Statement of Interest
Over the past six years, NASD’s Market Regulation department has built and operates two major “break detection” systems – the Advanced Detection System (ADS) and Securities Observation, News Analysis, and Regulation (SONAR) – for surveillance of the Nasdaq and several other markets. These systems rely for their effectiveness on the detection of instances of scenarios of regulatory interest – episodes in market activity where some violation may have occurred – many of which comprise relationships among transactions, market participants, securities, issuers, and other subject entities. Break Detection Systems are powerful tools for detecting errors, violations, or other anomalous conditions and activities. However, they are limited to the immediate activities which they find in the input data stream. Background knowledge, aggregation of detection over a priori identifiers (brokers, issuers, etc.) can start to draw a picture of an underlying intentional pattern and practice. Without powerful but tractable models of populations of breaks, we are limited to counts and percentages as a decision tool for investigators. As target activities become more complex and varied, and as the cost of regulation continues to rise, NASD feels increased need for such models to cull and derive the greater benefit from its break detection systems.