Abstract: Facing the risk events of slow development of e-commerce, disorder of e-commerce market and collapse of e-commerce industry which are all caused by e-commerce credit issues, this paper uses the Bayesian network theory to construct the credit risk network model of e-commerce in China, in order to analyze the impact of both sensitivity factors and key factors on risk events. The results are shown by using GeNie simulation software: three crucial credit risk factors are higher technical fraud, lower economic efficiency and lack of national regulatory effort; the five sensitivity factors are uneasy new comers, reduction of transaction willingness, higher technical fraud, lower economic efficiency and information asymmetry. The study shows that the change of the key factors of credit risk will result in the change of sensitivity factors in credit risk, thus leading risk events to happen at last.

Key words: electronic commerce, credit risk, key factor, sensitivity factor, Bayesian network

摘要： 针对电子商务信用问题可能会导致的电子商务发展缓慢、电子商务市场秩序混乱以及电子商务产业崩溃等风险事件,应用贝叶斯网络理论构建中国电子商务信用风险网络模型,分别分析电子商务信用风险的敏感性因素与关键因素对风险事件的影响程度。运用GeNie仿真软件求得结果:最关键的信用风险因素是技术诈骗率提高、经济效率降低以及国家监管力度不足等三个因素,敏感性因素为新者难做、交易者的交易意愿降低、技术诈骗率提高、经济效率降低以及信息不对称等五个因素。研究认为信用风险关键因素的变化会波及信用风险敏感因素的变动,进而引发风险事件的产生。

关键词: 电子商务, 信用风险, 关键因素, 敏感性因素, 贝叶斯网络