Detecting money laundering using filtering techniques: a multiple-criteria index

Money laundering is a dynamic activity attempting to circumvent anti-money laundering (AML) actions. We propose a money-laundering detection approach encompassing three separate detection measures applied simultaneously, providing a consolidated index to minimize circumvention. The index incorporates three detection measures: (1) deviations in trading volume and frequency; (2) unusual payments to or receipts from an atypical trade partner; and (3) Benford's Law, based on the number of times a specific digit occurs in a particular position in numbers to detect financial fraud. Finally, we design a numerical test that any reasonable detection approach should satisfy. Our results successfully discover possible fraud planted in the simulated data.money laundering, financial fraud, data mining techniques, outlier detection,