基于KMV模型的中小企业板上市公司财务预警研究

发布时间：2018-01-16 14:05

本文关键词：基于KMV模型的中小企业板上市公司财务预警研究　出处：《青岛理工大学》2013年硕士论文　论文类型：学位论文

【摘要】：我国已步入后金融危机时代，，受到2010年欧债危机的影响后，以出口为主的很多中小企业纷纷倒闭，出口受到阻碍，银行的信贷违约案例增加，对金融业造成冲击。中小板上市公司成立的时间短，规模较小，导致业绩评价困难，更易出现如操纵市场、内幕交易等行为，爆发财务危机的可能性也更大。因此需要建立积极有效的财务预警模型系统，用以合理避免企业的财务危机，保护投资者的权益，维持资本市场的稳定，尤其是能够解决目前中小板上市公司普遍存在的银行不良资产率高的问题。目前我国对财务预警模型领域的研究依赖传统的财务指标，因为我国的违约数据库系统尚未建立，部分的国内学者试图研究将KMV模型引入至预警模型，但是这方面的研究大多限于在我国KMV模型能否有效适用方面，实际上信用风险度量层面的最新研究成果尚未真正与我国企业的财务预警模型相结合。本文在前人研究的基础上，将传统的财务指标预警模型与信用风险度量模型相结合，引入KMV的违约距离与预期违约率指标，以期提高财务预警模型的预测精度。本文在阐述Logistic逻辑回归模型及KMV模型的基础上，把KMV模型中的违约距离（DD）引入到Logistic模型，建立一套新的财务预警模型，以时间因素为纵向，比较加入违约距离前后两种模型在预测的准确性的区别，考察违约距离对财务预警模型预测和解释能力的影响，再分析违约距离变量对模型2的影响。首先根据我国资本市场的实际情况对KMV模型的参数修正：一是股东权益市场价值,我国存在股权分置现象，除了流通股外还存在非流通股，非流通股大多处于股权分置改革的限售期。非流通股的定价依据净资产定价法；二是违约点DP，当公司的资产价值低于某一临界值时，对债权人和公司股东会出现违约，与这一临界值相应的资产价值称为违约点DP。其次对KMV模型的输出结果违约距离及依据违约距离计算得出的预期违约率做相关性分析和显著性检验。经过相关性分析，DD和EDF呈负相关关系，而显著性检验说明DD和EDF在0.05的显著性水平下均存在显著性差异。再次对22个基础指标进行筛选。本文将通过正态性检验及显著性检验筛选出能够对模型有较好代表的自变量，再通过因子分析剔除具有多重共线性的自变量，选出更具有代表性的变量。最后构建回归模型。基于上述的研究，利用Logistic回归模型针对2010年和2011年分别构建基于财务指标的预警模型和引入违约距离的预警模型，通过对两种判别模型在横向和纵向的对比分析，得出结论：引入KMV模型可以提高财务预警模型的解释和判别能力，提高预测的准确性；经检验新的预警模型对中小企业板上市公司危机预警效果良好。
[Abstract]:Our country has entered the post-financial crisis era, by the European debt crisis in 2010, many small and medium-sized enterprises mainly export have closed down, export has been hindered, the bank credit default cases increased. The small and medium board listed companies have short time and small scale, which lead to the difficulty of performance evaluation, such as market manipulation, insider trading and so on. Therefore, it is necessary to establish an active and effective financial early-warning model system in order to reasonably avoid the financial crisis of enterprises, protect the rights and interests of investors, and maintain the stability of the capital market. In particular, it can solve the problem of high non-performing assets in banks. At present, the research on financial early-warning model in China depends on the traditional financial indicators, because the default database system has not been established in China. Some domestic scholars try to introduce KMV model into the early-warning model. However, most of the research in this area is limited to whether the KMV model can be effectively applied in our country. In fact, the latest research results in the level of credit risk measurement have not been really combined with the financial early-warning model of Chinese enterprises. Combining the traditional financial index early warning model with the credit risk measurement model, this paper introduces the default distance and the expected default rate index of KMV in order to improve the prediction accuracy of the financial early warning model. Based on the description of Logistic logical regression model and KMV model, this paper introduces the default distance (DDD) of KMV model into Logistic model. A new financial early warning model is established to compare the prediction accuracy of the two models before and after adding the default distance, taking the time factor as the vertical. The influence of default distance on forecasting and explaining ability of financial early warning model is investigated, and the influence of default distance variable on model 2 is analyzed. First, according to the actual situation of the capital market in China, the parameters of the KMV model are revised: first, the market value of shareholders' rights and interests, there is the phenomenon of split share structure in China, in addition to circulating shares, there are also non-tradable shares. Most of the non-tradable shares are in the restricted period of the split share structure reform. The pricing of non-tradable shares is based on the net assets pricing method. The other is the default point DP.When the company's asset value is below a certain critical value, it will default on creditors and shareholders. The corresponding asset value corresponding to this critical value is called default point DP. Secondly, the correlation analysis and significance test of the default distance between the output of KMV model and the expected default rate calculated according to the default distance are made. Through the correlation analysis, DD and EDF are negatively correlated. The significance test showed that there were significant differences between DD and EDF at the significant level of 0. 05. Again, 22 basic indexes were screened. In this paper, the independent variables which can represent the model were screened by normal test and significance test. Then factor analysis is used to eliminate the independent variables with multiple collinearity, and the more representative variables are selected. Finally, the regression model is constructed. Based on the above research. The Logistic regression model is used to build the early warning model based on financial indicators and the early warning model based on default distance for 2010 and 2011 respectively. Through the comparative analysis of the two discriminant models in the horizontal and vertical, it is concluded that the introduction of KMV model can improve the interpretation and discriminant ability of the financial early-warning model and improve the accuracy of prediction; The new warning model is proved to be effective for the crisis warning of SMEs listed companies.
【学位授予单位】：青岛理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F275;F832.51

【参考文献】