The evidence
Prediction market tax software, compared.
Our landing page says no mainstream tax tool knows what a prediction market is. That is a claim about named companies, so here is the whole test behind it: what we put in, what each tool did, and where our own product currently stops.
Last checked September 2026. US federal treatment only.
01: The software you already use
Six tools, one Kalshi fill, zero that recognized it.
The test was the same in each: take a settled Kalshi event contract and try to reach a Section 1256 position without computing it yourself first. None of the six got there, and the reasons differ enough to be worth reading.
None of the six is a bad product. Every one of them is built for wages, a brokerage 1099, or a crypto wallet, and an event contract is none of those. The gap is a category gap, not a quality gap, which is also why waiting for them to close it is a poor plan for this filing season.
02: The tools built for this
We are not the only one, and we will not pretend otherwise.
A handful of products now take event contracts seriously. Here is where each one stops, ours included. The list covers what we could reach and test; it is not a census of everything that exists.
What it computes
Rebuilds per-lot cost basis from raw Kalshi and Polymarket fills, fees included, and computes the 60/40 split contract by contract. Both venues net into one federal position.
Filing and prior years
Compares carrying a loss forward against the three-year carryback under §1212(c), and identifies which prior years are worth amending. Forms and filing arrive in time to file 2026 taxes.
What it computes
Compares four treatment methods on a free P&L, and generates Form 6781 as a PDF on its paid tier. Reads Kalshi by API and Polymarket on-chain.
Filing and prior years
Produces documents to download. It does not file, and we found no amended-return path.
What it computes
Imports a Polymarket wallet and calculates the USDC gains and losses. The Section 1256 question is outside what the category is built to answer.
Filing and prior years
Exports into a filing product. No Section 1256 position, so nothing to amend.
The honest reading of that table: on producing this year’s numbers we are one of several options, and on the years you have already filed we have not found another. Our forms and filing are not live yet, and a tool that already hands you a PDF beats us on that specific job today.
03: Where the money is
Four things a P&L does not answer.
The three-year carryback, priced against the alternative
A net §1256 loss can be carried back three years against prior §1256 gains under §1212(c), and it is an election, not a default. Whether that beats carrying the loss forward depends on what you filed in those years. A tool that hands you this year’s forms has not asked the question.
The years you already filed
The money in a losing year is often sitting in a return that is already in. Getting it out means an amended return, which is a decision about prior years rather than a form for this one.
A December 31 mark that still exists
Section 1256 treats an open position as if you closed it at the final price of the year. Once January arrives, that price is no longer served by either venue, so it has to have been captured at the time. No export downloaded later brings it back.
Both venues, netted
You file one return. Kalshi knows nothing about your Polymarket book and Polymarket knows nothing about your Kalshi book, so a per-venue figure is never the figure that goes on the form.
04: How we modeled the $1,285
What that figure is, and what it is not.
- It is a median, not an average
- A median is the middle case. A handful of very large carrybacks would drag an average upwards and describe nobody, which is exactly the number a marketing page is tempted to print.
- It is modeled on 2023 Kalshi winners, not read off filed returns
- The population is traders who had Section 1256 gains in a prior year, because that is the precondition for the carryback under §1212(c) to be worth anything at all. Nobody in the model filed a return; it prices what the election would have been worth to them.
- It is federal, and it is not your number
- No state treatment anywhere in it. Your own figure depends on your brackets in the carryback years, what else you had in them, and what you actually traded. Connect an account and Realize computes from your own fills instead, which is the only figure that means anything.
The mechanics behind the election, worked through with an example, are in the guide on amending a return for prediction-market losses.
Asked and answered
- Does TurboTax handle Kalshi or Polymarket?
- Not as a recognized instrument. There is no prediction-market flow, so Section 1256 treatment means finding the Contracts and Straddles interview and entering a figure you computed elsewhere. The paths people reach instead, a self-entered 8949 or Miscellaneous Income, produce a different tax answer on the same trades.
- Can crypto tax software do my Polymarket taxes?
- It can account for the USDC movement, which is not the same question. Crypto tax tools treat what they find as a property disposal rather than as a Section 1256 contract, and they do not read Kalshi, so a trader on both venues has no way to net them into the one figure a return needs.
- Is Realize the only prediction-market tax tool?
- No. At least one other tool is built for event contracts and generates Form 6781 as a download. What we have not found elsewhere is the part that reaches backwards: comparing the three-year carryback of a net Section 1256 loss under §1212(c) against carrying it forward, and identifying which already-filed years are worth amending.
- Which treatment is correct for an event contract?
- Nobody can tell you that yet. There is no IRS guidance and no court ruling on whether prediction-market event contracts qualify for Section 1256. We compute the position under each reading and show the work behind every figure so that you and your preparer can decide which one you are willing to sign.
Run it against your own trades.
Connect Kalshi or Polymarket with a read-only key and Realize rebuilds the position from your real fills, every figure traceable back to the trade behind it. Tracking is free.
Primary sources
The federal authorities behind the figures on this page. Realize is not a tax advisor, so these are the law itself rather than our summary of it.
- IRC §1256 The Section 1256 regime: the 60/40 long-term / short-term split and the year-end mark to market
- IRC §1211(b) The $3,000 annual cap on net capital losses against ordinary income ($1,500 married filing separately)
- IRC §1212(c) The elective three-year carryback of a net Section 1256 loss against prior-year Section 1256 gains
- IRS Form 6781 Where Section 1256 gains and losses are reported, and where the carryback election is made
- IRS Form 1040-X The amended individual return, which is how a prior year is corrected