Reduce false positives and improve AML detection


Read a customer story: HSBC Reducing losses from fraudulent transactions

Challenging times are ahead for the European finance sector in the coming half year, with new demands for compliance regarding anti-money laundering (AML) detection and customer risk assessment.

In less than six months, all changes related to the European Union’s Fourth Anti-Money Laundering Directive will be legacy in full effect, meaning that member states must be compliant with more strict measures and incorporate them into their national legislation.

Financial institutions are eager to curb illicit financial transactions that can lead to losses, legal responsibility and reputational damage. In the area of AML detection, over the last years, legacy detection systems have generated many false positives that must be weeded out of the data before action can be taken on the suspicious activities.

This is a time-consuming process for investigators, which can stand in the way of customers’ expectations of rapid bank services. Because of this, many institutions have a strong need to strengthen Compliance and Know Your Customer (KYC) practices without increasing the cost and time factor in detection work.

AML challenges

A simple example of the challenges regarding AML detection is name screening against identified terrorists on blacklists, which can result in time-consuming false positive alert investigations. For example: “Sasha” is a male name in Russia but derived from many different names, and can be female in other parts of the world, resulting in a wrong identification.


It is also easy for fraudsters to go into hiding by making minor changes. Get a new passport, adjust a first name, use an alias and you might go “under the radar” on the blacklist.

Today, digital accounts can be created in five minutes with payments immediately authorised. Customers will not accept that it takes any longer, which makes it harder to implement AML detection. Mobile and online payment systems also presents a challenge, since fraudsters will be seeking out technology “gaps” to abuse for money laundering.

Analytics to the rescue

The solution to all this malady relies heavily on the strengths of data analytical competencies and tools. Alerts need to be more precise and patterns of transactions must be presented in a visual overview that can be easily interpreted in order to dismiss or approve transactions.

Flows of transactions between company entities or individuals must be mapped to get a complete picture. Ultimate beneficiary owners (UBO) must be identified. At last but not least, all these processes must be readily documented to live up to regulations.

The good news is that it is possible to improve AML detection with less false positives and without an increase in cost. Advanced analytics, predictive models and machine learning capabilities can bring detection to a new level.

When a customer opens a new account, proper risk assessment methodology must be applied. Risks can be estimated in real time, through advanced weight-factors models. If the behaviour or circumstances get to change, a new score is calculated. In order to get better results, a continual process of checking out the customer is highly recommended: Daily, weekly, monthly - until the client dies or leaves the bank. On top of this, governance on the models is required, and advanced analytics will help to identify variables within the KYC data that will impact your detection.

Looking at transactions is the next step in detecting AML without crying wolf. Most of the detection systems in use today is based on static profiles on the history of transactions. This is not sufficient any longer! A hybrid approach to detection is needed. Building new scenarios should include anomaly detection, text mining, advanced peer group capabilities or even predictive models. This is the only way to improve the disclosure rate of your suspicious activity reports (SAR), leading to quick and efficient detection of real cases.

The new regulations are a tall order for financial companies, and it will be difficult for many. But the benefits of a well-functioning anti-money laundering setup will be substantial and help strengthen overall compliance procedures.

Suggested read: an info site about monitoring suspicious activity, making fast decisions and staying in compliance.

Read a customer story: HSBC Reducing losses from fraudulent transactions

About Author

Christopher Ghenne

Christopher Ghenne is a financial crime specialist with over 10 years’ experience in this field. He has worked in the banking, insurance, government and telecommunications sectors helping organisations comply with Anti Money Laundering regulations and protect their customers and shareholders from losses. Working across the EMEA and AsiaPac regions, Christopher is responsible for developing propositions (including AML transaction monitoring, AML Optimisation and FATCA) for specific markets, working to strategically develop financial crime capabilities.

1 Comment

Leave A Reply

Back to Top