The 2026 Gambling Loss Cap: How OBBBA Can Tax You on Money You Never Made

If a prediction-market platform, sportsbook, or casino classifies your activity as gambling for tax purposes, a rule change starting in the 2026 tax year makes that treatment meaningfully worse than most people realize. It’s worse than simply “losses aren’t fully deductible.” Here’s the exact mechanism, with real numbers.

The rule before 2026

Under Section 165(d), gambling losses have always been deductible only up to the amount of your gambling winnings, and only if you itemize deductions on Schedule A. If you take the standard deduction, which is roughly 90% of taxpayers, gambling losses aren’t deductible at all, meaning you’re taxed on your gross winnings with no offset. That part hasn’t changed. What changed is what happens even if you do itemize.

What OBBBA changes, starting 2026

The One Big Beautiful Bill Act adds a new cap: itemizers can now deduct only 90% of their gambling losses, not 100%, even when losses don’t exceed winnings. That 10% haircut sounds small until you run the numbers on a breakeven year.

The math that surprises people

Say you won $100,000 gambling in 2026 and lost $100,000 gambling in 2026, a wash economically. Here’s what happens under the new cap, itemizing:

• Gross winnings (taxable income): $100,000
• Loss pool after the 90% cap: 90% × $100,000 = $90,000
• Deductible losses: the lesser of that pool and your winnings, so $90,000
• Taxable income: $100,000 − $90,000 = $10,000

You broke even. You still owe tax on $10,000 of income that never existed: the mechanical result of the 90% cap applied to anyone who itemizes and has losses close to their winnings, which describes a lot of active traders in a volatile year.

How the 90% and the winnings cap stack

There was a real question about how the two limits interact: do you take 90% of losses first and then cap at winnings, or cap at winnings first and then take 90%? They give the same answer when your losses are at or below your winnings, but they diverge once losses run higher. The IRS proposed regulations (REG-113229-25) settle it: apply the 90% to your losses first, then cap the deduction at your winnings. Those regulations are proposed rather than final, so it could still shift, but that is the order WagerMints uses, and it shows the arithmetic so you can check it.

Why this makes gambling treatment the one to avoid, if you have a choice

Compare this to the other three methods covered in our overview post. Under capital gains or ordinary income treatment, that same breakeven year owes tax on approximately $0: gains and losses just net against each other, no cap, no haircut. Under gambling treatment with the 2026 cap, the exact same trading activity produces $10,000 of phantom taxable income. If your activity could reasonably be characterized more than one way, and for most prediction-market trading it genuinely can, this is the gap that makes checking all four methods worth the ten minutes it takes, rather than defaulting to whatever a platform’s tax summary happens to assume.

If you don’t itemize, it’s worse, not better

None of this is a new discount: it’s a cap on a deduction you get only by itemizing in the first place. If you take the standard deduction, gambling losses remain fully non-deductible, cap or no cap, and you’re taxed on 100% of your gross winnings regardless of how much you lost. The 90% cap only makes a bad situation for itemizers a little closer to the situation non-itemizers were already in.