Part I: Machine learning against financial crime
The use of machine learning in anti-money laundering – especially transaction monitoring – is catching the eye of financial executives the world over. Read on to know more:
Banks today are investing in multiples of billions of dollars annually to fight increasingly sophisticated techniques of money laundering and other financial crimes being perpetrated round the globe. In 2020, for example, financial institutions spent an estimated $214 billion on financial crime compliance. What’s more, as consulting giants McKinsey & Co. writes,
“The resulting regulatory fines related to compliance are surging year-over-year as regulators impose tougher penalties. But banks’ traditional rule- and scenario-based approaches to fighting financial crimes have always seemed a step behind the bad guys, making the fight against money laundering an ongoing challenge for compliance, monitoring, and risk organisations.”
This is where machine learning comes in.
Recent advancements in machine learning have helped banks improve their anti-money laundering schemes considerably. This has stemmed from increasing research and investment in the transaction monitoring aspect of these programs, as well as backing from regulators and governing bodies. Legislature in the United States such as the Anti-Money Laundering Act of 2020 and the subsequent National Illicit Finance Strategy are promoting financial institutions to test and adopt innovative approaches in fighting financial crimes whilst reducing obstacles from guidance, examination practices and regulations.
“This momentum in the fight against financial crimes is creating keen interest in ML among industry leaders. Earlier this year, McKinsey invited the heads of anti–money laundering and financial crime from 14 major North American banks to discuss adopting ML solutions in transaction monitoring. More than 80 percent of the participants had begun the process of adopting ML solutions, with most expecting to dedicate serious efforts to implementing ML solutions within their AML programs in the next two to three years.”
When it comes to application, machine learning, in theory, can be used across the entirety of the AML value chain. The true value of ML in AML, however, according to the experts at McKinsey, can be found in combining machine learning with other advanced algorithms such as gradient boosting, random forest or deep learning in the transaction monitoring aspect of banks’ AML efforts.
Financial institutions today use rule- and scenario-based approaches and statistical tools for transaction monitoring. McKinsey writes, “rules and thresholds are driven primarily by industry red flags, basic statistical indicators, and expert judgement. But the rules often fail to capture the latest trends in money-laundering behaviour.
Machine learning models, on the other hand, leverage more granular, behaviour-indicative data to build sophisticated algorithms. They are also more flexible in quickly adjusting to new trends and continually improving over time. By replacing rule- and scenario-based tools with ML models, one leading financial institution improved suspicious activity identification by up to 40 percent and efficiency by up to 30 percent.”
Machine learning proves to be most advantageous in AML in aspects where there is a high degree of freedom in choosing data attributes, as well as sufficient availability of quality data (such as in scenarios where there is rapid movement of funds and a large number of attributes to consider). ML is also rather useful in identifying the dynamics and relationships between risk factors.
[Continue in reading in Part II]
Know more about the syllabus and placement record of our Top Ranked Data Science Course in Kolkata, Data Science course in Bangalore, Data Science course in Hyderabad, and Data Science course in Chennai.