The results
500,000+
fewer alerts per year in some countries
Immediate
results from analysis of the bank's historic data
Approved
by the Central Bank and other global regulators
Key takeaways
- A global financial institution with ~$2T in assets and over 3,000 branches across Europe, used ThetaRay to modernize its numerous rules-based AML systems.
- In some countries where the bank operates, the number of alerts declined by over 500,000 per year after ThetaRay established normal behavior from the data.
- ThetaRay found "unknown unknowns": risks and behavior that the bank's legacy systems had not previously identified.
- By introducing more robust and dynamic transaction monitoring across its global framework, the bank gained approval from its Central Bank and other global regulators.
Why did the bank look beyond rules-based AML?
In a proactive initiative, a leading global financial institution (~$2T in assets) with over 3,000 branches across Europe, set out to modernize its numerous rules-based anti-money laundering (AML) systems. Recognizing the limitations of its legacy solutions, the transaction monitoring team pinpointed machine learning as a critical tool for driving significant enhancements and achieving rapid results.
The existing rules-based approach was also producing unmanageable levels of false positives, resulting in increasing compliance costs.
Project scope
The bank was looking to implement a machine learning solution with a risk-based approach to its AML compliance, enabling a global framework to support its international operations. The bank also intended to implement the risk-based approach across other areas of its business, such as capital markets and retail banking, for continuity.
Client objectives
- Implement a machine learning transaction monitoring solution that would be accepted by the Central Bank of the country where the bank is headquartered.
- Ensure compliance with local and global regulators.
- Test that the machine learning solution is able to detect traditional money laundering activity plus more nuanced financial crime such as human trafficking.
- Deploy the solution to cover all lines of business over time.
The results: fewer false positives, better risk detection
ThetaRay analyzed the bank's historic data and was immediately able to provide results.
Significant reduction in false positives
In some countries where the bank operates, the number of alerts declined by over 500,000 per year. By establishing normalcy based on the data, ThetaRay's transaction monitoring solution was able to detect abnormal behavior. This ensured that only true positives were alerted, reducing the time and cost for compliance teams.
Improved risk detection
ThetaRay's solution was able to find "unknown unknowns": risks and behavior that were previously unidentified by the legacy systems. These findings gave the bank greater insight into its risk and enabled the transaction monitoring and compliance teams to make more informed decisions when reviewing customers and transactions.
Approval from the Central Bank
Most importantly, by introducing more robust and dynamic transaction monitoring across the bank's global framework, the financial institution gained approval from its Central Bank and other global regulators. The bank is now recognized as a leader in financial crime detection.
Frequently asked questions
Why did the bank replace its rules-based AML systems?
The bank's rules-based AML approach was producing unmanageable levels of false positives and increasing compliance costs. The transaction monitoring team identified machine learning as a critical tool for driving significant enhancements and achieving rapid results.
How much did ThetaRay reduce alerts for the global bank?
In some countries where the bank operates, the number of alerts declined by over 500,000 per year. ThetaRay established normal behavior from the data so that only true positives were alerted.
Did regulators accept the bank's machine learning transaction monitoring?
Yes. By introducing more robust and dynamic transaction monitoring across its global framework, the bank gained approval from its Central Bank and other global regulators.
What are "unknown unknowns" in AML transaction monitoring?
In this project, "unknown unknowns" were risks and behavior that the bank's legacy systems had not previously identified. ThetaRay's machine learning surfaced them, giving the compliance teams greater insight into customer and transaction risk.
Can machine learning transaction monitoring detect human trafficking?
One of the bank's objectives was to test that ThetaRay's machine learning solution could detect traditional money laundering activity plus more nuanced financial crime such as human trafficking.
