Learn / Strategy
Kalshi–Polymarket arbitrage, honestly.
Short answer
Cross-platform arbitrage means buying Yes on one venue and No on another for less than $1 combined. It only works if both contracts settle on exactly the same outcome and the gap survives fees, spread, size and the time your capital is locked up. Most visible gaps do not.
How cross-platform arbitrage is supposed to work
If two platforms list the same event, one contract must pay $1. Buy Yes on the venue where Yes is cheap and No on the venue where No is cheap. If the two prices add up to less than $1, you collect $1 no matter what happens and keep the difference.
That is the whole idea, and it is why “risk-free Kalshi Polymarket arbitrage” videos and bots are everywhere. The trouble is in three words: the same event, less than $1 and no matter what.
A worked example: 3 cents becomes 1 cent
Suppose a game is listed on both platforms with identical rules. Kalshi’s best Yes ask is 46 cents and Polymarket’s best No ask is 51 cents. On screen, that is 97 cents for a guaranteed dollar: a 3-cent gap. Now add each platform’s published taker fee (Kalshi at a 0.07 rate rounded up to the cent; Polymarket’s sports rate of 0.05; see fees explained).
| Leg | Price | Cost | Fee | Total |
|---|---|---|---|---|
| Kalshi Yes | 46¢ | $46.00 | $1.74 | $47.74 |
| Polymarket No | 51¢ | $51.00 | $1.25 | $52.25 |
| Both legs | 97¢ | $97.00 | $2.99 | $99.99 |
Guaranteed payout: $100. Profit: about one cent on $99.99 of capital, before deposit, withdrawal or currency conversion costs. Near 50 cents, fees on two legs absorb roughly 3 cents per pair, so a visible 3-cent gap is close to break-even.
Wider gaps can survive. At 45 and 50 cents the same fees leave $2.01 on $97.99, about 2.05%. If the event settles in 90 days, that is roughly 8.3% annualized, with your capital split across two platforms the entire time.
Risk 1: the contracts are not the same
This is how “risk-free” trades lose on both legs. Similar titles can hide different rules:
- Overtime and ties. Imagine one market settles on the regulation-time result and the other includes overtime. You buy Yes on “Team A wins in regulation” and No on “Team A wins.” Team A wins in overtime. Both legs lose.
- Deadlines and time zones. “By December 31” can mean different cut-offs on different venues.
- Resolution source. An official announcement, a named data provider and media reports can disagree, especially on close calls.
- Cancellation and postponement. One venue may void a postponed game while the other waits for a rescheduled one.
Read both rule pages line by line. Our guide to comparing Kalshi and Polymarket prices lists what to match first.
Risk 2: you cannot fill both legs at the shown price
The best ask may only be good for a handful of contracts. Larger orders walk up the book and the gap shrinks. Worse, you usually fill one leg first. If the price moves before the second fills, you are holding an ordinary directional position. This is called legging risk, and it is why arbitrageurs size to the thinner side of the book.
Risk 3: capital, access and settlement time
- Money is locked in both platforms until settlement and cannot be redeployed.
- Moving money between a USD exchange and a USDC-based venue has costs and delays.
- Account eligibility and market availability depend on the platform and your location.
- Disputed or delayed resolutions keep capital stuck longer than planned.
An arbitrage checklist
- Match rules: outcome, deadline, resolution source, ties, overtime, cancellation.
- Use executable asks on both sides, not last trades or midpoints, with timestamps from the same moment.
- Check depth on both books; size to the smaller.
- Subtract both platforms’ current fees, plus transfer costs.
- Annualize the remaining return over the time to settlement.
- Decide what you do if only one leg fills.
If the trade still looks good after all six steps, it might be real. Most screenshots of “free money” fail at step 1 or step 4.
Price gaps are research leads
A persistent gap that does not survive costs can still be informative: it may tell you which venue reacts faster or where liquidity sits. Kelix Lines shows available prices for the same sports event side by side with timestamps and labels stale or missing quotes. It does not claim a gap is an arbitrage. If you want to automate the scanning, read how to build a prediction market bot first.
Frequently asked questions
Is Kalshi–Polymarket arbitrage risk-free?
No. It is only riskless if both contracts resolve identically. Different resolution sources, deadlines or tie rules can make both legs lose. Execution risk, fees and platform access add further risk.
How big does a gap need to be?
Larger than the combined fees on both legs plus spread and slippage. Near 50 cents, fees alone can absorb about 3 cents per pair, so a 3-cent visible gap can be close to zero after costs.
Why does capital lock-up matter?
Your money sits in both legs until settlement. A 1% profit locked for six months is a small annualized return and carries platform and counterparty exposure the whole time.
Can a bot do this automatically?
A bot can scan prices, but matching contracts still needs rule-by-rule review. Automatic matching on similar titles is the most common way “arbitrage” bots lose money.
Educational content, not financial, legal or tax advice. Prediction markets involve risk, including loss of capital. Examples are illustrative calculations, not live quotes or trading results. Kelix is independent of Kalshi, Polymarket, Robinhood and sportsbooks.