prediction-markets-arb

This repository is a BSc thesis project from the University of Ljubljana, providing an academic foundation for arbitrage detection across multiple prediction markets. It goes beyond simple cross-market price comparisons into combinatorial arbitrage—which spans multi-outcome and correlated events to find arbitrage opportunities that basic scanners miss. The TypeScript implementation suggests structured, engineerable code. The academic background ensures comprehensive mathematical grounding, likely including linear programming optimizations for position sizing. Topics include Kalshi and Polymarket, focusing on capitalizing on price discrepancies between (and within) platforms. This is especially valuable for traders seeking complexity beyond BTC 5-minute bets, potentially covering political, tech, or economic event spheres. The linear-programming keyword hints at sophisticated optimization for maximising risk-free profit. Cloning this could provide a springboard for developing a comprehensive arbitrage strategy engine that spans multiple correlated markets, a sophisticated yet under-served niche.