基于贝叶斯网络模型对我国证券公司操作风险的实证研究

The Substantial Evidence Research on Operation Disk of Domestic Securities Companies Based on Baytsian Network

20世纪90年代以来,国内证券公司相继发生操作风险而导致重大损失事件,一些证券公司被迫关闭或被接管。证券行业与监管当局高度重视操作风险,将操作风险与市场风险、信用风险并列作为内部控制的重点。国内逐步开发金融市场,证券公司面临着巨大挑战。控制和管理操作风险的水平是反映证券公司竞争力的重要内容。该文以2001～2009年国内证券公司导致重大损失事件的数据为依据,采用贝叶斯网络法建立简单模型,得出操作风险的损失发生概率。通过实证分析表明,贝叶斯网络模型适用于我国证券公司操作风险管理,并可用于假设分析,能够更有效地降低证券公司的操作风险。

Since the 1990s of the last century, great operation risk losses take place in domestic securities companies in succession, and some securities companies are closed or taken over. Securities industry and supervision authorities pay close attention to operation risk, regard it, as well as market risk and credit risk, as the focal points of domestic control. The domestic financial market develops in an all-round way progressively, and the securities companies are facing the enormous challenges. The control and management level of the operation risk is the important index of the competitiveness of securities companies. The full text is based on the loss data of the large-scale loss incident of domestic securities companies of the years from 2001-2009, and it adopts Bayesian Network method to set up simple model, and obtains the happening probability of loss of operation risk. Through analyzing in the real example, Bayesian Network model holds good for the disk management of domestic cecurities companies, and using with hypothesis analysis, it reduces the operation disk of securities companies effectively.