Dec 11, 2021

How Simple Rules and Rigid Systems Hinder Effective Detection

There have been positive developments in the anti-money laundering (AML) industry, but a lack of flexibility and precision in monitoring has hindered our capacity to catch criminals. Transaction monitoring systems still have a hard time detecting true suspicious activity given the industry’s unacceptably high false-positive rate of 99%. With a false positive rate this high, spending in the US and Canada on compliance associated with catching money launderers is estimated to be $49.9 billion for 2021 as companies plug the gaps left by sub-par technology with people. Despite the massive investment in human capital, trillions of dollars are laundered every year.

Blunt detection methods have been a particular source of pain in my experience. Monitoring strategies are still created with overly simplified rules which don’t reflect how a compliance analyst thinks through dynamic scenarios. Many rules also do not take into account mitigating factors in the dataset leading to unnecessary alerts. One risk typology that is heavily impacted by this is structuring. If a user breaks up large cash deposits into smaller cash deposits to avoid filing certain government forms, that would be indicative of structuring. However, if they were to have filed these reports, it would indicate to an analyst that they might be less likely to have been trying to avoid information sharing. Many financial institutions do not take into account this behavior when detecting structuring, leaving analysts to manually check for mitigating behavior, despite the fact that this behavior is embedded in the same data that is analyzed to create that alert.

Even these simple rules are slow to be implemented, leaving gaps in the industry’s ability to assess evolving criminal activity. In a certain part of the US, there could be instances of cash activity at electronics companies that were indicative of trade-based money laundering schemes. This holds true for a certain set of zip codes and a certain type of industry. Identifying this activity would just require some small tweaks to an existing structuring rule. However, in many legacy systems, I have seen that even small changes take months to deploy, leaving analysts behind and firms exposed to risk.

Monitoring should be based on the pattern of transactions making sense for that specific entity and their holistic relationship with the institution. If analysts and officers think this way then we should be offering a better solution that thinks this way too. This ability to mirror the analyst’s mind is an important part of implementing a truly effective transaction monitoring program. We should be able to account for various mitigating factors and historical interactions. We should be able to react quickly to changing risk landscapes. By thinking historically and holisticallywe can do better.

Get started today.

Talk with our team to learn how we can scale your AML systems, with less