Cross-Platform Prediction Market Arbitrage (Kalshi / Polymarket / PredictIt)

How It Works The same real-world event (e.g., 'Will the Fed raise rates in July?') may be listed on multiple prediction market platforms at different prices. If Kalshi prices YES at 0.62 and Polymarket prices YES at 0.55 for an economically identical contract, buying on Polymarket and selling on Kalshi (or equivalent hedged positions) locks in a ~7-cent edge per dollar at risk.

Entry/Exit Conditions The scanner continuously monitors equivalent markets across Kalshi, Polymarket, and PredictIt. A trade is entered when the price differential exceeds a threshold sufficient to cover withdrawal fees, transaction costs, and capital lockup duration. Positions are held to resolution unless early exit is favorable.

Verified Results No verified results are disclosed. The repository appears to be a scanning tool (arbitrage scanner) rather than a fully automated execution system. The Max-Brunelle/market-arb-portfolio repo provides additional architectural documentation for a similar two-platform (Kalshi + Polymarket) implementation, with honest disclosure of what is and is not built.

Implementation Notes Two repos address this strategy: pjmerica/pred-arbitrage (scanner for Kalshi, Polymarket, PredictIt) and Max-Brunelle/market-arb-portfolio (architecture documentation for Kalshi + Polymarket). The PMXT unified API (pmxt-dev/pmxt, 1,896 stars) provides a standardized interface for trading across platforms, significantly reducing implementation complexity for cross-platform strategies.

Risks Market contracts that appear equivalent may have subtly different resolution rules, creating basis risk. Capital must be deployed on multiple platforms simultaneously, increasing counterparty and platform risk. Withdrawal delays (especially from PredictIt) can tie up capital for weeks. Regulatory differences between platforms (Kalshi is CFTC-regulated, Polymarket is offshore) create asymmetric risk profiles. Liquidity may be insufficient on one leg.