Some of the methodology in this article seems flawed. e.g. They rank banks by the total value of suspicious activity reports (SARs) each has filed, and infer h

Some of the methodology in this article seems flawed. e.g. They rank banks by the total value of suspicious activity reports (SARs) each has filed, and infer higher amounts as evidence of broken AML. I don't dispute the underlying accusations, but one could also argue SAR filings are a signal of compliance (at least compared to other banks that have undergone less scrutiny in the past and might be less vigilant in their reporting). Neither do I see any attempt to normalize the data (e.g. by number of customers).