Polymarket vs Kalshi: fees, access & the price gaps between them

Polymarket and Kalshi are the two largest prediction markets — one an offshore crypto order book, the other a CFTC-regulated US exchange. They list many of the same events at different prices, and that difference is where cross-venue arbitrage lives.

Why does the same event trade at different prices on Polymarket and Kalshi?

The two venues cannot share liquidity: different countries, different money (USDC vs USD), different users. A political headline can move Polymarket's crypto-native crowd minutes before Kalshi's US retail follows — or the reverse. With no common market maker forcing convergence, gaps of 1–5 cents on the same question are routine.

The gaps persist for structural reasons rather than because nobody has noticed them. Capital cannot move freely between the venues: separate KYC, separate funding rails, and a fiat-to-stablecoin conversion in between. Money committed to one leg is locked until that market resolves, so an arbitrageur cannot recycle capital quickly enough to close every gap. Academic work on this has reached the same conclusion — the mispricings are persistent, not fleeting, precisely because the barriers to closing them are real.

How do you arbitrage the gap between Polymarket and Kalshi?

When Polymarket prices YES at $0.47 and Kalshi prices NO at $0.505 for the same event, buying both costs $0.975 against a certain $1.00 payout — a 2.5% gross edge whichever way the event resolves, if both markets resolve on the same rules.

Two adjustments turn that into a real number. Kalshi's taker fee on a 50c contract is about 2c per contract, which is most of a 2.5% edge on its own — fees are decisive, not cosmetic. And the quoted price holds only for the size resting at the top of the book; filling a large order walks you into worse levels. Always price the hedge net of fees and at the size you actually intend to trade.

Can one person trade both Polymarket and Kalshi?

Generally no, and this is the catch that makes most Polymarket-vs-Kalshi spreads unusable. Kalshi is a CFTC-regulated US exchange open to US persons; Polymarket restricts them. A hedge spanning the two therefore needs accounts in two jurisdictions — not one person with two tabs open.

The pairs a single account can actually place are Kalshi + PredictIt if you are in the US, or Polymarket + Limitless if you are not. Using a VPN to open a restricted account does not solve this; it breaches the venue's terms and the consequence usually arrives at withdrawal. This is why every PROVISUM signal is labelled with the jurisdictions it requires before you look at its return.

Which is better for arbitrage, Polymarket or Kalshi?

Neither, individually — the pair is the product. But their characters differ in ways that matter to a hedger. Polymarket lists a wider set of questions, including long-tail crypto and culture markets that have no Kalshi equivalent, so many of its markets have no cross-venue partner at all. Kalshi's strength is standardised, well-defined contracts — economic data, weather, sports — where the resolution source is explicit, which makes it far easier to confirm that two markets truly ask the same question.

That last point is the practical one. The most common way to lose money here is not a bad price; it is two markets whose wording matches while their resolution rules do not. Kalshi's contract specifications make that check quicker to perform.

Where are they disagreeing right now?

For most people the choice of venue is settled by regulation before preference gets a say. For arbitrage specifically, the interesting question is not which venue is better; it is where they disagree today — and that changes every hour. PROVISUM re-prices both venues every 10 minutes, checks the order books at size, and flags the pairs whose resolution rules deserve a second look before you act.

Keep reading: Crypto arbitrage: real cross-exchange spreads, measured · Prediction market arbitrage: the complete guide · Polymarket arbitrage: finding mispriced markets · Oracle risk: when both legs of a hedge lose · Liquidity: why the quoted edge is not the real one