November 22, 2019
16:00 - 17:00
ISBA - C115 (Seminar Room Bernoulli)
Applied statistics workshop
Cameron Todd and Jonathan Moriau, Intys, Brussels
"How to automatically identify UBOs (Ultimate Beneficiary Owners) on a worldwide scale?"
A B S T R A C T
The purpose of the project was to design, build and test a scalable solution to identify ultimate beneficiary owners (UBO) in network of 50million + companies and beneficiaries. The solution can help companies and regulatory bodies to identify any ultimate owners that may operate in fraudulent, money laundering and terrorist financing activities. The definition of a UBO of a legal entity is the natural person who is the beneficiary of at least 25% of the legal entity’s capital or 25% voting rights. To help with building a scalable solution we researched into various graph technologies and the benefits they have over a traditional relational database in network problems such as this. We proposed various algorithms to identify and validate the UBO’s including loop detection, connected components, search, geometric series, Monte Carlo simulations. The current solution can identify 85% of UBO’s owning at least 25% of the legal entity’s capital at 100% accuracy, the remaining 15% can have on average small errors in the calculation because of complex loop ownership structures in the network. The final work of the project will include the development of Monte Carlo simulation and game theory methods to estimate the voting power of an entity.