How Rule Experimentation Tools Improve AML Transaction Monitoring

How do you know if your anti-money laundering (AML) transaction monitoring rules are working as intended? 

If you work at a regulated financial institution (FI), you’re required to conduct tests that prove the effectiveness of your AML policies. Regulators expect FIs to conduct regular transaction monitoring rule testing to improve their ability to detect suspicious activity, adapt to new money laundering typologies, and maintain efficient and effective AML programs.  

In this article, we discuss how rule experimentation tools can simplify and improve the testing process. 

Key Points 

1. Above-the-line, below-the-line, and new rule testing make transaction monitoring more effective and efficient 
2. Manual rule experimentation processes can be expensive, time-consuming, and difficult 
3. Automated rule experimentation tools make testing less expensive, faster, and easier 

How Does Rule Experimentation Improve AML Transaction Monitoring? 

There are two primary ways to test transaction monitoring rules: 

1. Above-the-line (ATL) and below-the-line (BTL) testing: in these exercises, you would simulate existing rules, testing higher thresholds (above the line) and lower thresholds (below the line) on a sample of transaction data. 
2. Testing new rules: to add a new rule covering an emerging typology, you would first run the rule on historical data in a sandbox environment. 

The aim of both exercises is to make your AML program more effective and efficient. This improves regulatory compliance and increases risk coverage.   

Regulatory bodies promote regularly measuring the effectiveness of your transaction monitoring rules and technology. According to the FATF,  
 
“By putting in place measurements of effectiveness, regulated entities will be encouraged to be more outcome oriented, and also ensure that the adoption of new technologies is fit for purpose and continue to perform adequately over their life cycle.”  

How Do AML Teams Usually Test Transaction Monitoring Rules? 

For most FIs, testing rules is a manual process. It may involve collaborating with internal business analysts and/or data scientists. AML operators and managers face the headache of working with millions of lines of transaction data in a spreadsheet. Manual rule experimentation processes cause a variety of problems:  

1. Expensive: the costs of the internal resources required to conduct manual tests add up quickly. 
2. Time-consuming: manual rule testing exercises often take weeks, if not months. 
3. Difficult: many FIs don’t have an army of data scientists on staff, and if they do, they probably stay quite busy. This means most AML operators and managers end up running manual tests themselves. Regulated entities also need to document rule tests, which is difficult to do manually.  

What Can I Do Now to Improve Transaction Monitoring Rule Experimentation? 

Even without using an automated rule experimentation tool, there are several things you can do now to improve your testing operations: 

1. Invest in testing: if you can’t get a new solution approved right away, you can at least escalate any issues and build a business case for more testing resources.  
2. Improve processes: you can evaluate your existing testing procedures to discover feasible opportunities for efficiency gains.  
3. Training: you can regularly address testing with your staff to share methodologies for better performing rule tests. 

How Does a Transaction Monitoring Rule Experimentation Tool Improve the Process? 

A rule experimentation tool improves testing processes in three key areas:  

1. Less Expensive: with a rule experimentation tool, there’s no need to rely on costly internal data science or business analysis resources. These tools automate statistical analysis, making the process less expensive. 
2. Faster: a rule experimentation tool can run tests automatically, in minutes not weeks, with only a few clicks. 
3. Easier: rule experimentation tools make demanding data compilation tasks unnecessary, as it has already integrated available data sources. 

How Is a Transaction Monitoring Rule Experimentation Tool Used? 

Automated rule experimentation tools empower AML teams to conduct better tests with a simple, straightforward process:  

1. Create test environment: a rule experimentation tool can mirror historical data in a separate tenant of the transaction monitoring system. 
2. Update rule: rule experimentation tools feature simple configurations for testing various rule thresholds or new rules. 
3. Set test parameters: with a rule experimentation tool, setup is simple, and it takes only a few minutes. 
4. Run test: rule experimentation tools run tests on historical data at the click of a button. 
5. Analyze results: a rule experimentation tool can identify differential alerts, i.e., the additional alerts that triggered during the test. Comparing these alerts with prior results will demonstrate the rule’s effectiveness and/or efficiency. 
6. Adjust in live environment: once the rule change is deemed more efficient (triggers fewer unproductive alerts) or more effective (triggers more productive alerts), the transaction monitoring system can make the change. If the first test did not achieve the desired results, the tool can repeat the test with a different configuration. 
7. Document: an effective rule experimentation tool will automatically produce audit trails detailing the results of each test. This provides the documentation necessary to justify rule changes. 

Example: Above-the-Line and Below-the-Line Rule Testing 

Mountaintop Bank is preparing for its next regulatory exam. The AML transaction monitoring team decides to test the thresholds of their rules.  
 
First, the team creates a test environment in their transaction monitoring system. They identify a volume rule to test. They have previously set the rule threshold at $5,000.  

For the above-the-line test, the transaction monitoring team adjusts the threshold up 10%, from $5,000 to $5,500. Then they run the test in the sandbox environment. They test the rule again with a 20% increase, changing the threshold to $6,000.  

The team follows the same process below the line. They adjust the rule threshold down 10%, from $5,000 to $4,500, and run a test. They test again at a 20% decrease, changing the threshold to $4,000.   

The transaction monitoring team analyzes the test results. Each adjustment resulted in a different rule efficiency score, as illustrated below. A higher score means the rule finds more suspicious activity and/or generates fewer false positives at that threshold. Because the rule was most efficient at the threshold 10% above the line ($5,500), the team decides to make this adjustment in the live transaction monitoring system. The system now finds more suspicious activity and produces fewer unproductive alerts, so the AML team can focus their attention where it’s most needed.  

Example: Testing a New Rule  

Mountaintop Bank has uncovered a unique money laundering typology in their casework. The transaction monitoring team creates a new rule to monitor it.  

The new rule triggers an alert if five transactions to a counterparty in a particular high-risk jurisdiction occur in one week. The team makes the configurations in the rule manager, then simulates the rule in the sandbox environment.  

The rule experimentation tool applies the new rule to a sample of Mountaintop Bank’s transaction data from the previous ninety days. The rule generates four differential alerts. The transaction monitoring team analyzes the alerts and determines that most of the activity is suspicious enough to open a case. They apply the new rule in their system for additional monitoring. 

Hawk’s Transaction Monitoring Rule Experimentation Technology 

Hawk has developed a transaction monitoring rule experimentation tool that automates testing processes for banks. With this technology, AML professionals can perform what used to be time-consuming and resource-intensive rule tests in a matter of hours, by themselves. It takes the headache-inducing manual work out of rule experimentation. Using this powerful tool will help banks optimize their transaction monitoring rules for efficiency and effectiveness, improving their risk coverage and AML compliance. 

Contact us to learn more about our transaction monitoring rule experimentation technology.