Workflow
证券市场智能化仿真系统
icon
Search documents
AI赋能金融监管:深交所智能监测系统找到多项违法违规线索
Di Yi Cai Jing Zi Xun· 2025-11-28 14:32
Core Insights - The Shenzhen Stock Exchange (SZSE) has developed advanced technologies for monitoring abnormal trading behaviors, addressing the complexities of modern market activities [2][3][4] - The "Intelligent Monitoring and Detection Technology for Abnormal Trading Behaviors in Securities Markets" project has successfully completed its implementation phase, enhancing regulatory capabilities [2][3] - The industry regulation model aims to improve the accuracy and completeness of regulatory responses, leveraging AI to address the challenges of traditional methods [7][8] Group 1: Intelligent Monitoring Technologies - The project has produced three core technologies: investor trading behavior classification, insider trading detection, and a securities market simulation system [2][3][4] - The investor trading behavior classification technology utilizes a comprehensive label system to analyze trading behaviors across six dimensions, significantly improving the precision of investor profiling [3] - The insider trading detection technology employs a large-scale trading behavior graph to identify suspicious trading patterns, enhancing regulatory efficiency and transitioning to a proactive monitoring approach [4] Group 2: Securities Market Simulation System - The securities market simulation system integrates historical trading data to create parameterized behavioral models, allowing for the simulation of various trading strategies [5][6] - This system addresses the limitations of traditional empirical research by providing tools for risk assessment and strategy validation, thus shortening the development cycle for trading strategies [6] Group 3: Industry Regulation Model - The industry regulation model, developed in collaboration with Huawei, aims to enhance the understanding and accuracy of regulatory inquiries through an intelligent Q&A platform [7][8] - The platform has been deployed for internal use and is being trialed by regulatory and market institutions, showing significant improvements in response accuracy and completeness [7][8] - The model serves as a pilot for future task models in the industry, supporting various regulatory functions and contributing to the legal foundation of the capital market [8]