Arbitrage Analysis in Polymarket NBA Markets

Battle Count: 83

Highly relevant to quantitative trading in prediction markets and market microstructure research. The paper provides rigorous empirical evidence on arbitrage frequency, duration, and executability constraints in a decentralized CLOB environment. Key findings on liquidity-bounded extraction (76.9% of opportunities constrained to ~14.8 shares) directly inform algorithmic execution strategies. The limits-to-arbitrage framework application and the distinction between theoretical and practical profitability are directly applicable to quantitative strategy design. However, the retail-scale constraint limits institutional applicability.

Moderate complexity. Requires: (1) high-frequency API polling infrastructure with 3.6-5.5 second cycles across hundreds of markets, (2) state reconstruction with forward-fill imputation and deduplication for mirrored liquidity, (3) unified state machine for combinatorial analysis across asynchronous order books, (4) strict execution constraint filtering, and (5) episode grouping with trust ceilings. The core arbitrage detection logic is mathematically straightforward (price sum comparisons), but the data engineering pipeline for continuous market state reconstruction is non-trivial.

The paper provides detailed methodology descriptions including API endpoints, polling intervals, state alignment procedures, and execution constraints. However, no code repository or processed dataset is explicitly mentioned as available. The data collection relies on Polymarket's public CLOB API, but the 75 million snapshot dataset and processing pipeline are not stated to be publicly released. Reproduction would require significant data engineering effort to replicate the polling infrastructure and state reconstruction logic.