Does direct indexing actually save you money?
Instead of one index fund, you own the hundreds of stocks inside it — so the losers can be harvested for tax losses. It’s a real strategy with a real, narrow use case. The pitch is wider than the use case.
Consumer education, not investment advice. Paul Powell is not currently a licensed financial advisor.
For a specific person, yes: a large taxable account, real capital gains to offset, a high bracket, and no intention of ever selling. For everyone else the arithmetic runs the other way. The fee premium arrives every year, in every market. The tax benefit is front-loaded, shrinks as the account matures, and much of it comes back when you sell.
One cost is certain. The other one is a forecast.
Here’s the mechanic, in plain terms. A normal index fund holds the whole index in one wrapper. When some of those stocks fall, you can’t do anything about it — the fund is one position, and the losers are hidden inside the winners. Direct indexing unpacks the wrapper. You own the individual stocks in a separately managed account, so the ones sitting at a loss can be sold, the loss booked against your taxes, and something similar bought in their place. That’s tax-loss harvesting, done at the stock level. It is legitimate, and it does work.
Now the price. Broad index ETFs commonly charge somewhere between 0.03% and 0.10% a year. Direct indexing solutions Morningstar has reviewed commonly run 0.20% to 0.40% a year — and Morningstar’s own comparison found the large platforms’ versions priced roughly 0.35 percentage points a year above the cheapest ETFs. Minimums at established providers commonly sit between $100,000 and $250,000, though newer platforms have pushed that down.
So the strategy has to clear a fee hurdle of roughly a quarter to a third of a percent every year, forever, out of a tax benefit that — this is the part the brochure buries — does not last.
Morningstar’s research measured it. The average estimated tax benefit ran about 0.69% a year over a portfolio’s first five years, with the 10th and 90th percentiles at roughly 0.40% and 1.18%. In years six through ten, the average fell to about 0.09% a year, with the 10th percentile slightly negative. The reason is not complicated: harvesting requires losers, and in a rising market you eventually run out of them. Early on, the account turned over aggressively. Later, there’s nothing left to sell at a loss.
Two more mechanics decide whether any of it reaches you. First, the wash-sale rule. The IRS is explicit: sell a security at a loss and buy a substantially identical one within 30 days before or after, and the loss isn’t deductible — it just adjusts your basis in the replacement. So the harvest is real, but it’s a substitution, not a free lunch.
Second, and larger: a harvested loss is mostly deferral, not forgiveness. Every loss you book lowers your cost basis, which means a bigger gain waiting at the other end. Morningstar’s analysis of typical investor profiles concluded that direct indexing provides “no net tax advantage … once the portfolio is sold,” and that the benefit lands mainly with people who donate the appreciated shares or leave them to heirs. For high-income filers with sizable gains who do exactly that, the same analysis found the tax reduction can approach 2% a year. That’s a genuinely good outcome — for a genuinely narrow group.
And you need gains to offset in the first place. Morningstar’s look at IRS filing data found that most taxpayers report none — roughly 89% of filers in the $80,000 income bracket reported no capital gains the prior year, about 65% in the $300,000 bracket, and even 30% of those above $1 million. With no gains to absorb them, harvested losses offset only $3,000 of ordinary income a year. That’s a $3,000 deduction bought with a fee charged on the entire balance.
Completely. That’s the point.
There is nothing wrong with direct indexing. No hidden commission, no surrender schedule, no trick. It’s a real strategy that some investors should genuinely use. What deserves scrutiny is the modeling — “tax alpha” figures are commonly calculated at the highest federal bracket, which is not most people’s bracket, and are often projected forward on the assumption that the tax savings get reinvested and compound. Both choices push the number up. Neither is disclosed in the headline.
One side of this trade is certain and the other is a projection. The extra fee is charged on your whole balance, every year, whether the market rises or falls, whether losses exist to harvest or not. The tax benefit depends on your bracket, your gains, the market’s path, and what you eventually do with the shares. There’s also an exit problem nobody mentions: after a decade you don’t own one fund you can sell in an afternoon — you own hundreds of appreciated positions with a tax bill attached to unwinding them. Ask what leaving costs before you arrive.
Make them model it at your bracket, not the top one
You don’t have to know whether direct indexing is right for you. You have to make the person recommending it show their work — which is the same standard institutions apply to every strategy that costs more than the plain version. Four facts settle it: is the account taxable, do you have gains to offset, what is the all-in cost against the simple index alternative, and what does the exit look like in year ten.
Get all four in writing. This is exactly what the 12 written due-diligence questions are for, and it’s Step 04 of the Method. If the advisor is recommending a strategy that raises their own fee, that’s also a good moment to check their public record on BrokerCheck and Form ADV — the Form ADV fee section will tell you what the program actually charges, in the firm’s own words.
“Please show the all-in annual cost of this direct indexing account — advisory fee, platform fee, and any underlying costs — next to the cost of holding the same index in an ETF. Then show the projected tax benefit modeled at my actual marginal rate, for years six through ten as well as years one through five, and tell me what liquidating the account would cost me in year ten.”
An advisor who has run that comparison for you already will hand it over. An advisor who hasn’t is recommending a more expensive product without knowing whether it’s worth more. Which of those two is sitting across from you?
You may also be interested in
The 12 Questions
The written due-diligence questions every advisor must answer — and what each answer means.
Read the guide →What’s inside your “all-inclusive” wrap fee
One bundled percentage, and the stack of underlying costs that keeps running underneath it.
Read →The fund fee you never see on a statement
Expense ratios come out before the return is ever reported. No bill, no line item, no receipt.
Read →Done paying a premium for a projection?
The Method makes every recommendation show its work — six steps, in order, free.