Bellman–Ford negative-cycle arbitrage detection over live FX and crypto order books — and a rigorous accounting of why almost every cycle it finds is not a trade. Certificates of absence, analytic quantisation-error bounds, and 97% warm-start reuse on live data.
python graph-algorithms numpy websockets forex cryptocurrency quantitative-finance algorithmic-trading arbitrage numerical-analysis shortest-path market-microstructure bellman-ford differential-testing fastapi negative-cycle-detection
-
Updated
Aug 6, 2026 - Python