我国制造业上市公司财务预警模型的实证研究

发布时间：2017-12-28 13:26

本文关键词：我国制造业上市公司财务预警模型的实证研究　出处：《中国海洋大学》2014年硕士论文　论文类型：学位论文

【摘要】：经济全球化使世界形势产生了剧烈的变化,经济的迅猛发展在给企业带来前所未有机遇的同时也带来了巨大的挑战,然而与之相匹配的完善的全球经济管理体制尚未形成,企业经营中所面临的诸多不确定性日益增加。世界顶尖企业诸如安然、宝丽来、安达信等的轰然倒塌,美国次贷危机的全面爆发公司,遭遇财务危机甚至破产的案例层出不穷,面对竞争激烈的市场环境,各类公司都加强了对其自身财务状况稳定性的关注,越来越注重完善公司财务预警系统。基于我国实际情况,财政部企业司于2009年也曾提出“企业从自身实际出发,探索建立企业财务预警机制”、“注重发挥企业财务信息的预警作用,加强企业财务风险监测预警工作”等要求。一般而言,任何企业的财务危机恶化都会经历一个过程,因此,在我国建立一个完善的、敏感的制造业上市公司的财务预警模型,对财务危机尽早反映、将其消灭于萌芽之中,对一般投资者、政府和银行等国家金融机构、注册会计师、债权人及潜在的债权人、经营管理者等利益相关者来说,显得尤为重要。论文以我国制造业上市公司作为研究对象,将公司因财务状况异常而被特别处理(ST或*ST)作为企业陷入财务危机的标志,选取2011-2013年我国沪、深两市首次被ST的27家财务危机公司作为样本组,同时采用一一配对的方法逐年选择27家财务健康公司作为配对组；初步选定反映企业偿债能力(包括短期偿债能力和长期偿债能力)、盈利能力、股东获利能力、发展能力、营运能力和现金流量能力等六大方面的62个变量指标,并通过柯尔莫哥洛夫-米诺夫正态分布检验(K-S检验)、显著性检验(独立样本T检验和曼-惠特尼-威尔克森检验),筛选出显著性的变量指标,再对显著性的变量指标提取主成分以降低变量的个数和消除多重共线性；然后建立了ST前1-5年的Logistic回归模型并对其拟合和预测效果进行检验,ST前1-5年的预测能力分别为80%、94.4%、75.0%、59.1%、63.6%；最后论文指出了财务预警模型的功能和适用范围。利用中国海洋大学图书馆的资源,在论文写作前期大量搜集、鉴别和整理国内外文献,反复推敲选定该论文题目以及写作思路。接下来论文围绕如何对我国制造业上市公司进行提前财务预警的问题,对财务预警的研究对象(即财务危机)、样本选取、模型构建和模型的运用等要素进行综合系统地分析,从而找出解决的可行方案。对于国内外财务预警研究的现状综述,界定财务危机的内涵框架,财务预警的现有模型和财务预警模型在企业中的运用,论文采用了描述研究法。另外,通过实证研究方法,利用SPSS统计分析手段,建立了Logistic回归的财务预警模型,并对其预测能力进行检验。论文在我国制造业上市公司中,连续选取三年的样本数据,并在ST前1-5年逐年配对选择财务健康公司作为配对组；然后针对每个公司选取了62个财务指标,分别建立了ST前1-5年的Logistic回归模型；数据容量较大、信息量丰富。初步选定的变量指标包括六大方面的62个财务指标,先后对其剔除了异常值、正态分布检验、显著性检验(符合正态分布的用独立样本T检验、不符合正态分布的用曼-惠特尼-威尔柯克森检验)和主成分分析；变量指标全面、处理得当。论文分别建立了ST前1-5年的Logistic回归模型,从而能够提前、及时发现企业存在财务危机的可能性,起到一定预警的效果。
[Abstract]:The economic globalization makes the world situation changed, the rapid development of economy and the hitherto unknown opportunities in brings to the enterprise has brought great challenges, however perfect global economic management system matching has not yet formed, enterprise faces many uncertainties increasing. The world's top companies such as Enron, Arthur Andersen, polaroid, the collapse of the outbreak of the U.S. subprime mortgage crisis, suffered a financial crisis or even bankruptcy cases emerge in an endless stream, facing the fierce competition in the market environment, various companies have strengthened the status of its own financial stability concerns, more and more attention to improve the company's financial early warning system. Based on the actual situation of China, the Ministry of finance of the Ministry of finance also proposed in 2009 that "enterprises start from their own reality, explore the establishment of enterprise financial early warning mechanism", "pay attention to giving full play to the early warning function of enterprise financial information, and strengthen the monitoring and early warning of enterprise financial risks" and so on. In general, any company's financial crisis will go through a process, therefore, to establish a perfect financial early-warning model, sensitive Manufacturing Listed Companies in China, the financial crisis as soon as possible, will reflect the nipped in the bud, to ordinary investors, government and banks and other financial institutions of the state, registered accountants, creditors and potential creditors, managers and other stakeholders, it is particularly important. Based on the listed companies of China's manufacturing industry as the research object, the company because of the abnormal financial condition by special treatment (ST or *ST) as the enterprise into a financial crisis, selects 2011-2013 China's Shanghai and Shenzhen two city for the first time by the 27 financial crisis companies ST as samples, using the method of paired each year choose 27 financial health companies as the matched group; preliminary selected reflect the enterprise solvency (including short-term solvency and long-term solvency), 62 variables are indicators of profitability, shareholder profitability, development ability, operation ability and cash flow ability six aspects, and through the Karl Region - Minov normal distribution test (K-S test), t-test (independent samples T test and Mann Whitney test, Wilkerson) screened significant variables, and variables of significant extraction The main components in order to reduce the number of variables and eliminate multicollinearity; and then establish a Logistic regression model of ST before 1-5 years and to test the fitting and prediction, prediction ability of ST before 1-5 were 80%, 94.4%, 75%, 59.1%, 63.6%; finally, the paper points out that the financial early-warning model and function applicable scope. The Ocean University of China library resources, large collection, in writing papers early identification and sorting of literatures at home and abroad, the thesis selected batted and writing ideas. This paper focuses on how to advance the financial early-warning problem of manufacturing listed companies in China, the research object of financial early-warning (i.e. financial crisis), a comprehensive system of sample selection, construction model and use factor analysis, so as to find out the feasible solution. To summarize the current situation of financial early warning research at home and abroad, define the connotation framework of financial crisis, the existing models of financial early warning and the application of financial early-warning model in enterprises, the paper adopts descriptive research method. In addition, through the empirical research method, the SPSS statistical analysis method is used to establish the financial early warning model of Logistic regression, and its prediction ability is tested. The manufacturing industry listed companies in China, with continuous sample data for three years, and 1-5 years ago in the ST year by year selected financial health companies as the matched group; then selects 62 financial indicators for each company, respectively, the regression model was established Logistic ST 1-5 years ago; the rich amount of information, large data capacity. Variables selected 62 financial indicators including six aspects, has removed the abnormal value, normal distribution test and significance test (with normal using independent samples T test, the distribution does not meet the normal distribution by using the Mann Whitney Weill Kirk Sen test) and principal component analysis; the variable index comprehensive, proper treatment. The Logistic regression models in the first 1-5 years of ST were established respectively, so that we can find the possibility of financial crisis in advance and in time, and play a certain early warning effect.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F425;F406.7

【参考文献】