摘要
越来越多的投资者通过证券市场对上市公司进行投资，上市公司如果出现财务危机而投资者不能及时掌握相关信息，预警危机发生，将会面临巨大的风险，造成经济损失，甚至影响整个社会的稳定。通过参考国内外文献，选取财务、公司治理、总体经济和经营效率51个变量，2002～2005年中国A股上市ST与非ST公司378个样本，利用DEA数据包络分析法、Logit模型和类神经网络模型，并进行拟合优度与假说的检测及实证．确定并构建出本研究中最优的跨期财务危机预警综合模型。
Nowadays, in China, more and more investors invest in listed companies through the stock market. If financial crisis occurs in a listed company, investors of any kind, without obtaining relevant information in time, will face enormous risk, which will probably result in personal financial losses and even in threat to the stability of the whole society. While building models, on the basis of those research and literature reviewed, besides financial ratio variables, we add corporate governance variables, macroeconomic variables and business operation efficiency variables sequentially and progressively to establish synthesized models. Then the related data of 126 ST companies and 252 non-ST matching companies listed as A-Shares in China, covering the time span from 2002 to 2005, is processed to help construct models through statistical procesures accordingly. By employing DEA, Logit model and neural network model, we apply the data progressively, yet employ the models respectively. An optimum comprehensive financial crisis predicting model is constructed and confirmed through above mentioned processes. At the end, the optimum Logit Model IV is further verified to be robust in predicting financial crisis by applying the data of 1379 A-Share listed companies in the year 2006 as testing sample.
出处
《山东大学学报（哲学社会科学版）》
CSSCI
北大核心
2007年第6期130-136,共7页
Journal of Shandong University(Philosophy and Social Sciences)
