To implement a risk-based regulatory approach, this paper aims to make an assessment on customers' money laundering risk and conducts some applications.
During the transition of a regulatory approach from “rule-based” to “risk-based”, this paper considers that the area of a customer, types of business and the industries to which the customer belongs are the main indicators to judge money laundering risk of a customer. Based on the statistical analysis of 221 typical money laundering cases, first-class index weights are given by using the entropy weight method and then by combining with the membership function, this paper determines a customer’s money laundering risk levels. On the basis of the entropy weight method, this paper uses the C5.0 algorithm to construct a decision tree model and then carries out application research on customer money laundering risk assessment to verify the effectiveness of the entropy weight method and the decision tree model.
This empirical research found the weights of three key money laundering indicators: customer areas, business types and corresponding industries.
Asserting that current money laundering risk assessments of customers are marginal at best, this paper contends from the perspective of practice, and applies the entropy weight method and the decision tree model for money laundering risk assessment of customers.
