Abstract:A new partnership between BMLL and Features Analytics aims to overhaul market abuse detection by integrating deep historical order book data to reduce surveillance false positives.

BMLL, an independent provider of historical Level 3 data, has formed a strategic alliance with AI surveillance firm Features Analytics to tackle inefficiencies in market abuse detection. The collaboration addresses the overwhelming volume of false positive alerts generated by legacy rule-based systems.
Financial institutions currently face significant operational drag from surveillance systems that flag non-suspicious activity. Features Analytics claims its “eyeDES” technology can reduce these false positives by 90%, a figure that becomes actionable when paired with BMLLs granular order book reconstruction capabilities.
While the initial focus includes global equities and ETFs, the technology has implications for the broader financial ecosystem, including Foreign Exchange (FX). The move mirrors recent trends; the London Stock Exchange Group (LSEG) recently launched a similar cross-venue platform targeting FX and MiFID instruments.