Tracking money laundering matters: But don’t cast too wide the webs of suspicion

I WOULD like to introduce you to Alyssa Thomas, a 6-year old girl from Ohio. She was placed on the terror watch list, causing airport havoc whenever her family would fly. How does a 6-year old end up in the terrorist watch list? When her father was asked by news reporters “What could she possibly have done?”, he retorted “she may have threatened her sister, but I don’t think that constitutes a Homeland Security trigger”.

In intelligence profiling, the case of Alyssa Thomas is not rare at all. It is called a false positive (that is, a case that is not really suspect following manual analysis – or common sense). But the issue with false positives is that they tend to become a persistent white noise, acting against the extraction of useful information. In the context of tackling money laundering, false positives are the rule; not the exception. Their handling and management consumes considerable analyst resources that are wasted on examining cases that aren’t really suspect.

An average rate of “success” in dealing with money laundering through screening transactions electronically is around 1 per cent. This means that from the 100 suspect cases that are flagged by monitoring systems, only one turns out to be really suspect after manual analysis takes place. And this costs a lot of money. A bank that has four to five dedicated analysts for reviewing suspect cases spends about $500,000 per year in examining false money laundering suspicions. It’s not only a waste of time and money; staff become demoralised and many don’t even see the point of it all. But there is another issue for consideration here: algorithms and queries in transaction monitoring systems define who is suspect for money laundering by using all sorts of techniques, transaction data and demographic rules; and in the vast majority of occasions, they fail.

If such profiling is difficult for money laundering then for terrorist financing it seems almost impossible. For instance, the terrorist attacks in the London Underground are estimated to have cost around £8,000 to pull off. How could anyone monitor this low-level activity within the vast volume of transactions taking place daily? Banks cannot profile terrorist financing behaviour and even the inter-governmental Financial Action Task Force admits in its guidelines that “it should be acknowledged that financial institutions will probably be unable to detect terrorist financing as such.” You can appreciate the difficulties in dealing with terrorist financing by considering the scale of confiscation of terrorist assets. For example, the money confiscated for the financing of the Taliban amounted to $2,648 according to the 2008 US Terrorist Assets Report. In subsequent annual reports, the figure was omitted altogether. Meanwhile, only 7 per cent of the rest of the world’s specialised financial intelligence units differentiate suspicious transaction reports related to terrorist financing in their annual reports. From the remaining 93 per cent there is silence.

Why do we seem to be ignoring these fundamental problems? We seem to have convinced ourselves at a high level of the validity of absurdities: that tackling terrorist financing can and should be mixed up with anti-money laundering efforts while intelligence agencies remain underfunded; or that feedback from financial intelligence units to financial institutions is not really critical and necessary (it has yet to be formalised and made compulsory).

The initial momentum of the G8 expressed in the Financial Action Task Force’s constitution has had important effects on anti-money laundering compliance, but there are many issues that need to be looked at more closely. Semantic problems in defining money laundering come into play and generate yet more variation. What does get reported varies in its structure, cases and classifications, not to mention content. In fact, to make any sense of what is actually going on, or to construct a meaningful statistical analysis from currently available data is a challenge in itself. A significant number of countries don’t even have an established financial intelligence unit for dealing with money laundering.

As re-regulation is bound to trigger unanticipated consequences for businesses and potentially steeper financial fines for non-compliance, it is becoming increasingly evident that the answers to “Who’s suspect for money laundering?” are expanding, not only through legislative initiatives but also through the use of technology-based monitoring systems. But most of the time, the suspects turn out to be perfectly legitimate. “Who’s really suspect?” is another matter altogether.

Dr Dionysios S. Demetis has a PhD on anti money laundering from the London School of Economics where he has also taught classes in information systems management. His Gresham College lecture, Unweaving the Anti Money Laundering and Counter Terrorist Financing System: Complexity, Simplicity and Paradox is this Thursday at 6pm at Barnard’s Inn Hall.

The money confiscated for financing the Taliban was declared as $2,648 in 2008