a public charity or donor-advised fund removes the capital gain entirely and supports a deduction of up to 30% of adjusted gross income. Target the lowest-basis lots, where the embedded gain per dollar is largest. Gifting chips away at a frozen portfolio each year. It rarely unwinds one by itself.
Hold for the step-up in basis
Cost basis resets for your heirs under current law, which erases the embedded gain entirely. For money you never intend to spend, especially later in life, holding can be the right answer. It carries real costs in the meantime, from decades of drift to whatever fees continue. For a 45-year-old’s taxable savings, waiting for the step-up is less a plan than a decision never to use the money.
Options that restructure the portfolio
These paths change what you actually hold, and each comes with friction, fees, or both.
Exchange into a new ETF under Section 351
A newer path contributes your whole basket into a brand-new ETF at launch and hands back fund shares, with the gain deferred rather than realized. One transaction resolves the operational mess, but the doorway is narrow. The exchange is a one-time event at the launch of a newly seeded ETF, so you join a sponsor’s calendar rather than your own. Recent launches have set minimums between $150,000 and $5 million depending on sponsor and custodian, and the contributed portfolio must pass diversification tests.
Then weigh what you hold afterward. Conversion ETFs have charged between 0.09% and 0.54% per year for broadly diversified stock exposure that conventional index funds price at 0.03% to 0.09%. Selling the new ETF later triggers the very gain you deferred, so the fee spread works like a standing charge on the deferral. Treasury is also reviewing these transactions to determine whether certain arrangements produce abusive tax outcomes, which leaves regulatory risk on the table.
Extend into a long-short overlay
The industry’s own remedy for frozen portfolios adds leverage. A manager shorts some stocks and buys extra longs around your existing holdings, and the added trading manufactures fresh losses that can fund a gradual transition. The same study that measured the freeze also promotes this extension, which is worth knowing when you weigh the pitch. The costs deserve equal billing. Management fees near 1% commonly apply to the extended positions, financing adds its own spread, and minimums often start between $1 million and $3 million. Tracking error grows with the leverage ratio, and unwinding the overlay quickly can surrender most of the accumulated tax benefit. Exchanging one hard-to-exit strategy for a more complex strategy that is also hard to exit deserves real hesitation.
Sell down under a gains budget
The general-purpose path works at any account size and requires no eligibility tests. Decide how much capital gain you can absorb each year at acceptable rates, accounting for the 0%, 15%, and 20% brackets and the 3.8% surtax on investment income. Then sell in deliberate order. Highest-basis lots go first because they raise cash with the least gain. Weaker businesses go before stronger ones. Charitable gifts and lower-income years fold in as they arrive. Spread across several tax years, a planned unwind usually costs far less than the single liquidation your statement implies. Tax rules change, so build any multiyear plan with your tax professional.
How individual stock research changes the unwind
A gains budget answers when to sell. What to sell first, and what never to sell, is a research question, and most advisory firms are not built to answer it. A model-driven firm may even liquidate an incoming account of 350 stocks and move the proceeds into its funds, which lands the entire tax bill in one year. Some firms might decline the account altogether.
We see a portfolio with hundreds of positions as an opportunity. Magnifina invests using original research on individual companies, so a frozen direct indexing account looks like raw material to us. A few dozen of those holdings could be businesses we’d consider owning for their own merits. The rest would be sold on a schedule set by the gains budget, with the weakest and highest-basis positions going first. The concentration that drifted in with the winners is managed deliberately, and comprehensive financial planning ties each year’s sales to your actual tax picture.
Where to go from here
Direct indexing is a specialized tool that Wall Street oversold to many investors unsuited for it. If your portfolio has become a stuck pile of stocks that tracks its index a little worse each year, leaving it alone or liquidating everything are poor choices. A patient unwind built around which holdings deserve to stay beats both. If you want help deciding what to keep, what to sell, and when, start with our intro survey. It asks four questions, takes about a minute, and shows whether we are a good fit.
Unwinding Direct Indexing FAQ
Why did my direct indexing account stop generating tax losses?
Harvesting sells stocks below cost and replaces them, which resets basis lower with each round. After a few years of rising markets, nearly every position sits above its basis, so nothing remains to sell at a loss. Research on the strategy shows most of the benefit arrives in the first five years.
Should I turn off direct indexing once the harvesting stops?
Turning off the service or transferring the positions in kind usually costs nothing in tax and stops the fee. The embedded gains and the drift remain, so treat it as a first step while you decide on a full unwind plan.
Can I transfer a direct indexing portfolio to another advisor without selling?
Usually yes. Positions can move in kind through a standard account transfer without realizing gains. The practical question is whether the receiving firm can manage hundreds of individual stocks or would simply liquidate them, so ask that directly before you move.
Is a Section 351 exchange a good way to exit direct indexing?
It can be for portfolios that meet the minimums and diversification tests, since the gain defers into a single new ETF. Weigh the fund’s ongoing fee against plain index funds, the fixed launch windows, and the fact that selling the ETF later still triggers the deferred gain.
How much tax will unwinding a direct indexing portfolio cost?
That depends on the embedded gains, your bracket, and the method. Spreading sales across several tax years under a gains budget, pairing them with charitable gifts, and keeping the holdings worth owning can reduce the cost substantially compared with liquidating all at once. A tax professional should review any multiyear plan.
