No-Arbitrage Political Cross-Market Arbitrage
Process-first quant bot that hunts no-arbitrage violations between related Polymarket election markets and sizes positions with fractional Kelly.
This bot, named 'arbiter', focuses exclusively on Polymarket political prediction markets. It systematically scans for no-arbitrage violations between related markets — for example, if the sum of probabilities for mutually exclusive election outcomes exceeds 1 (or falls below 1), that indicates a pricing inconsistency. The bot calculates the implied probabilities and identifies the arbitrage. It sizes positions using a fractional Kelly criterion to manage risk and maximize long-term growth, but only trades if a backtested out-of-sample edge is confirmed. By default it operates in paper-trading mode, emphasizing survival over profit. The codebase is process-first, meaning it logs every decision and refusal to trade. The implementation is in Python and includes a test suite for validating the no-arbitrage detection logic.