The Excess Business Loss Limitation for 2026: Why OBBBA Just Lowered the Cap on Deducting Business Losses.
OBBBA made the excess business loss limitation permanent and reset its threshold, so for 2026 a noncorporate taxpayer can deduct only $512,000 of net business loss on a joint return against nonbusiness income, down from $626,000 in 2025. Everything past the cap becomes an NOL carryforward usable against no more than 80% of future income. Here is the §461(l) math and why it bites the cost-segregation and short-term-rental crowd hardest.
A business owner finishes a strong year, then places a heavily depreciated asset in service, a cost-segregated building or a round of equipment bonused under §168(k), and expects the resulting loss to erase a seven-figure income. Then the software flags a number nobody planned for. The excess business loss limitation for 2026 caps how much of a net business loss a noncorporate taxpayer can deduct against nonbusiness income at $512,000 on a joint return, and every dollar past that is disallowed for the year. The One Big Beautiful Bill Act made this limit permanent and, in the same stroke, lowered it. Here is how §461(l) works in 2026 and why the cap is smaller than it was last year.
Start with where the rule sits. A loss from a pass-through business or a rental has to clear four gates before it reaches your return: your basis in the activity, the at-risk rules of §465, the passive activity loss rules of §469, and finally §461(l). The first three decide whether a loss is allowed at all. Section 461(l) is the last gate, and it does something different: it takes the losses that already survived the other three, adds them up, and asks whether the total is simply too large to deduct against your wages, interest, dividends, and capital gains this year. Above the threshold, the answer is no.
Why the excess business loss limitation for 2026 is lower than 2025.
The limit was a Tax Cuts and Jobs Act creation, scheduled to expire after 2028. The One Big Beautiful Bill Act, signed July 4, 2025, removed the sunset and made §461(l) permanent. It also re-based the inflation adjustment back to the original $250,000 and $500,000 starting points, which stripped out the inflation that had compounded since 2018. The result is a threshold that actually went down. For 2025 the cap was $313,000 for most filers and $626,000 for joint returns. For 2026, Rev. Proc. 2025-32 sets it at $256,000 and $512,000. A joint filer lost $114,000 of headroom from one year to the next, which is the rare case of an inflation-indexed number moving the wrong way.
How the cap is computed on Form 461.
An excess business loss is the amount by which your total deductions from trades or businesses exceed your total gross income and gain from those businesses, plus the threshold. You figure it on Form 461, and two adjustments matter. The deductions are counted without regard to any net operating loss under §172 or the qualified business income deduction under §199A, so a prior-year NOL does not stack into this year's calculation. And capital gains and losses get special handling: business capital losses are left out of the deduction side entirely, while business capital gains count as income only up to your total capital gain net income. The losses that count are from your trades or businesses, not from being an employee. Section 461(l)(3)(B) leaves wages from performing services as an employee out of the business side of the computation.
Whatever clears the cap is deductible now against any income. Whatever does not is disallowed for the year. The threshold is per taxpayer, not per business and not per spouse, so a married couple running two separate ventures still shares one $512,000 ceiling. Here is what that looks like in a big-loss year.
- Net loss from trades and businesses (after basis, §465, §469)
- −$1,200,000
- Business gross income and gain to net against
- $0
- Excess business loss threshold, MFJ 2026 (Rev. Proc. 2025-32)
- $512,000
- Business loss deductible this year (capped at the threshold)
- $512,000
- Excess business loss disallowed for 2026
- $688,000
- Carried to 2027 as an NOL (offsets ≤ 80% of taxable income)
- $688,000
- Amount disallowed under the 2025 threshold ($626,000) instead
- $574,000
- Extra loss trapped by the 2026 reset
- $114,000
- Federal + Utah tax deferred on the disallowed loss at 41.5%
- ≈ $285,520
Tax year 2026, married filing jointly. Assumes the $1,200,000 is a net nonpassive business loss that has already cleared the basis, §465 at-risk, and §469 passive limits, with no offsetting business capital gains. Threshold per Rev. Proc. 2025-32; $626,000 was the 2025 MFJ figure. The $688,000 disallowed amount becomes a net operating loss under §461(l)(2) and §172(b), usable against no more than 80% of taxable income under §172(a)(2). The 41.5% rate is a 37% federal marginal bracket plus Utah's 4.5% individual rate (H.B. 106, effective tax year 2025). This is a deferral, not a permanent loss of the deduction.
The number to sit with is the $688,000. The couple did not lose that deduction, but they cannot use it this year, exactly the year they generated it to offset a high income. They get $512,000 against their wages and portfolio income now, and the rest waits. Run the same $1,200,000 loss against last year's $626,000 cap and only $574,000 would have been disallowed. The 2026 reset alone pushed an extra $114,000 of deduction into the future, worth roughly $47,000 of tax deferred at a 41.5% combined rate.
The disallowed loss is not gone, it is deferred.
Section 461(l)(2) treats the disallowed amount as a net operating loss carryover to the following year under §172. That is the saving grace and the catch at once. The deduction survives, but it lands in the post-2017 NOL regime, where a carryforward can offset no more than 80% of taxable income in any later year under §172(a)(2). So a large disallowed loss can take several years to fully absorb, and in each of those years 20% of your income stays taxable no matter how big the carryforward is. A loss you expected to use against a 37% bracket this year may come back against a lower bracket later, or trickle out so slowly that the time value of the deferral is the real cost. The deduction is not lost. Its timing and its rate are.
Three situations where the cap surprises people.
- ◆The cost-seg or short-term-rental year. The strategies that generate six- and seven-figure first-year losses are precisely the ones that run into this cap. A cost segregation study with 100% bonus depreciation, or the short-term rental loophole that frees a large loss against W-2 income, routinely throws off more than $512,000. The §469 work gets the loss to non-passive, and then §461(l) caps how much of it lands this year.
- ◆A sale or a genuinely bad year stacked with capital losses. Operating losses plus a business that sold assets at a loss feel like one big deduction, but business capital losses are excluded from the §461(l) computation and run through their own $3,000-a-year capital loss limit instead. The two regimes do not combine, and taxpayers routinely overestimate how much will be deductible.
- ◆Two spouses, two businesses, one cap. The joint threshold is double the single amount, not double per business or per spouse. A couple where each runs a separate venture, or where one has an active loss and the other has a profitable one, still measures the aggregate against a single $512,000 ceiling.
The planning is about controlling the size and timing of the loss before year end, not fixing it on the return. Electing out of bonus depreciation under §168(k)(7) for a class of property, or using §179 with its taxable-income limit instead, can keep a deduction from overshooting the cap in a single year. Staggering placed-in-service dates across two tax years spreads the loss across two thresholds. Recognizing business income or a Roth conversion in the same year raises the income the loss is measured against. These are the same levers behind the real estate strategies high earners use, just read in reverse: the goal there is to maximize the current-year loss, and §461(l) is the reason the biggest possible loss is not always the most valuable one.