摘要
文章以2002~2004年深沪两市首次被ST的118家上市公司作为研究对象,同时选择同行业且资产规模相近的118家盈利公司作为配对样本。以上市公司突然出现重大亏损、连续两年亏损、股东权益低于注册资本或每股净资产低于面值等几种情形作为界定“财务危机”的标准。利用ST公司财务危机前1年的数据,初步选择了20个财务指标,对ST公司与配对样本的均值、标准差、T统计量、Z统计量进行分析,寻找ST公司与非ST公司在财务指标上的差异,并采用主成分分析法的思路进一步筛选财务指标。最后运用二元逻辑回归(Logit),建立起上市公司财务危机前1年的预警模型,且进行了预测。
This paper takes the 118 listed companies as the study objects which were labeled ST on the Stock Exchanges of Shenzhen and Shanghai from 2002 to 2004. Meanwhile, another 118 payoff companies in the same trade which are of similar capital and size are chosen to be the partnership samples. The standards for financial crisis are sudden great loss of listed company; continuous two years of loss; shareholder's rights and interests below registered capital or net asset of each share below par value. 20 financial indexes are chosen from the previous data of one year before the financial crisis of the ST companies. The average value, standard derivation, T statistic quantum and Z statistic quantum are analyzed between ST companies and the partnership samples in order to find the difference in financial indexes between them. Main contents analysis is also used to filter those financial indexes. Finally, by using Logit, an early- warning model for listed companies one year before financial crisis is established and then some prediction is made.
出处
《山东理工大学学报(社会科学版)》
2006年第3期40-43,共4页
Journal of Shandong University of Technology(Social Sciences Edition)
关键词
财务危机
配对样本
LOGIT
预警模型
financial crisis
partnership samples
Logit
an early- warning model