Network analysis for financial crime risk assessment: the case study of the gambling division in Malta

Abstract

ABSTRACT The present study aims to support existing risk assessment tools by proposing an innovative network-oriented methodology based on ownership information. The approach involves calculating company-level indicators that are then transformed into red flags and used to rate risk. To this end, we collect data on companies active in the division of gambling and betting activities in Malta, and further combine them with information on enforcement actions imposed on Maltese companies, their beneficial owners, intermediate shareholders and subsidiaries. Correlation analysis and statistical testing were performed to assess the individual relevance of company-level indicators, while machine learning methods were employed to validate the usefulness of the indicators when used collectively. We conclude that the intelligent use of ownership information and proper analysis of ownership networks greatly supports the detection of firms involved in financial crime, hence the recommendation to adopt akin approaches to improve risk assessment strategies.