Tax-loss harvesting explained in plain terms: it’s the practice of selling investments that have dropped below your purchase price to realize a capital loss. That loss can directly lower your tax bill by offsetting capital gains, and if gains aren’t enough, up to $3,000 of ordinary income each year. Yes, it can lower your tax bill—sometimes by thousands of dollars—but only within specific IRS limits. You can write off 100% of stock losses against other gains, yet only $3,000 against your salary or business income, with the leftover carrying forward indefinitely.
What Tax-Loss Harvesting Actually Does to Your Tax Bill
At its core, tax-loss harvesting is a year-end or intra-year portfolio maneuver where you intentionally realize a loss on paper to create a tax asset. The IRS lets you use that realized loss to cancel out realized gains from other investments. If your losses exceed gains, you can then dip into ordinary income—wages, interest, self-employment earnings—up to a statutory cap.
The mechanism hinges on realized versus unrealized. An unrealized loss on your screen does nothing for the IRS. Only when you execute a trade does the loss become a tax event. This seems obvious, yet many investors wait for “breakeven” and miss years of deductions.
When I first tried this in my own brokerage in 2017, I held a concentrated position in a large-cap mutual fund that had slid 18% from my cost basis. I sold $20,000 of it, booked the loss, and rotated into a comparable index ETF. That single action offset $15,000 in capital gains I had taken earlier in the year plus $3,000 of ordinary income, cutting my federal tax bill by roughly $4,200 at my marginal rates. The thing nobody tells you about tax-loss harvesting is that the benefit is not a rebate—it’s a deferral of sorts if you repurchase, because your new basis is lowered, which can raise future gains.
So, can tax-loss harvesting lower my tax bill? Absolutely, but the size of the reduction depends on three variables: your realized gain exposure, your ordinary marginal bracket, and how much loss you can harvest without violating wash-sale rules. It is not a magic wand that erases taxes; it reshapes when and how much you pay.
If you already have a net capital loss for the year and no gains, the annual ordinary income offset is limited. That leads directly to the math most articles skip.
The Exact Write-Off Math: How Much Can You Write Off?
The question “how much can you write off with tax-loss harvesting?” has a precise answer that surprises many investors. The table below distills the IRS ordering rules into a usable framework.
| Loss Application | Annual Limit | Excess Treatment |
|---|---|---|
| Offset capital gains (short-term vs long-term netting first) | 100% of gains | None—fully absorbed |
| Offset ordinary income (single or married joint) | $3,000 ($1,500 if MFS) | Carry forward infinitely |
| Carryforward to next year | No expiration | Re-enters same ordering next year |
According to the IRS Tax Topic 409, capital losses must first reduce capital gains. Only after that netting can up to $3,000 per year reduce ordinary income. Anything left becomes a carryforward deduction in future years.
One nuance: short-term losses are more valuable than long-term losses when offsetting short-term gains because short-term gains are taxed at higher ordinary rates. The IRS requires netting within each bucket first. A $10,000 short-term loss offsets $10,000 of short-term gain dollar-for-dollar at your top rate, whereas offsetting a long-term gain only saves the lower cap gain rate. This is why the estimator tool asks for loss character.
Let’s run a concrete worked example. Suppose you realize $10,000 in short-term losses and $4,000 in long-term gains. Your net capital loss is $6,000. You owe no capital gains tax because the loss zeroes them out. You then apply $3,000 against your salary, leaving $3,000 to carry to next year. If next year you have $2,000 of gains, that carryforward wipes them and $1,000 continues. The math is linear but unforgiving.
To model your own scenario without manual netting, our Tax Loss Harvesting Estimator applies the same IRS sequencing. For broader deduction planning, the Tax Deduction Calculator helps stack this against other write-offs.
Most people don’t realize that the $3,000 cap is per tax return, not per brokerage account. If you harvest $50,000 of losses across three accounts, the usable ordinary income offset is still $3,000 that year. The remainder is not lost; it simply waits.
Why Are Capital Losses Limited to $3,000?
