No Losses Left to Harvest: Direct Indexing Tax Alpha Reset

When a direct indexing account runs out of harvestable losses, there are only four things that genuinely work: add new money, which brings in lots at today's prices; let dividend reinvestment keep seeding fresh basis; move the appreciated shares out through charitable giving or a step-up at death rather than selling them; or accept that the tax phase of the strategy is over and stop paying for it. Everything else on offer, and there is a lot on offer, is either a repackaging of one of those four or a leveraged product that manufactures losses at a cost high enough to eat the benefit.

This is the least discussed part of direct indexing and the question people arrive at around year four or five, usually after noticing that the harvesting reports have gone quiet while the fee has not. Educational content only, not investment or tax advice. Confirm anything here with your own CPA before acting on it.

Why do direct indexing accounts run out of losses?

Because harvesting is a one-way ratchet on your cost basis. Every time you sell a position below what you paid and buy a replacement at the current price, you book a loss and reset that holding's basis lower. Repeat that a few hundred times through a market that rises over time and you gradually convert a portfolio of mixed lots into one where almost every share is worth more than you paid. There is nothing left underwater to sell.

The pattern is consistent across every provider that discusses it honestly. Harvesting is heaviest in year one, when an account funded with cash holds lots sitting right at their purchase price and any dip pushes them below. It stays productive through the first few years, especially across volatile stretches. Then it thins out, and in the later years most of what remains comes from newly purchased lots rather than the original portfolio: reinvested dividends, fresh contributions, shares from corporate actions. All of those enter at current prices, which is precisely why they can fall below them.

The fee, meanwhile, is flat. A 0.40% account costs the same in year eight as it did in year one while the thing you are paying for has quietly shrunk to a fraction of its original size. That asymmetry is the entire problem, and it is why "how long will I hold this" belongs in the decision at the start rather than as a discovery in year five.

I have been direct indexing for four years and have no losses left. What actually works?

Four things, in rough order of how much they help relative to what they cost.

New contributions are the cleanest fix. Money added today buys shares at today's prices, so those lots sit right at the waterline and any subsequent dip makes them harvestable. This is the mechanism every provider quietly relies on, and it is why direct indexing works best for someone still accumulating rather than someone who funded an account once and left it. If you are adding meaningfully each year, the account keeps a rolling supply of fresh basis and the decay is much slower than the theory suggests.

Dividend reinvestment does a small, continuous version of the same thing. Every reinvested distribution is a new lot at the current price. On a broad US equity portfolio yielding somewhere over 1%, that is a modest but genuinely permanent trickle of new basis, and it is a large share of what late-stage accounts still manage to harvest. It also happens to be the single most common source of accidental wash sales, so if you are running this yourself, know exactly which accounts have reinvestment switched on.

Move the appreciated shares out instead of selling them. This is the one that changes the arithmetic rather than extending it. Donating long-held appreciated shares to a donor advised fund or a charity generally lets you deduct the fair market value without realizing the embedded gain, which is the most efficient exit a low-basis position has. Shares held until death currently receive a step-up in basis for the heirs. Both of these turn a deferred tax bill into one that is never paid, which is the outcome the whole strategy was implicitly aiming at. A direct indexing account late in its life is, structurally, a very good pool of appreciated lots to give away.

Or stop paying for it. If you are not adding money, not giving shares away, and the harvesting reports have gone quiet, you are paying a premium for a benefit that has largely been consumed. Winding down is awkward because selling hundreds of low-basis positions realizes exactly the gains you spent years deferring, but many platforms will let you transfer the shares out in kind and simply hold them. You keep the portfolio and stop paying the management fee on a service that is no longer producing much.

Do the loss-generating products actually help?

Some platforms answer harvest depletion with long short or leveraged structures. The mechanics are real: adding short positions creates holdings that gain when the market falls and lose when it rises, so in any market direction something in the book is underwater and harvestable. These strategies genuinely produce far more realized losses than a long-only account.

The problem is the price. Frec's Long Short tier, for example, runs 0.50% to 1.30% a year plus 0.23% to 0.86% in financing costs, against 0.09% to 0.35% for its plain direct indexing, and it starts at $100,000. So you may be paying somewhere north of 1% all-in to generate losses whose after-tax value, on the most conservative published estimates, is a fraction of a percent. That can still work for someone in a very high bracket with large recurring gains, which is exactly who these products are sold to. For everyone else the fee swallows the benefit, and you have taken on leverage and short exposure to get there.

