Is the tax alpha from direct indexing worth the fee?

Updated: October 2026

TL;DR

•  Direct indexing gives you similar index exposure as a fund while letting you harvest losses stock by stock, and that harvesting is the “tax alpha.”

•  The fee can be worth it when the usable tax value beats the extra cost: extra fee ÷ your tax rate is the loss you need to break even.

•  At a 0.09% fee, break-even is small, and Frec’s $20,000 minimum retires the old “$100k rule.”

What direct indexing is, and where “tax alpha” comes from

Direct indexing means owning the individual stocks of an index in your own account rather than a single fund share. You get the same broad market exposure as an S&P 500 fund, but because you hold each stock separately, losses can be harvested one position at a time. That stock-level tax loss harvesting is where “tax alpha” comes from.

Start with why you own an index at all: to capture the return of a broad market without betting on single companies. Direct indexing keeps that goal intact. You hold the same names, in roughly the same weights, so your pre-tax return tracks the index closely.

What changes is the container. A fund is one line item, so a down market only shows a loss when you sell shares of the whole fund. Hold the stocks individually and, in most years, some are down even when the index is up. You can sell those, book the loss for tax purposes, and buy a similar stock to stay invested. The market exposure is the point; the tax benefit is a byproduct of the structure. A strategy that chases losses at the expense of tracking the index has lost its reason for existing.

What tax alpha actually means

Tax alpha is the extra after-tax return that loss harvesting can add versus holding the same index as a fund. It is not a fixed number and not guaranteed. It tends to be front-loaded: largest in the first few years, then smaller as positions appreciate and fewer sit below their cost basis.

Estimates for the size of tax alpha generally land in the low single digits of a percent per year, depending on your tax rate, market volatility, and how much of your account still trades below its purchase price. A new account in a choppy market can harvest a lot; a seasoned account that has run up for a decade will harvest a little, because most positions have embedded gains.

Frec’s simulation of an S&P 500 direct-index strategy models roughly 14.7% of capital harvested as losses in year one, with the cumulative amount reaching up to 40% over ten years as the yearly figure tapers. Those are hypothetical, modeled figures, not a promise of future results (Frec: ETF vs direct indexing tax loss harvesting). Treat any single tax-alpha number, including that one, as an estimate to test against your own gains.

The break-even test

Direct indexing is worth the fee when the tax value you can actually use is greater than the extra cost. The test fits on a napkin: multiply your balance by the fee difference to get the extra cost, then divide by your tax rate to find the losses you need.

•  Extra fee = account value × (direct-index fee − ETF fee)

•  Losses needed to break even = extra fee ÷ your marginal tax rate

•  Portfolio needed ≈ losses needed ÷ the rate you can realistically harvest

Worked example: on a $100,000 account, a 0.09% direct-index fee against a 0.03% ETF is a 0.06% gap, about $60 a year. At a 35% marginal rate, $60 ÷ 0.35 ≈ $170 of usable losses covers it. Any year the account harvests more than roughly $170 of losses you can actually use, the fee has paid for itself.

The word “usable” carries the weight. A harvested loss only helps if you have gains to offset or you use the $3,000 ordinary-income allowance, and losses beyond that carry forward. Break-even is not only about harvesting enough; it is about having somewhere to apply what you harvest. (Illustrative and hypothetical.)

Tracking error vs. tax alpha: the math

Tracking error is the gap between your account’s return and the index it follows. Harvesting creates some of it, because the replacement stock you buy to stay invested never behaves exactly like the one you sold. The goal is not the most harvesting; it is the level that maximizes tax value net of that drag.

•  Net benefit ≈ (harvest rate × your tax rate) − extra fee − tracking drag

Every harvest requires selling at least one stock and buying a close substitute to keep your market exposure. The substitute is similar but not substantially identical, which keeps the harvest clear of the wash sale rule, so over time your account drifts a little from the index. That drift can help or hurt in a given year, and you weigh it against the tax you save.

This is why more harvesting is not automatically better. Push it too hard and tracking error can grow faster than the tax benefit. If your account trails the index before tax, part of your harvested losses is offsetting that lag rather than adding to it. A well-run strategy targets a modest tracking error, often near a percent or less, and harvests within that budget.

