Cross-Market Political Arbitrage with Kelly Sizing
Hunts no-arbitrage violations between related election markets on Polymarket, sizes positions with fractional Kelly, and requires a backtested out-of-sample edge before trading.
This quant bot focuses exclusively on political prediction markets, scanning for arbitrage opportunities between correlated markets (e.g., candidate win vs. party win). It detects no-arbitrage violations where the sum of probabilities across mutually exclusive outcomes diverges from 100%, or where conditional probabilities imply inconsistent pricing. The bot uses a process-first approach, refusing to trade unless a backtested out-of-sample edge is statistically validated.
Position sizing is handled by fractional Kelly criterion to manage risk and maximize long-term growth, with paper-trading as the default mode to ensure survival over profit. Entry is triggered when an arbitrage opportunity is detected and the edge exceeds a minimum threshold. Exit occurs when the mispricing corrects or at market resolution.
The implementation is built in Python using the Polymarket API for order execution and includes a backtesting module with historical data. The developer emphasizes rigorous risk management and statistical rigor over aggressive returns.