Follow the money: Rethinking geographical risk assessment in money laundering

The new EU Money Laundering Regulation (EU 2024/1624) explicitly defines financial secrecy as a geographical risk factor that obliged entities must take into account when applying their customer due diligence obligations to customers from third countries in the future. According to the regulation, financial secrecy arises, for example, when countries hinder the exchange of information, do not maintain registers of beneficial owners or have strict banking secrecy. These factors overlap with the indicators of the Financial Secrecy Index, thus opening up the possibility of assessing geographical risks in money laundering prevention in a more evidence-based and less politically biased manner.

Traditional high-risk country lists (so-called ‘blacklists’) fall short: they are politically biased, binary (yes-no) and thus overly simplistic, and risk having discriminatory effects. These lists usually target small countries or those with lower incomes, while large financial centres are often overlooked. This is why they have been criticised by the IMF, for example, and why the FATF has promised improvements. A data-driven assessment can compensate for this distortion and draw attention to those countries and transactions that actually pose a high risk of money laundering.

Against this backdrop, on 26 September, I had the honour of presenting a geographical risk model for money laundering using data from the Financial Secrecy Index at a roundtable on ‘Anti-Money Laundering in the Berlin-Brandenburg Region’. It is based on the results of the EU-funded TRACE research project (Horizon 2020) and is explained in more detail in a scientific article accepted for publication by Cambridge University Press (see a working paper version here). My slides from the roundtable can be found here.

The Financial Secrecy Index assesses 141 countries on the basis of 20 indicators. The secrecy score reflects the extent of financial secrecy in each country on a scale of 0 to 100, thus providing an objective basis for geographical risk models. The underlying database is structured according to scientific standards and makes an unrivalled wealth and depth of comparative legal data and gap analyses publicly available in over 120 data points per country. The index method was statistically validated by the Joint Research Centre of the European Commission in 2018.

The model presented combines the secrecy score with the transaction volume and uses this to calculate a risk score for each transaction or suspicious activity report. This allows suspicious activity reports to be prioritised – particularly important given the large volumes that are received daily by the FIU or generated by transaction monitoring by obliged parties. This model can serve as a safety net to ensure that no big (money laundering) fish slip through the net of other risk assessments.

As an example, we applied this method to the FinCEN Files, a dataset containing over 18,000 suspicious activity reports from the United States. Our model can help not only FIUs, but also banks and other obliged parties to fulfil their reporting obligations in a more targeted and efficient manner and reduce the risk of fines. Supervisory and law enforcement authorities can sort suspicious activity reports according to urgency and substantiate national risk analyses with data (as we outline in greater detail in an article in the European Journal of Criminal Policy and Research – open access, here).

The model is modularly expandable – for example, for sectors such as real estate or cryptocurrencies – and can be continuously improved via feedback loops. Using anonymised data packages from FIUs, supervisory authorities or obliged parties, the model could be further refined and calibrated to achieve the best results and reduce false positives. We are therefore open to partnerships with authorities, supervisory bodies and the private sector or obliged parties in order to further improve anti-money laundering and risk assessment. A data-based geographical risk assessment could lead to a more targeted use of resources and thus make the fight against money laundering significantly more effective.

The European Union could certainly be even more ambitious in this regard: according to the EU Money Laundering Directive, the above-mentioned consideration of financial secrecy as a geographical risk factor is only mandatory for non-EU member states (‘third countries’). Of course, obliged entities can go beyond this and apply the same risk parameters to transactions or business relationships within EU countries. After all, the idea that financial secrecy and money laundering risks are not a problem in the EU is far from reality.