Using data analytics to distinguish legitimate and illegitimate shell companies

Shell companies can be a legitimate entity but can also been used for illicit activities such as money laundering. Users of shell companies have included illegal arms dealers, drug cartels, terrorists and cyber-criminals, as well as legitimate businesses. To assist in distinguishing between legitimate and illegitimate uses of shell companies, we develop a data-driven model to detect shell companies that are being used for money laundering. We use a hybrid approach combining graph analytics and supervised machine learning. The resulting detection models have an impressive classification accuracy ranging between 88.17% and 97.85%. We found no prior study that developed such models to detect illicit shell companies using publicly available information as done with our models. Beneficiaries of this work include government officials and compliance professionals, particularly accountants, tax officials and anti-corruption agencies.