Cross-Market Political Election Arbitrage with Fractional Kelly Sizing

How It Works Political prediction markets generate many related questions: 'Does Candidate A win?', 'Does Party X win the presidency?', 'Does Party X win the Senate?'. The joint probabilities implied by these markets must satisfy logical consistency constraints. When they do not — e.g., the candidate market implies a higher probability than the parent party market allows — a no-arbitrage violation exists. The `nayrbryanGaming/arbiter` bot systematically hunts these violations across Polymarket political markets.

Entry Conditions The bot enforces a strict rule: it refuses to trade without a backtested out-of-sample edge. This is a significant quality gate. Once an edge is confirmed, positions are sized using fractional Kelly criterion (a fraction of the theoretically optimal Kelly bet) to control drawdown. The 'process-first' philosophy documented in the repo description prioritizes capital survival over aggressive profit maximization.

Exit Conditions Not explicitly detailed, but cross-market arb in political markets typically resolves at election outcomes. The fractional Kelly sizing suggests the strategy tolerates holding positions to resolution rather than actively unwinding.

Verified Results & Implementation Notes The repo is paper-trading by default ('survival over profit' framing), indicating the author is still in validation phase. The explicit requirement for out-of-sample backtesting before live trading is a strong methodological signal. Fractional Kelly is an academically grounded position sizing method. Repo size is 180KB — small, suggesting a focused, clean implementation rather than a bloated framework.

Risks Political markets are subject to sudden information shocks (candidate drops out, scandal breaks) that can make previously logical constraints invalid. Related markets may resolve on different timelines, creating basis risk. Liquidity in niche political sub-markets may be insufficient to execute both legs at quoted prices.