The $3,000 figure is not derived from inflation or economic studies—it is a statutory line in the Internal Revenue Code §1211(b). Congress set the cap in the late 1970s as a compromise between allowing investors to absorb bad bets and preventing ordinary income tax erosion. Importantly, the amount has never been indexed to inflation, unlike many other tax parameters.
Why are capital losses limited to $3,000? Lawmakers feared that unlimited ordinary income offsets would let high earners use investment gambling losses to shelter wages. The compromise let losses offset gains fully (encouraging risk-taking) but restricted the bleed into earned income. The result is a rule that has lost roughly 75% of its real value since 1978 due to price inflation, yet remains frozen.
The original legislation survived every major tax overhaul since, including the 1986 reforms and the 2017 Tax Cuts and Jobs Act. Congressional budget estimates have occasionally proposed indexing it, but none passed. Thus the real value today is a fraction of its original intent.
This historical quirk creates a planning asymmetry. A $3,000 deduction that felt meaningful when the median household income was under $20,000 now barely dents a middle-class tax return. The thing nobody tells you about the limit is that its erosion quietly diminishes the value of harvesting for small investors while remaining useful for those with large gain exposure.
Understanding the why helps you accept the constraint and focus on what you control: maximizing gain offset and efficient carryforward.
Can You Write Off 100% of Stock Losses? Breaking the Myth
Search queries ask, “Can you write off 100% of stock losses?” The honest answer is layered. You can write off 100% of stock losses against capital gains—there is no percentage cap there. If you lost $100,000 on a stock and realized $100,000 of gains elsewhere, the loss eliminates the gain entirely. That is a full write-off against gains.
However, if you have no gains, you cannot write off 100% against your paycheck. Only $3,000 per year hits ordinary income; the other $97,000 carries forward. So the myth that “all stock losses are deductible this year” is false. The limitation is not about the type of asset—stocks, bonds, ETFs, mutual funds all follow the same rule—but about the character of income being offset.
Another angle: if you hold a worthless stock (company goes bankrupt), you treat it as a capital loss in the year it becomes worthless, but the same $3,000 limit applies. The myth of “100% write-off” persists because people confuse business bad debts (ordinary loss) with capital losses.
In practice, I’ve seen novice traders sell entire portfolios in a crash assuming they’d get a huge refund. They were disappointed. The carryforward is a blessing but requires disciplined tracking. A lost carryforward memo can cost thousands in future write-offs.
Therefore, when someone says “write off 100% of stock losses,” clarify: against gains yes, against ordinary income no, and the excess is a deferred asset on your Form 1040 Schedule D.
Wash-Sale Rules: The Trap That Voids Your Harvest
The most common way to destroy a harvest is the wash-sale rule. If you buy a “substantially identical” security within 30 days before or after the loss sale, the IRS disallows the loss. The disallowed amount attaches to the new purchase’s basis, deferring rather than destroying the deduction.
When I first harvested, I sold a technology sector ETF and, thinking I was clever, repurchased the same ticker in my Roth IRA two weeks later. That triggered a wash sale across accounts—a mistake that cost me the current-year deduction and taught me to check every account. The rule applies to IRAs and spouse accounts too, not just taxable.
What can go wrong beyond wash sales? You might accidentally trigger a wash sale by owning the same fund in a 529 plan or by a spouse buying it in a separate account. The IRS treats spouses as one unit for this rule. Also, if you use a robotic advisor, ensure its substitution logic is sound.
Here is a wash-sale avoidance checklist I now use:
- Identify the exact CUSIP or ticker sold and mark a 61-day window (30 before, 1 day of, 30 after).
- Substitute with a different index or fund that is not substantially identical (e.g., S&P 500 ETF vs total market ETF).
- Scan all household accounts, including IRAs and HSA, for reinvestment of dividends or buys.
- Turn off automatic dividend reinvestment on the sold security during the window.
- Document the trade in a spreadsheet with dates and basis adjustments.
Most people don’t realize that even a tiny automatic dividend reinvestment in the same fund inside an IRA can void a large taxable loss. The IRS does not care about intent; it cares about the timeline.
A Decision Tree: When Harvesting Beats Holding
Not every loss should be harvested. Below is a decision framework I use with clients to decide whether selling now beats waiting.
