Cross-Platform Binary Arbitrage (Polymarket ↔ Kalshi)

How It Works Polymarket and Kalshi frequently list markets on the same real-world events (elections, economic data releases, Fed decisions). Because the platforms have different user bases, liquidity profiles, and fee structures, the implied probabilities for identical outcomes diverge. The strategy buys YES (or NO) on whichever platform prices the outcome lower and simultaneously sells the equivalent position on the platform pricing it higher, creating a locked spread.

Entry Conditions A trade fires when the cross-platform spread exceeds the combined fee load (Polymarket CLOB fees + Kalshi fee structure + potential gas costs). Multiple implementations exist: `Sourav19o7/polymarket-kalshi-arb` is a detector + execution assistant; `Steeve-Crypto/Quant-CLI` is a full quant trading framework; `pjmerica/pred-arbitrage` extends the scan to PredictIt as well, creating a three-platform scanner.

Exit Conditions Positions resolve at market settlement. Alternatively, if the spread compresses before resolution, both legs can be unwound for a realized gain. Active unwinding requires monitoring both platforms continuously.

Verified Results & Implementation Notes Three separate repositories implement variants of this strategy, suggesting the opportunity is real and known. `pjmerica/pred-arbitrage` also scans PredictIt, which has slower retail pricing, potentially offering wider and more persistent spreads. The `Sourav19o7` repo (64KB, created June 19 2026) is very small and likely a thin detection layer rather than a full execution system. `Steeve-Crypto/Quant-CLI` at 1857KB is more substantial.

Risks Market question wording differences between platforms can mean markets are not truly equivalent, creating basis risk at resolution. Kalshi operates under CFTC oversight, introducing regulatory asymmetry. Capital must be deployed on both platforms simultaneously, reducing capital efficiency. Withdrawal/deposit delays mean capital cannot be redeployed quickly.