我国上市公司价值创造与价值损害的判定分析  

An Empirical Research on Predicting Value Created and Value Destroyed for Chinese Listed Firms

作　　者：翁洪波[1] 吴世农[1] 

机构地区：[1]厦门大学管理学院

出　　处：《经济管理》2006年第8期32-41,共10页Business and Management Journal ( BMJ )

基　　金：国家自然科学基金项目(70372035)的资助。

摘　　要：本文以1999年底前在沪市上市的429家A股公司为研究对象,分别以样本公司的2002年度EVA绝对值和相对值(权益资本效率)为标准将样本公司划分为“创造价值”、“模糊状态”或“价值损害”三种类型,然后从这三类公司中各随机抽取80％的样本公司为建模样本,根据其1999～2001年的有关财务变量、公司治理变量和行业变量,分别应用Fisher和Logistic判定分析方法建立我国上市公司价值创造预测模型,并使用剩余20％样本公司作为检验样本验证模型。研究结果表明,基于上市公司的(T-1)年度的财务数据创建预测模型误判率较低；Logistic模型的判定准确性较高；EVA相对值判定模型的准确度略高于EVA绝对值判定模型；财务指标、公司治理指标和行业特征对公司价值创造具有显著影响。The paper selects 429 listed finns in Shanghai Stock Exchange which were developed before 1999 as samples, and group the listed finns as ＂value creating group＂ , ＂unclear group＂ and ＂value destroyed group＂ according to the EVA and EVA / Equity ratio in 2002 of the listed finns. First, we select 80% of the listed finns randomly from each group. Then we use publicly available data in the fiscal year 1999, 2000 and 2001 to develop Fisher Discriminate model and Logistic Regression model respectively to predict the value creating of Chinese listed finns. In the end, we use the rest listed finns to test the validity of each model. The results show that the input variables work well to predict whether the listed finns create or destroy value before one year. We also find Logistic Model is better than Fisher Discriminate Model in predicting EVA value when using the same information collection, addition to financial ratios, corporate governance variables and industry variables did affect the prediction model.

关 键 词：经济增加值 判定分析 实证研究 上市公司 价值创造 价值损害 

分 类 号：F279.246[经济管理—企业管理]