Measure for measure - how insight from analytics is central to a strong transaction monitoring system

With trade and commercial transactions continuously increasing at an exponential rate across the globe, appropriate anti-financial crime detection is proving ever more challenging for financial institutions. Consequently, financial institutions more than ever need to ensure their existing transaction monitoring capabilities are accurate, effective and efficient, so as to maintain Financial Crime Risk protection.

Gone are the days when an analyst would sit and review all transactions of a particular type, no more do institutions rely exclusively on the experience of their teams to capture suspicious transactions. Instead come an abundance of providers offering products and approaches to help firms achieve AML regulatory compliance and resiliency against Financial Crimes.

Analytics has proven to be a fundamental tool to support manual review of transactions. We must ask then, how much can analytics truly help automate and optimise transaction monitoring (TM)?

Fundamentally, for TM analytics to work, there has to be an agreed methodology, approach and explain-ability. There need to be rules, algorithms and scoring systems on which the analytics can be triggered. The TM analytics need to be built based on accurate and dynamic data sets and must be open to fine tuning in line with new regulations, learnings, patterns and behaviours. One way to iteratively improve TM analytics is the introduction of network and behavioural analytics to drive the identification of trends and patterns in large data sets.

Additional to a robust methodology and approach, is quality. A centralised data lake that is both extensive and well managed is critical. The integrity of the data lake is immediately compromised if there are doubts about the data quality or completeness and so the governance around the data lake goes hand in hand with its existence. Data owners need to be agreed and made accountable for the quality, maintenance and any associated cleansing. Competent and knowledgeable data analysts and scientists with a deep understanding of Financial Crime strategies can drive higher quality data management and the subsequent TM analytics.

So, it’s all about data, I leave you with a relevant reoccurring question, could machine learning replace traditional rule-based transaction monitoring systems?

Accenture Digital Risk and Compliance have supported clients in TM analytics, TM scenario development and optimisation, as well as data engineering and predictive model building.

My thanks to Dahlia Belloul, Management Consultant, Accenture, for contributing this article.

?Disclaimer: This content is provided for general information purposes and is not intended to be used in place of consultation with our professional advisors. Copyright ? 2021 Accenture. All rights reserved. Accenture and its logo are registered trademarks of Accenture