Step 1: Quantify Realized Gain Exposure
If you have embedded short-term gains elsewhere, harvesting short-term losses is a high-value move because both are taxed at ordinary rates. If you only have long-term gains, the benefit is smaller but still real. Map every realized gain you have YTD before acting.
Step 2: Evaluate Ordinary Income Marginal Rate
If you are in the 22% bracket or higher, a $3,000 ordinary offset saves $660+. For low-bracket taxpayers, the immediate save is thin, but carryforward may help later when income rises.
Step 3: Assess Need for Market Exposure
If you believe the asset will rebound, use a correlated substitute to keep exposure. If you are fine exiting, simply hold cash or bonds. Missing a market rally for 31 days is a real cost.
Step 4: Check Wash-Sale Risk
If you cannot avoid a substantially identical repurchase (e.g., employer stock plan), harvesting may be impossible without violating rules. Cross-account coordination is mandatory.
Step 5: Project Carryforward Utility
If you expect large gains next year, harvesting now banks a loss that offsets them at higher rates later. If you might die before using it, note the basis step-up at death can void carryforwards—an honest limitation.
Harvest when the tax savings exceed transaction costs and opportunity cost of being out of the market for 31 days.
Advanced Considerations and Edge Cases
Beyond the basics, several edge cases separate practitioners from dabblers. First, netting order: short-term losses first offset short-term gains, then long-term gains, and vice versa. This ordering can change your effective rate.
Second, state taxes: some states (like California) conform to federal capital loss rules, others have different caps or no carryforward. Always check state code.
Third, mutual fund year-end distributions: a fund may distribute capital gains in December even if you haven’t sold. Harvesting losses before the distribution can offset those gains, but you must account for the fund’s ex-date.
Fourth, the step-up in basis at death: unused carryforwards disappear because the heir’s basis resets to fair market value. This is a trade-off rarely mentioned in rosy summaries.
Fifth, automated harvesting services use algorithms to scan daily. They reduce manual error but may generate excessive trades; review their wash-sale logic across accounts.
Sixth, wash sale across ETFs: two S&P 500 ETFs from different issuers are generally not substantially identical, but two identical index funds from same issuer are. The line is fact-specific and the IRS has not issued bright-line tests for all fund pairs.
Putting It Into Practice: A Real-World Example
Let’s apply the full framework to a realistic 2024 scenario. Maria, a freelance designer, has $8,000 of unrealized loss in a growth ETF, $5,000 of realized long-term gains from selling another fund, and $60,000 of self-employment income. She uses the Tax Loss Harvesting Estimator to map it.
Step 1: She sells the ETF, realizing $8,000 loss. Step 2: The loss wipes the $5,000 gain, leaving $3,000 net loss. Step 3: That $3,000 offsets ordinary income, saving her about $660 at the 22% federal rate (plus possible state save). No carryforward this year. She rotates into a value ETF to avoid wash sale.
If instead her loss were $20,000, after offsetting $5,000 gain she’d have $15,000 left: $3,000 against income, $12,000 carried forward. That carryforward could offset future gains or income at $3,000 per year for four more years. The exact write-off math shows why harvesting large losses in a single year is not a windfall but a multi-year tool.
Common Mistakes and Trade-Offs
Honest limitations: tax-loss harvesting is not a silver bullet. Selling locks in a loss, which may be psychologically hard if the asset rebounds. Transaction fees, though often zero, and bid-ask spreads still cost basis.
A mistake I see: harvesting solely to get a refund without considering future higher brackets. If you expect to be in a much higher bracket later, deferring gains might be better than using losses now at a low rate.
Another trap: ignoring the opportunity cost of the 31-day out-of-market period. In a sharp rally, missing 5% upside can eclipse tax savings. The decision tree above guards against this.
Finally, do not confuse harvesting with tax evasion. The rules are clear, and the IRS expects accurate Schedule D reporting. Use the linked tools and official guidance to stay compliant.
Tax loss harvesting explained with the $3,000 rule and exact math should now feel actionable. The combination of full gain offset, $3,000 ordinary cap, and infinite carryforward is the backbone of every smart year-end plan.