The fight against money laundering has led to the creation of many policies aimed at curbing related criminal activities.
With financial institutions forced to comply with these policies or be fined, it is understandable why risk and compliance teams are intent on finding the best solutions for Anti-Money Laundering (AML) transaction monitoring.
However, there are a lot of issues associated with AML transaction monitoring, and one of the most difficult to deal with is the occurrence of false positives. So, in this article, we will cover:
- What false positives and false negatives are with regard to anti-money laundering,
- How organizations can reduce false-positive rates and,
- The balance you need to maintain for a highly effective AML transaction monitoring system.
Let’s get to it.
What are False Positives in AML and Why Do They Happen?
Separate from the better-known “Know Your Customer” (KYC) policy is the “Know Your Transaction” (KYT) policy, which is an extremely fundamental component of any buttoned-up AML compliance program.
The KYT policy, in application, involves the establishment of behavioral “red flags” within a rule-based transaction monitoring system. When these established behavioral red flags are seen, an alarm is raised, and an alert investigation process begins.
Manual monitoring is almost impossible given the broad scope and scale of data that fit the term “transaction.”
By definition, a transaction is anything that denotes a completed agreement between a buyer and a seller where money is exchanged for goods, services, or financial assets. Transaction monitoring software programs are deployed to spot these tricky to identify behavioral red flags automatically.
Now, what happens when these behavioral red flags are not what they seem after further review?
For instance, when a suspected payment between two parties turns out to actually be an innocent series of transactions? This is a prime example of when the transaction monitoring rules set up to catch suspicious transactions deliver what is known as a false positive.
Unfortunately, false positives come with the territory. In fact, financial institutions are incentivized to implement overly broad rules which lead to more alerts so as not to accidentally miss something legitimate.
This means that there will inevitably be alerts for transactions that aren’t truly suspicious, and with zero context accompanying the transactions, both legal and illegal transactions are directed towards the AML alert review process for manual review.
False Positives vs. False Negatives
As mentioned, false positives are innocent transactions flagged as illicit in transaction monitoring scenarios.
On the other hand, false negatives are illegal transactions unintentionally allowed to pass through without triggering the AML alert review process.
As a result, the system ignores these legitimate money laundering transactions, which is clearly a failure that could lead to losses and compliance issues down the line.
High rates of false positives and/or false negatives are a risk to any AML compliance program and the organization.
3 Consequences of High False-Positive Rates
To understand the magnitude of risk a high false-positive rate brings, you need an idea of what a false-positive rate is on its own.
Think of a case where the implemented rule-based transaction monitoring system is set to flag all transactions worth $10,000 and above for AML review.
Then, to get hold of perpetrators using “structuring” techniques, an additional rule is placed demanding AML alerts on transactions just shy of this mark, like ones that are over $9,700.
Where ten transactions are flagged, and nine of these transactions are false positives, your false-positive rate is 90% (9/10 * 100). If there are four false positives, your false-positive rate is 40%.
A false-positive rate indicates the efficiency of your implemented anti-money laundering transaction monitoring systems and techniques.
Sadly, organizations with high transaction monitoring false-positive rates suffer consequences, some of which includes the following:
The Immense Costs Of Manual Reviews
Administratively and financially, a high false-positive rate costs an organization a lot, and these costs lay in the AML case management process. For every high-risk behavior detected, a Suspicious Activity Report (SAR) is generated.
As an organization, you inevitably set personnel aside to handle this. When several SARs are generated on one customer, the information is forwarded to law enforcement, so the case can be investigated appropriately.
During the investigation the case manager will work closely with law enforcement officials, answering questions on suspected customers and being on standby to provide any additional information as requested.
Obviously, this is a large burden as limited administrative human resources are tied up for a long period of time.
Where a case turns out to be a false positive, these expended resources translate to wasted resources.
As such, an increasing number of false positives means more wasted resources, which can have a devastating impact on the institution over time.
Negative Impact on Customer Experience
While wasting internal resources on dead-end investigations is costly for an institution’s operational budget, the damage done to the customer relationship as a result of false positives is even more dangerous for the bottom line.
Repeatedly flagging customers for AML investigations leaves an organization at the risk of losing good customers.
No one wants to be treated like a criminal, and such false accusations can cause irreparable damage to an organization’s brand reputation.
Time Factor And Lapses
Manual AML alert review processes are also cumbersome, with deeper investigations involved for an extended period of time to properly study suspects.
According to FinCEN files covering periods between 2011 and 2017, large financial institutions took an average of 166 days to file SARs on suspicious transactions.
A high false-positive rate means you engage in a lot of these AML case management processes and stay at risk of missing reporting deadlines.
