Tax Loss Harvesting for Small Portfolios: Is It Worth It? (Run the Math First)
A $12,000 index fund position sitting 8% underwater generates a $960 realized loss. At a 22% marginal rate, selling it and buying a near-identical fund puts $211 back in your pocket at tax time — about twenty minutes of work for the price of a tank of gas.
That’s the honest starting point for anyone asking whether tax loss harvesting small portfolios is worth the effort. This post gives you the actual formula, four worked scenarios across account sizes from $10,000 to $150,000, the three costs that never show up in the calculators, and the specific portfolio size where the strategy stops being a rounding error and starts being real money.
The 30-second answer: the tax loss harvesting small portfolios formula
Everything about this strategy collapses into one line:
Year-one tax savings = min(realized loss, capital gains + $3,000) × your applicable tax rate
Two pieces of that formula do all the work. First, realized losses offset realized capital gains dollar for dollar with no ceiling. Second, once your gains are wiped out, the IRS lets you deduct a maximum of $3,000 per year of net capital loss against ordinary income — $1,500 if you’re married filing separately. That limit is set by statute and did not change for 2026. Anything above it carries forward indefinitely, keeping its original short-term or long-term character, until it’s used up (IRS Topic No. 409).
Here’s the part most calculators quietly skip: harvesting a loss also lowers your cost basis in the replacement position. You have not erased a tax. You have moved it to a later year. The real profit is the rate spread — you deduct today at your ordinary income rate and you pay later at the long-term capital gains rate. Deducting at 22% and paying back at 15% nets you 7 cents on the dollar, plus whatever the deferred cash earns in the meantime.
Four scenarios: what tax loss harvesting small portfolios actually pays
Assume a buy-and-hold index investor in a taxable brokerage account with no realized gains to offset — the default situation for most DIY investors. Assume a garden-variety drawdown and that every dollar of harvested loss above $3,000 gets banked as a carryforward.
| Taxable account | Drawdown | Harvestable loss | Deductible this year | Marginal rate | Year-1 savings | Carried forward |
|---|---|---|---|---|---|---|
| $10,000 | −10% | $1,000 | $1,000 | 22% | $220 | $0 |
| $25,000 | −15% | $3,750 | $3,000 | 22% | $660 | $750 |
| $60,000 | −20% | $12,000 | $3,000 | 24% | $720 | $9,000 |
| $150,000 | −20% | $30,000 | $3,000 | 24% | $720 | $27,000 |
Look at the “Year-1 savings” column and the whole debate resolves itself. Between the $60,000 account and the $150,000 account — a 2.5x difference in assets and a $18,000 difference in harvested losses — the immediate tax benefit is identical. The $3,000 deduction cap flattens the curve the moment your losses exceed it, which for a 20% drawdown happens at roughly $15,000 of taxable equities.
The corollary matters more than the table itself: for a portfolio under about $20,000, the ceiling almost never binds, so your entire loss is deductible in year one and the strategy is simple and clean. For a portfolio well over that, you’ll be sitting on a carryforward you can only spend $3,000 at a time — a $27,000 carryforward takes nine tax years to burn off unless you generate offsetting gains. Which, incidentally, is exactly what makes those banked losses valuable later when you rebalance a 3-fund portfolio and are forced to realize gains in the taxable sleeve.
Three costs that never make it into the calculator
1. The wash sale rule, and its ugly cousin. If you buy the same or a “substantially identical” security within 30 days before or 30 days after the sale, the loss is disallowed and rolled into the basis of the replacement shares (IRS Publication 550). The trap that catches DIY investors is that the window covers all your accounts, including your IRA — and a wash sale triggered by an IRA purchase permanently destroys the loss rather than deferring it. Automatic dividend reinvestment inside the 61-day window is the single most common way people accidentally blow this up.
2. Basis reduction, which turns a benefit into a loan. Harvesting is a deferral, not a discount. If you’re currently in the 0% long-term capital gains bracket — taxable income up to $49,450 single or $98,900 married filing jointly for 2026 — you are deducting losses today against income you might have shielded anyway, and reducing basis on shares you could have sold tax-free later. In that scenario harvesting can be mildly negative. This is the same sequencing question we worked through in our breakdown of tax loss harvesting vs Roth conversion: the order in which you use limited tax capacity is often worth more than any single move.
3. Tracking error and friction. To stay invested for 31 days you need a replacement fund that isn’t substantially identical but tracks close enough that you don’t miss a rebound. Swapping a total US market fund for an S&P 500 fund is the standard move, but the two are not the same portfolio, and if the small-cap tail rips during your 31 days you can lose more in performance than you gained in tax. On a $10,000 position, a 0.5% divergence over a month costs $50 — a quarter of the entire benefit in scenario one.
When tax loss harvesting small portfolios starts to pay for itself
The academic ceiling on this strategy is well documented. Chaudhuri, Burnham, and Lo, publishing in the Financial Analysts Journal in 2020, tested harvesting across the 500 largest US securities from 1926 to 2018 and found a before-transaction-cost tax alpha of 1.08% per year at 15% long-term and 35% short-term rates. Once they enforced the wash sale rule, that fell to 0.82% per year. Vanguard’s 2024 research puts the range at 0.47% to 1.27%, and adds the caveat that matters most here: the figure must be scaled by the size of your taxable equity assets relative to your whole portfolio.
