基于现金流的上市公司财务危机预警研究

发布时间：2018-03-27 21:30

本文选题：Logistic逻辑回归分析模型　切入点：径向基神经网络模型　出处：《北方工业大学》2012年硕士论文

【摘要】：近些年来大量公司蜂拥上市,但是成功上市并不意味着企业的胜利,在证券市场上,很多上市公司由于经营不善,市场环境无法预测等原因而带上了ST的帽子,不但给投资方造成巨大的经济损失,还对证券市场的稳定发展埋下不稳定的因素,而如何预测上市公司的财务危机成了关联方急于解决的问题,本文意在建立一种有效的企业财务危机预警模型来解决这一问题。而在建立企业财务危机预警模型时,如何选择合适的指标建立模型一直是解决这一问题的前提,有效的财务指标通常来源于上市公司的资产负债表、利润表和现金流量表。每年大量会计报表丑闻的曝光,更多公司对资产负债表数据中和利润表中透漏的信息产生了严重的怀疑,更能反映企业真实财务运行状况的现金流指标越来越受到学者们的关注。本文首先从理论上验证了现金流指标在构建财务危机预警模型时的优越性和适用性,并利用线性方法即Logistic回归模型和非线性方法神经网络(径向基函数神经网络法)模型来构建基于现金流指标的企业财务危机预警模型,并阐述了所构造模型给上市公司带来的价值和效用,意在在繁冗的数据信息环境里,提取出有效的信息,在瞬息万变的市场环境中为企业提供及时的决策。本文共选取200多家运行良好的上市公司的财务数据和被特别处理(文中用ST代表)公司的财务数据,样本数据来自于多个行业,根据有关文献选取并构建了现金流量指标构建财务预警指标体系,并运用因子分析方法剔除指标间的共线性影响,分别采用Logistic逻辑回归分析对表现良好的上市公司与出现财务危机的公司进行了分类判断,并构建了财务危机预警模型,最终对检验的结果进行了有效的验证；又采用径向基神经网络模型建立财务危机预警模型,也进行了检验结果的验证,并对两个模型进行了比较。根据Logistic逻辑回归分析与神经网络模型的检验结果进行了应用上的评价,最后具体阐释了企业构建财务危机预警模型的程序和原则。
[Abstract]:In recent years, a large number of companies have flocked to the stock market, but the successful listing does not mean the success of the enterprise. In the stock market, many listed companies wear the name St because of poor management and unpredictable market environment. Not only does it cause huge economic losses to investors, but it also buries unstable factors in the stable development of the securities market. How to predict the financial crisis of listed companies has become a problem that related parties are eager to solve. The purpose of this paper is to establish an effective financial crisis warning model to solve this problem. Effective financial indicators usually come from listed companies' balance sheets, profit statements and cash flow statements. More companies are seriously sceptical of the information leaked in the balance sheet data and in the profit statement. More and more scholars pay more and more attention to the cash flow index which can reflect the real financial operation of the enterprise. Firstly, this paper theoretically verifies the superiority and applicability of the cash flow index in the construction of financial crisis warning model. Using linear method (Logistic regression model) and nonlinear neural network (radial basis function neural network) model to construct the enterprise financial crisis warning model based on cash flow index. The paper also expounds the value and utility of the model to the listed companies, which is intended to extract effective information in the redundant data and information environment, and to provide timely decision-making for the enterprises in the rapidly changing market environment. In this paper, we select the financial data of more than 200 well-run listed companies and the financial data of specially processed companies. The sample data come from many industries. According to the relevant literature, this paper selects and constructs the cash flow index to construct the financial early warning index system, and uses factor analysis method to eliminate the co-linear influence between the indexes. The Logistic logical regression analysis is used to classify the listed companies and the companies with financial crisis, and the early warning model of financial crisis is constructed. Finally, the results of the test are validated effectively. The radial basis function neural network model is also used to establish the financial crisis warning model, and the test results are verified. According to the results of Logistic logistic regression analysis and the test results of neural network model, the paper evaluates the application of the two models, and finally explains the procedure and principle of constructing the financial crisis warning model.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F275;F832.51;F224

【相似文献】