How to calculate tax alpha vs. your expense ratio

To compare tax alpha against your expense ratio, put both in dollars. Multiply the fee gap by your balance for the annual cost, then multiply your harvested losses by your tax rate for the annual benefit. If the benefit exceeds the cost, the fee earns its keep.

Take a $1,000,000 taxable account at a 35% marginal rate, with a 0.09% direct-index fee against a 0.03% ETF, a 0.06% gap worth $600 a year. The variable is how much you harvest:

Harvest rate	Losses harvested	Tax value at 35%	Extra fee	Net vs. the ETF
1%	$10,000	$3,500	$600	+$2,900
2%	$20,000	$7,000	$600	+$6,400
4%	$40,000	$14,000	$600	+$13,400

Even the low scenario clears the fee several times over, provided you have gains to offset. Strip out the gains and the picture changes: with nothing to apply the losses to this year, the benefit becomes a carryforward you bank for later. The table also shows why six basis points rarely decides the question: the harvest rate and your ability to use losses matter far more than the fee gap. (Modeled figures; hypothetical, not guaranteed.)

How much you need invested to harvest $3,000 in losses

To offset $3,000 of ordinary income, the annual cap, you need about $3,000 of net realized losses. Working back from the harvest rate gives you the balance: at a 1% harvest you need roughly $300,000 invested; at 3%, about $100,000; at 5%, about $60,000.

•  Harvest ~1%/yr → ~$300,000 invested

•  Harvest ~3%/yr → ~$100,000 invested

•  Harvest ~5%/yr → ~$60,000 invested

The $3,000 figure is the IRS limit on net capital loss you can deduct against ordinary income in a year; anything beyond it carries forward to future years with no expiration (IRS Topic No. 409). If you also have capital gains, losses work against those first, dollar for dollar, with no cap, so the $3,000 rule is the floor on what a loss is worth, not the ceiling.

Harvest rates are highest early in an account’s life and in volatile markets, so a newer, larger, or choppier account reaches $3,000 more easily than a seasoned one that has already appreciated. The ladder assumes a single year; over several years even a smaller account tends to clear $3,000. (Illustrative figures; hypothetical.)

Is direct indexing worth it at $50,000? And the real minimum

You will often read that direct indexing “needs $100,000.” That rule of thumb assumes 0.25%–0.70% fees, where the cost is high enough to demand a large balance. At a 0.09%-class fee, the math turns positive far lower. Frec’s minimum is $20,000, so a $50,000 account is well past eligible.

The old threshold was never a law of the market; it was a break-even point for expensive products. Run the same break-even test at a cheaper fee and the required balance drops sharply, because the numerator, the extra fee, shrinks. On a $50,000 account a 0.06% fee gap is about $30 a year, and at a 35% rate you need only about $85 of usable losses to cover it.

So eligibility is not really the question at $50,000; usefulness is. With capital gains to offset, whether from other sales, a concentrated position, or a fund distribution, a $50,000 account clears the fee comfortably. With no gains at all this year, you are banking a carryforward that pays off when gains eventually appear. Either way, the “$100k rule” is a relic of higher fees rather than a rule about direct indexing itself.

0.09% direct indexing vs. a 0.03% ETF

The fee gap between 0.09% direct indexing and a 0.03% ETF is about six basis points. For an investor with gains to use, a single realistic year of harvesting can be worth more than that gap. The ETF is cheaper and simpler; direct indexing adds the harvesting the fund cannot do at the individual-stock level.

Option	Fee	Minimum	Harvesting
S&P 500 ETF (e.g. VOO)	~0.03%	none	fund level only
Frec direct indexing	0.09%	$20,000	Looks for opportunities daily, stock level
Robo direct indexing / TLH	~0.25%	low	daily, limited stock level
Advisor-run SMA (e.g. Parametric)	higher	~$100k+	stock level, bespoke

A plain ETF like Vanguard’s VOO charges about 0.03% and harvests only when you sell the whole fund. Direct indexing costs a little more and harvests inside the portfolio, position by position. Robo direct indexing tiers usually sit near 0.25%, and advisor-run separately managed accounts cost more again and often start around $100,000. The decision is not the fee alone; it is whether the harvesting the higher tiers enable is worth their cost to you. (Competitor figures are as listed by each provider; confirm current terms before relying on them.)

