Domain and Learning Models to Support Automated and Comprehensive Post-trade Regulation Compliance

点击翻译按钮获取中文摘要

Canadian financial organizations are required to comply with the regulations specified by the Investment Industry Regulatory Organization of Canada (IIROC) to be able to perform any trading operations in Canada. According to IIROC regulations, investment advisors should be registered in order to make sure they comply with a set of rules related to business conduct, financial operations and trading practices. Consequently, every trade performed by an investment advisor on behalf of a customer should comply with the IIROC regulations. This process is undertaken by a compliance supervisor. Currently, compliance supervisors rely on a risk-based approach using the data generated by a wealth management software in order to detect trades that might violate the rules. However, this approach does not scale with the sheer number of trades made on a daily basis. In addition, there is a shortage in terms IIROC regulation expertise and the impact of IIROC supervisor is not uniform. There is a need for an approach to perform a systematic, comprehensive and timely trade validation. The objective of this research project is to enable the automation of the trade compliance supervision process. In order to achieve this objective, this project will explore how to integrate a domain model for IIROC built using a model-driven engineering with advanced machine learning techniques.