基于神经网络模型的财务困境预测实证分析  

An Empirical Study of Financial Distress Prediction with Neural Network Model

作　　者：马喜德[1] 

机构地区：[1]厦门大学金融系,福建厦门361005

出　　处：《山西财经大学学报》2005年第1期124-128,共5页Journal of Shanxi University of Finance and Economics

摘　　要：财务困境预测是金融领域一个倍受关注的研究课题,运用定量分析提前对企业(特别是上市公司)陷入财务危机进行预警,对于债权人、投资者以及监管部门来说都有极其重要的意义。通过研究我国A股上市公司,以公司因财务状况异常而被特别处理(ST)作为企业陷入财务困境的标志,利用BP神经网络模型提前三年进行财务困境预测,结果发现,预测准确率可以达到87%。Financial distress prediction, a much-focused subject in financial studies, provides early warning of financial distress for business firms, especially the publicly held ones. It is of great significance not only for creditors and investor, but also for supervisors and regulators. The study takes special treatment of companies of abnormal financial status as an index and utilizes BP neural network model to make predictions and it attains an accuracy of 87%.

关 键 词：上市公司 财务困境 神经网络 

分 类 号：F224.9[经济管理—国民经济]