Surveillance that stands up: market abuse controls across four regimes

Four regulators — the FCA, the SEC and FINRA, ASIC and Hong Kong's SFC — now ask the same question. Not whether you run market-abuse surveillance, but whether you can prove what it covers, by venue, asset class and channel.

Most surveillance estates grew product by product, so coverage is strong in listed equities and thin in OTC, and cross-product abuse falls between systems that never meet. By common industry benchmark, well over 90% of alerts are false positives, so real risk hides in the noise. The newer failure mode is the surveillance model itself: thresholds tuned years ago, suppressions never documented, and machine-learning triage that has never been validated to the model-governance standards regulators now apply.

The full paper compares the four regimes, sets out the rebuild order that works — risk assessment, evidence-based vendor selection, disciplined implementation, defensible calibration — and illustrates it with a delivered programme that took a global markets group to live ETD and OTC surveillance with an evidenced coverage matrix.