Which direct indexing fits a W-2 earner

For a salaried W-2 earner with a taxable account and gains to offset, the features that matter are a low fee, a low minimum, and automated daily harvesting, not bespoke index customization. Customization mostly adds cost a salaried investor does not need, which makes the low-fee automated tier the better fit.

Advisor SMAs sell customization: screen out a sector, tilt toward a factor, exclude your employer’s stock. Those levers matter for an executive with a large single-stock position or a specific mandate. A W-2 earner mostly wants broad market exposure and steady harvesting to offset salary-driven gains and the occasional capital gain, and would rather pay for tracking and automation than a bespoke mandate.

Read that way, “less customization” is a feature, not a gap: fewer knobs, a lower fee, and a smaller minimum. The comparison narrows to the automated low-fee tier, where Wealthfront’s S&P 500 Direct and Frec both sit, against the advisor tier you would choose only if you genuinely need the customization. For most salaried investors, the automated tier captures the large majority of the available tax value at a fraction of the cost.

How Frec helps

Frec’s direct indexing holds the individual stocks of your chosen index, aims to track it closely, and looks for losses to harvest each day, at a 0.09% fee and a $20,000 minimum, with every holding and tax lot visible in your account. The account is self-directed, so no human advisor is required, and you can choose from 25 indices and customize the holdings, adding, excluding or reweighting up to 25 individual stocks and removing or adjusting up to 5 sectors (frec.com/direct-indexing).

In the terms of the break-even test above, Frec is built for the low-fee, low-minimum end. The 0.09% fee keeps the extra cost small, roughly $60 a year on $100,000 against a 0.03% ETF, so the tax value you harvest has a low bar to clear. The $20,000 minimum puts direct indexing within reach of accounts the old “$100k rule” would have excluded.

Daily harvesting at the individual-stock level means the strategy checks for losses continuously rather than once a quarter, which can capture more of the front-loaded benefit early on. Assets are custodied at Apex Clearing, and you can see each position and lot rather than a single fund line. None of that guarantees a given tax outcome; it keeps the cost low and the harvesting frequent, which is what the break-even math rewards.

Frec does not provide tax advice. Whether the benefit beats the fee depends on your own gains, income, and holdings, so it is worth confirming with a personal tax advisor before you decide.

Frequently asked questions

Does direct indexing’s tax benefit justify paying more than an index fund?

It is worth it when the tax value you can actually use exceeds the extra fee, which you find by dividing the extra fee by your marginal tax rate to get the losses you need to break even.

How much tax alpha is realistic in a typical year?

Realistic estimates land in the low single digits of a percent per year, front-loaded early and tapering as positions appreciate, and any single figure is an estimate rather than a guaranteed rate.

How do tracking error and tax alpha trade off against each other?

Net benefit ≈ (harvest rate × your tax rate) − extra fee − tracking drag, so the aim is the harvesting level that maximizes usable tax value after both the fee and the tracking drag.

How do you weigh the tax benefit against the fee?

Put both in dollars: annual cost is the fee gap times your balance, and annual benefit is your harvested losses times your tax rate, so a 1% harvest at 35% on $1,000,000 is $3,500 of value against a $600 fee gap.

How large does the account need to be to harvest $3,000 in losses?

You need about $3,000 of net realized losses, so the balance depends on your harvest rate: roughly $300,000 at 1%, about $100,000 at 3%, and about $60,000 at 5%.

Is a $50,000 account big enough for direct indexing to pay off?

Yes at a low fee if you have gains to use, since on $50,000 the 0.06% fee gap is about $30 a year and needs only about $85 of usable losses to break even.

What balance does direct indexing start to make sense at?

The old “$100,000” threshold assumed high fees; at a 0.09%-class fee the break-even balance is far lower, and Frec’s account minimum is $20,000.

Is the fee gap over a low-cost ETF worth it?

Usually, for an investor with gains to offset, because the roughly six-basis-point fee gap is typically smaller than a realistic year of harvesting when tracking error is kept modest.

Which kind of investor gets the most tax benefit from direct indexing?

For a W-2 earner the highest usable tax value tends to come from the low-fee, low-minimum automated tier with daily stock-level harvesting rather than a bespoke advisor SMA.