Cross-Market Political Election Arbitrage (No-Arbitrage Violations)

How It Works In political prediction markets, related markets must satisfy logical constraints. For example, the probability of a candidate winning the presidency must be less than or equal to their probability of winning their party's primary. When these constraints are violated across separate but related markets, a near-arbitrage position can be constructed by trading the logically inconsistent spread.

Entry/Exit Conditions The bot scans Polymarket political markets for no-arbitrage violations — cases where a combination of positions across related markets guarantees a profit regardless of outcome. Positions are sized using fractional Kelly criterion based on estimated edge magnitude. The strategy requires a backtested out-of-sample edge before any live trading is permitted, with paper trading as the default mode.

Verified Results No live trading results are disclosed. The repository description explicitly states 'refuses to trade without a backtested out-of-sample edge' and defaults to paper trading, indicating a disciplined but unverified approach. The 'survival over profit' philosophy suggests conservative position sizing.

Implementation Notes Built as a process-first quant system. Fractional Kelly sizing is implemented to manage drawdown risk. The system requires a market relationship graph to map logical dependencies between markets (e.g., primary → general election chains). Political markets have low liquidity relative to crypto markets, so position sizing must account for market impact.

Risks Political market relationships can break down in unexpected scenarios (e.g., a candidate winning a primary but being disqualified). Liquidity is often thin in political markets, making large positions impractical. The strategy depends on correctly modeling market dependencies, which requires domain knowledge. Resolution disputes or rule changes by Polymarket can invalidate assumed relationships.