Cross-Market No-Arbitrage Violation (Political Markets)

How It Works Related prediction markets must satisfy no-arbitrage constraints. For example, the probability of Candidate A winning a primary plus the probability of Candidate B winning must not exceed 1.0 when they are mutually exclusive outcomes. When these constraints are violated — prices are inconsistent across related markets — a risk-free or near-risk-free profit can be locked in by trading both sides simultaneously.

Entry/Exit Conditions The scanner continuously checks sets of related political markets for price inconsistencies. A trade is entered when the sum of 'YES' prices for mutually exclusive outcomes exceeds 1 (or the sum of 'NO' prices falls below 0) by enough to cover fees and slippage. The position is exited at resolution or when prices converge to no-arbitrage bounds.

Verified Results The repository emphasizes a 'backtested out-of-sample edge' as a prerequisite for any live trade. The bot defaults to paper-trading mode, indicating the edge has not yet been publicly verified on live capital. The description explicitly states 'refuses to trade without a backtested out-of-sample edge,' which is a strong process discipline signal.

Implementation Notes Position sizing uses fractional Kelly criterion, which is a mathematically grounded approach to capital allocation that accounts for edge and variance. The system is described as 'process-first,' prioritizing risk management and survival. Built specifically for Polymarket political markets.

Risks No-arbitrage gaps in prediction markets can be illusory due to differing resolution rules, timing mismatches between related markets, or liquidity constraints preventing full position execution. Political market correlations can shift rapidly. Smart contract or platform risk on Polymarket (USDC, on-chain settlement) adds tail risk.