Transaction Monitoring Tuning & Optimization

Project Overview

A large international bank was under review by the regulator who identified many issues related to the client’s Actimize transaction monitoring system for multiple workstreams.

Challenges

• The client needed assistance in assessing whether the issues identified had been addressed properly and conformed to AML/BSA regulatory guidelines.
• He was requested to establish new tuning and optimization methodology across all rules and segments.
• Finally, the client wanted to measure the risk and impact of these issues through an analysis of historical data.

Project Solution

Matrix-IFS worked closely with the Global Head of AML to complete a tactical review of its SAM transaction monitoring system in order to develop risk-based suspicious activity monitoring coverage, data validation testing, AML model testing, segmentation, and scenario tuning.

• A multi-disciplined team was deployed in statistics, data science, and regulation to conduct focused data quality checks on 1.7 million records of transaction data in order to uncover and test a “golden set” of data to be used for further analysis.
• An exploratory data analytics was conducted with the use of data visualization tools and statistics to discover a hidden pattern in the data unknown to the client.
• Using Matrix-IFS’ data visualization approach, a hidden “network” of relationships among accounts, scenario thresholds, and alerts was uncovered.
• An overall tuning methodology was developed in addition to SAM tuning methodology per each and every active rule.
• Matrix-IFS provided a menu of possible threshold changes, all fundamentally defensible and supported by statistical analysis.
• Matrix-IFS transitioned the gains from this analysis into the next steps to create an end-to-end risk-based Transaction Monitoring Model.