Last week the UK’s Financial Conduct Authority decided to fine the British branch of the Bank of Beirut over £2m for failing to establish sufficient controls to guard against the possibility of money laundering or other financial crimes. In addition, while the Bank of Beirut sorts out its AML procedures, the FCA has forbidden it from taking on customers from what it considers to be “high-risk” jurisdictions for the next four months.
What counts as high risk? Any country scoring 60 or below on Transparency International’s Corruption Perception Index. To give you a sense of which countries this covers, take a look at the map below.
That’s right — nearly everyone. Of the 175 countries that TI establishes a rating for, nearly 80 percent are below the threshold set by the FCA. This includes 12 unfortunate countries whose placement below the threshold is within the CPI’s margin of error.
There are two reasons why this approach is silly. First, the CPI doesn’t actually measure money-laundering risk, but public-sector corruption — these are different concepts. Second, an arbitrary threshold is a poor way to judge the risk of money laundering.
The CPI doesn’t measure money laundering risk
Even though the two might be correlated, not all money laundering is driven by public-sector corruption and not all public-sector corruption results in money laundering. While we would expect politicians and civil servants in corrupt countries to divert state revenue to overseas bank accounts, there are many other sources of illicit financial flows which are purely private, including revenue generated from illegal activities and tax-dodging. Furthermore, most forms of petty corruption and bribery involve rents which never leave the country. This all seems clear, but none of the twelve sources used to construct the Corruption Perception Index actually involve perceptions of money-laundering risk.
The problem with thresholds
Bureaucrats love thresholds because they are easy to implement, but there is no reason to believe that risk rises sharply at the arbitrary threshold of 60 points on the Corruption Perception Index.
Other leading indicators of money-laundering risk show little going on at the 60-point cut-off. The Basel Institute on Governance has an annual composite indicator of money-laundering risk (the index itself partially comprises the CPI). As the graph below shows, a lower CPI score is correlated with a higher risk of money laundering, but there is no sharp increase at sixty.
The picture is similar if you look at another red flag: countries which are considered high risk or noncooperative by the Financial Action Task Force (FATF), otherwise known as the “blacklist.” While no country with a CPI score above sixty happened to be blacklisted in the last five years, there is no evidence that the risk of blacklisting increases exponentially at the cut-off. Finally, even though a significant proportion of countries with low CPI scores do end up being blacklisted by the FATF, this is not strong enough evidence to justify such a high degree of statistical discrimination by the FCA.
More transparency, please
This is not the first time the Financial Conduct Authority has used the CPI: in 2012 it briefly revealed, then repealed, its definitive list of high-risk jurisdictions in 2012. If regulators are going to use such imprecise measures to label countries as “high risk,” many countries will unfairly end up bearing the economic burden as banks and other financial institutions refuse to do business with them. It is time that such nonsensical measures be dropped or regulators start being more transparent about their reasons for using them in the first place.
The Stata code and the underlying data used to generate these graphs can be found here.