Considering that reporting deadlines are only placed at 30 days after suspicious activities start, and another extended 30 days of identifying if suspects are not known, difficulty in meeting these deadlines is inevitable.
What is even more mind-blowing is that 95% of these reports are false positives.
How Can You Achieve a “Healthy” False-Positive Rate?
Ironically, false-positive rates on the lower end signal that an institution is deploying an overly-specific set of rules, which in practice, means that fewer alerts will be triggered because the parameters are tighter.
While the risk and compliance team might be dealing with fewer false positives, they may also be missing legitimate threats because the rules for alerting are too narrow.
This phenomenon is termed the “false positive paradox;” even if a high false-positive rate is unwanted, a low false-positive rate is also undesirable.
Therefore, a balance is what every compliance team needs, and we are here to help you achieve this. So how do you push towards a healthy false positive rate?
1. Implement A Risk-Based Approach
A risk-based approach in AML means creating risk profiles for entities to be monitored and implementing rules and policies appropriately.
For your transaction monitoring, it involves creating risk profiles for customers, with individuals with high-risk profiles considered likely to be engaged in money laundering activities and, once caught, placed on sanction lists.
By creating a risk profile, you reduce the scope of data considered relevant for the AML alert review process and, somewhat intelligently, the number of false positives without risking a higher false-negative rate.
2. Properly Organize Your Collated Data
Organizing your data involves ensuring that every piece from your vast stream of information collected for analysis is distinguishable. For instance, and in consideration of one of the most factors that cause false positives, customer names may be structured more appropriately.
Rather than pulling full names relating to an individual as one unit, these names may be structurally pulled as first names, middle names, last names, and even titles.
This gives more definition to individual identities and reduces the number of individuals matched with those placed on sanctions lists or identified with high-risk profiles.
3. Use a Reliable Anti-Money Laundering Transaction Monitoring Solution
A high false-positive rate is an end product of an ineffective transaction monitoring process. Usually channeled through software programs, with ordinary rules-based systems, there is no context given to transactions while they are monitored.
This is the reason why all transactions fitting regulatory thresholds keep popping up for review.
Advanced solutions like Unit21 offer sophisticated transaction monitoring scenarios and dynamic rules that can easily be adapted, tested, and deployed, to suit a variety of use cases.
Our world-class transaction monitoring RegTech software enables risk and compliance teams to customize rules and create complex statistical models without pulling on expensive and limited engineering resources.
Thanks to increased team availability, you aptly control your AML alert investigation process, reduce costs through more relevant positives, and even deploy more appropriate risk management programs and solutions.
What's more, your SARs filing is not just done electronically but also automatically, meaning you drastically reduce manual administrative work.
4. Continuous Improvement Is Key
Criminal methods and technologies improve daily while your implemented AML compliance rules only change when you want them to. As methodologies improve, some become obsolete due to improved technologies within financial frameworks.
Continuous review of the money laundering environment helps you know what measures don’t necessarily have to be implemented and reduce alerts generated through them.
Once all these are done, how do you know that your processes are working, and how do you stay satisfied at what false positive rate? Our in-depth case study with our customer, Bakkt, sheds a very bright light on this.
Based on recent industry studies, we place a healthy false positive rate at around 15%. This rate is great within an industry with an average of 90% - 95% false-positive rate.
This case study also looks into what a healthy timeframe is for SARs management (20 minutes per SAR), and serves as a testimonial to the reliability and effectiveness of Unit21’s transaction monitoring solution.
Reducing False Positive Alerts: Key Takeaways
While costly and perturbing, false positives can’t and shouldn’t be entirely eliminated from your AML program. They signify a system that is broad enough to catch suspicious activities, even if not all of the transactions they flag end up being a real crime in action.
However, that’s not to say that financial organizations can’t work toward a more efficient and reliable system.
With all this in mind, here are some key takeaways:
- High false-positive rates create additional administrative and financial costs, as well as increasingly unfavorable customer relationships.
- Even with this, the false positive paradox means you are not permitted to totally eliminate false positives from your AML program as they serve as an indicator that the system works.
- Organizations need a proper AML scenario and rules implementation infrastructure with appropriate customer indexing and profiling.
- It is very difficult, if not impossible, to achieve this at scale given ever-changing threats and requirements without an advanced and reliable transaction monitoring solution.
With sanctions and fines waiting for you at almost every corner, implementing a highly accurate transaction monitoring solution can drastically improve efforts to combat AML.
Transaction monitoring can empower teams to assess historical/current customer data to provide a complete picture of customer activity.
Interested in seeing how Unit21’s transaction monitoring software can help you take control of your compliance? Schedule a demo to see it in action today.