January 2025
JEL Classification: C49, G28, K42
Keywords: Money laundering, Suspicious Transaction Reports, Italy, Quantile Random Forest, Imputation.
This study proposes a robust and innovative methodology for estimating the value of money laundering and other illicit financial conduct in Italy using information from suspicious transaction reports (STR) received by the UIF.
The methodology is based on the joint application of statistical methods and machine learning techniques, and on a rigorous selection process to identify the relevant STRs, i.e. those that, following the financial analyses conducted by UIF experts, present a high risk or have received positive investigative feedback from the investigative bodies.
The results provide an estimate of money laundering in Italy for the period 2018-2022 of around 1.5-2.0% of GDP, corresponding to approximately 25-35 billion euros per year.
