Are we poised for a breakthrough in the fight against money laundering?

COMARCH/SUPPLIED

Real estate purchasing, smurfing, correspondent banking—ways to launder money gain in finesse. And become harder and harder to trace there is hope though.

The total value of money laundering transactions is estimated at two to five per cent of global GDP, or approximately $1-2 trillion annually. Only 10 per cent of suspicious transactions reported by financial institutions are passed on to law enforcement agencies for further investigation. European banks will spend more than EUR 1 billion a year by 2020 on KYC management.

Those numbers leave no doubts with the scale we are facing here.

What are banks doing today to counteract the problem?

In 2017, Alix Partners prepared a study in which they surveyed 361 financial institutions around the world on their anti-money laundering (AML) sanctions compliance programs. The respondents were mostly working in wide range of AML and sanctions functions.

The results spotted the most challenging areas where transaction-monitoring systems took the first spot:

Apparently, big efforts are being made. But if it’s so good, then why is it…so bad? Take these examples:

• Dutch ING agreed to pay $900 million under the settlement to close the investigation into money laundering.
• An independent investigation has shown that the astronomical amount of $30 billion has been laundered by the Estonian branch of Danske Bank since 2013.
• The British HSBC bank paid a record penalty of $1.9 billion for participating in the money laundering scandal.

Analysing the examples above, as well as talking to many representatives of global and local AML departments, one could come to the following conclusions:

• Digital transformation changes the way banks operate—while AML procedures stay intact.
• Core-banking systems are subject to constant changes.
• Criminal and terrorist groups are becoming more and more sophisticated in attempting to break AML systems.
• Most AML systems of today are old, and rule based.
• AML procedures are largely based on subjective evaluation made by bank employees.
• Very often the human factor also turns out to be unreliable: errors/greed.

As a result, all energy today goes into manual verification, then it turns out to be a false alarm, but the real cases of money laundering go unnoticed. 

Is it possible to increase the effectiveness here and to mitigate the risk?

Mitigating risks can be done the old way—employ more AML analysts, define more rule but there is a question—what’s enough?

You can also think about a hybrid model in which you combine AI with human factor.

The technology detects anomalies in customer behaviour, estimates the risk, and then builds rankings, limiting them to most suspicious cases only—the ones experts from AML departments should really concentrate on.

This is what banks can gain from such an approach:

• Enhanced money laundering detection process—less false alarms, more ways to discover new money laundering patterns
• Reduced cost—lower operating cost through automated transaction monitoring
• Increased effectiveness—greater productivity of AML analysts
• Added value for shareholders, clients and regulators—AI based approach to AML

The current AML approach is more on the side of keeping up with money laundering, not preventing it. Only combination of AI and human factor, not isolation of both, could bring the greatest benefit. The question that financial institutions should ask themselves is: can we afford not to accept AI in our AML procedures?

For more information click here.