Run that scaling against real numbers and the picture sharpens:
| Taxable equity assets | At 0.82%/yr (wash-sale constrained) | Hours/year to run it | Effective hourly rate |
|---|---|---|---|
| $10,000 | $82 | 2 | $41 |
| $25,000 | $205 | 2 | $103 |
| $50,000 | $410 | 3 | $137 |
| $150,000 | $1,230 | 4 | $308 |
For context on what “small” means in practice: the Federal Reserve’s 2022 Survey of Consumer Finances found that among the 21% of families holding stock directly outside of funds and retirement accounts, the median holding was $15,000 — the lowest on record. Most people asking this question are somewhere in the first two rows of that table.
My working rule: below roughly $25,000 in taxable equities, harvest opportunistically, not systematically. Wait for a genuine drawdown of 10% or more, do it once, and move on. Between $25,000 and $100,000, a once-a-year December sweep plus one opportunistic trade during any sharp correction captures most of the available value. Above that, it’s worth a calendar reminder and a spreadsheet.
And a prerequisite that outranks all of this: if you still have unused space in a 401(k), HSA, or IRA, filling those beats harvesting losses in taxable by a wide margin. Our walkthrough of the tax-advantaged accounts order of operations covers why. The same goes for the placement question — deciding which assets sit in which account, covered in our look at asset location vs asset allocation, is a once-and-done decision that quietly outperforms most annual tax maneuvering.
The 20-minute version, step by step
- Turn off automatic dividend reinvestment in the taxable account before you do anything. This is the number one wash sale trigger and it takes one click.
- Pull up unrealized gain/loss by tax lot, not by position. A position that’s up overall can still contain lots purchased at a higher price. Every major brokerage shows this; make sure your cost basis method is set to specific identification rather than average cost.
- Sell only the losing lots. Note the exact dollar loss and whether it’s short-term or long-term — short-term losses are more valuable because they offset short-term gains taxed at ordinary rates first.
- Buy the replacement fund the same day. Staying out of the market to be safe is the expensive mistake; a broad-market alternative with a different index keeps your exposure intact.
- Set a 31-day calendar reminder. After that window you may switch back if you want, though there’s rarely a reason to.
- Record the carryforward. It goes on Schedule D and it’s yours indefinitely — but only if you actually track it. Losses lost to sloppy record-keeping are the most common way this strategy underdelivers.
Curious what a $660 tax refund compounds into over 20 years if you reinvest it instead of spending it?
What I’ve learned running this on my own accounts
I’m a software engineer, which means my instinct with anything like this is to automate it — write a script, pull the tax lots via API, flag anything below a threshold, done. I built roughly that a few years ago, mostly out of curiosity about whether the strategy financial Twitter treats as free money actually moved the needle in a portfolio my size. The honest answer: it works, and it’s smaller than advertised. In a good year I harvested enough to fill the $3,000 deduction; in flat years the script found nothing worth trading.
What surprised me was the behavioral side, which I’ve come to think is underrated in personal finance generally. Having a mechanical rule for selling losers made me much less twitchy during drawdowns, because a red position became a task instead of a threat. That’s not a return you can put in a spreadsheet, but for a DIY investor with no advisor talking them off the ledge, it’s arguably worth more than the 82 basis points. The rest of my approach is deliberately boring — index funds, tax-advantaged accounts filled first, rebalance on a band rather than a calendar — and tax loss harvesting sits exactly where it belongs in that stack: a nice-to-have, executed twice a year, never the reason for a trade.
Frequently asked questions
Is tax loss harvesting worth it on a $5,000 portfolio?
Rarely as a standalone activity. A 10% drawdown on $5,000 produces a $500 loss, worth $110 at a 22% marginal rate. That’s real but it’s not worth building a process around, and one accidental wash sale from a reinvested dividend erases it. If your brokerage or robo-advisor does it automatically at no extra cost, let it run. If it requires manual work, spend that hour increasing your contribution rate instead.
Can I harvest losses inside my 401(k) or IRA?
No. Losses in tax-deferred and tax-free accounts have no tax consequence because gains and losses inside them aren’t reported annually. Worse, buying a substantially identical security in your IRA within 30 days of a taxable-account sale disallows that loss permanently — it does not get added to the IRA’s basis. Keep the two sides of your portfolio operationally separate.
How long can I carry forward unused capital losses?
Indefinitely, until they’re fully used. Each year you apply them first against capital gains of the same character, then against the opposite character, then up to $3,000 against ordinary income. Short-term carryovers stay short-term and long-term stay long-term, which matters because short-term losses shield income taxed at higher rates. Unused losses do not transfer to heirs, so a very large carryforward on a small portfolio is a signal to plan for realizing gains rather than to keep harvesting.
This article is educational, not tax advice.
Capital loss rules interact with your full return, including state taxes, the net investment income tax, and any business income. Confirm your specific situation with a CPA or enrolled agent before acting.