Treat any product pitched as a fix for depleted harvesting with the same subtraction you would apply anywhere else: what does it cost, what is the realistic after-tax value of the losses it produces, and what happens if the market moves against the structure. If the answer to the first question is bigger than the second, the product is solving the provider's problem rather than yours.

Is there cost-basis zero-out risk after five to ten years?

"Zero-out" overstates it, but the direction is right. Harvesting does not drive your basis to zero. It drives it down toward the prices at which you last transacted, and in a rising market those are still well below today's value. What accumulates is a large unrealized gain spread across hundreds of small positions, which creates two practical problems that have nothing to do with harvesting.

The first is exit friction. Leaving a direct indexing account is not like selling a fund. You are either transferring hundreds of individual lots to another broker, which is doable but tedious and occasionally messy with fractional shares, or you are liquidating and realizing every deferred gain at once. Plenty of people discover this at the moment they want to leave.

The second is concentration drift. A sampled basket that has been harvested for years is no longer the index it started as. Every replacement trade moved it slightly, and the positions you were never able to sell, because they only ever went up, are now the largest weights in the account. You can end up with meaningful single-name concentration you never chose. It is worth measuring rather than assuming, and it is the same question we work through in how much of your portfolio should be in one stock.

Would a broader index have more harvesting opportunity?

In principle yes. A Russell 1000 or total-market basket holds more names, more dispersion between them, and more small and mid cap volatility, all of which produce more positions below cost in any given period. If you are choosing an index at the start and harvesting is the point, breadth helps.

In practice the effect is smaller than it sounds, for two reasons. The added names carry small weights, so the extra harvestable dollars are a modest share of the portfolio even when the count of losers rises a lot. And tracking a wider index with a sampled basket introduces more tracking error, which is a real cost paid in return dispersion rather than in fees. Switching indexes in an existing depleted account does not help at all, because the switch itself would realize the accumulated gains. This is a decision for the start of an account, not a repair for the middle of one. Frec publishes the widest retail index menu if you want to make it deliberately.

How do you know when harvesting has actually stopped working?

Ask your provider for realized losses harvested per year since inception, as a percentage of account value, and put it next to the fee you paid each of those years. That single comparison answers the question better than any general rule, and every platform can produce it.

What you are looking for is the crossover: the year in which the after-tax value of the harvest fell below the fee premium you are paying over a plain index fund. Convert harvested losses to money at your own marginal rate rather than a headline rate, and only count losses you could actually use, meaning you had gains to offset or room under the $3,000 ordinary income limit. A large pile of carried-forward losses you have no way to use is not a benefit, it is a receivable of uncertain date. Once the crossover has happened and you have no plan to add money or give shares away, the strategy has finished its job.

This is also when the timing question tends to become concrete for people who are approaching a liquidity event from the other direction. Harvested losses are worth the most in a year with large realized gains, so a business owner planning a sale has a genuine reason to keep harvesting capacity alive until the transaction closes. If that is your situation, the sequencing matters more than the harvesting does, and it is worth having a defensible read on what the business is actually worth before you plan the tax year around it.

What to do next

If the account is still young, the useful move is to keep it fed. Regular contributions are what keep a direct indexing account productive, and they do more for long-run tax alpha than any product upgrade a platform will sell you.

If it is mature and quiet, decide deliberately rather than by inertia. Either the appreciated shares have a destination that avoids the gain, which makes the accumulated position a feature, or they do not, in which case you are paying a management fee for maintenance rather than for tax value. Both are reasonable answers. Drifting is the expensive one.

And if you are still deciding whether to start, the honest framing is that this is a front-loaded benefit with a flat cost. The full economics, including what the published research actually says the benefit is worth, are in direct indexing tax loss harvesting. The rules that decide whether a harvested loss counts at all are in tax loss harvesting rules, and the fee and minimum comparison across every provider is on direct indexing platforms. If you want to see how a 30 or 50 stock construction would have behaved against the full index before you commit to anything, that is what the builder above is for.