经典的财务危机预警模型包括判别分析模型、logistic分析模型、神经网络模型、支持向量机模型、分类数模型等,这些经典模型最重要的缺陷是他们都是静态预测模型,都不能描述财务比率变量的时间序列特点;而且这些模型用来预测财务危机的数据都是单期的,没有考虑历史值对结果的影响。用时间序列判别分析的方法估计财务比率的演变过程;用在质量管理中经典的指数加权移动平均控制图模型构建公司财务危机的动态预测模型。实践表明预测效果良好。
TSDA(times series discriminants analysis),put forward in 1990s,is used to judge the evolving distances between variants.Difference from the static distance judgment method in classic statistics,TSDA is a dynamic judgment method.EWMA is a mature theory which is widely applied in quality management.This paper combines the two theories to build the prediction model of corporate financial distress.