联系我们

论文查询结果	返回搜索

论文编号：	2428
作者编号：	2120082129
上传时间：	2010/6/10 14:22:31
中文题目：	我国上市公司持续成长性预测研究
英文题目：	A Prediction Study on the Continued Growth of
指导老师：	王全喜
中文关键字：	上市公司;持续成长性;因子分析;Logistic回归模型;预测
英文关键字：	Listed companies;Continued growth;Factor analysis;Logistic regression model;Prediction
中文摘要：	近些年来，许多对上市公司的预测研究都是围绕着上市公司财务困境与风险而展开的，并取得了大量的研究成果。由于财务困境的预测研究关注的是企业未来陷入财务困境甚至破产的可能性，这有利于投资者、债权人、企业管理者等利益相关者做出合理决策规避相应风险，但对于投资者、债权人等利益相关者明确做出合理有效决策及管理层明确目标采取相应措施改进管理实践却作用有限。基于此，本文将研究视角拓展到上市公司财务困境预测研究的另一面，即上市公司的持续成长性预测研究，由于投资者等利益相关者更加关注企业的未来，因此，对上市公司成长性的预测研究也就更加重要并且具有现实意义。 考虑到行业及规模对企业成长性的影响，从样本数量及数据可获得性出发，选取制造业344家A股上市公司作为研究对象，其中包括68家ST公司及276家非ST公司。同时，从财务角度及资本市场预期角度出发，选择19个指标构建预测指标体系。依据该指标体系运用因子分析法对非ST公司成长性样本采用2006年-2008年三年的数据进行了经济学意义和统计学意义上的界定，选出24家非ST公司样本与ST公司样本共同构成企业成长性异常样本。在此基础上，为了检验模型的预测准确性与稳定性，将总体样本划分为建模样本组和测试样本组。最后运用因子分析法对变量指标进行有效降维，并在此基础上运用Logistic回归模型构建企业成长性预测方程，同时对预测模型的准确性和稳定性运用检验样本组进行检验。 实证研究结果表明：(1)关于企业持续成长性的预测，要侧重其前瞻性，预测指标体系构建需要动态和静态相结合来体现趋势性。(2)构建预测模型的样本划分及分割点选择要考虑到现实情况的需要。(3)建模样本组与测试样本组的总体判别正确率都接近于80%，说明财务指标包含着企业未来发展趋势的信息，运用Logistic回归模型对企业持续成长性进行预测分析是可行的。(4)除了考察可量化的财务因素外，企业的持续成长性也需要考虑企业内部的非量化因素主要可归纳为四个维度，即产品、市场、技术与制度。 根据以上研究结果，本文对投资者、债权人、管理层等不同层面的利益相关者提出建议，以期能够为利益相关者做出明确合理决策提供决策依据。
英文摘要：	In recent years, many prediction researches focus on financial distress and bankruptcy risk of listed companies and obtain lots of academic achievements. Since this kind of research on prediction mainly pays attention on the probability of financial distress and bankruptcy risk of enterprises, it is beneficial for investors, creditors, managers and other stakeholders to make reasonable decisions and avoid risk. However, it isn’t helpful completely for stakeholders to make straightforward decisions. As the related aspect of existing prediction research, this paper extends the view of research to the continued growth of listed companies due to the focus of stakeholders for the future and prediction research on continued growth of listed companied is much more important and meaningful. The paper draws on existing research results, considering the influence of industry and size onto enterprises growth, and selects samples of A share listed companies of manufacturing. The sample contains 344 listed companies (68 ST companies and 276 non-ST companies).At the same time, the paper chooses 19 indexes to construct the index system of prediction from the view of financial and anticipation of capital market .Further, the paper uses factor analysis to define the growth of non-ST companies according to the index system. And select 24 non-ST companies form the sample of abnormal growth together with ST companies. In order to test the accuracy and stability of the testing model, the sample is divided into modeling sample group and test sample group. The paper finally uses factor analysis to reduce dimension, and uses Logistic regression model to construct prediction equation to predict continued growth of listed companies. The empirical results give several conclusions: (1) The prediction research on continued growth of listed companies requires to consider dynamic and static aspects in order to construct index system reasonably. (2) The definition of prediction samples and point of distinguish should take into real condition. (3) The result indicates that financial indexes could be used to predict the growth of companies and the accuracy of modeling sample group and test sample group is nearly 80%, which means Logistic regression model has excellent function to predict continued growth of companies. (4) Besides above quantized factors, non-quantized factors in companies shouldn’t be neglected, and they can be summarized as product, market, technology and system. According to above conclusions, the paper put forward different advices for different stakeholders, so as to provide helpful advices for reasonable decision making of different stakeholders.
查看全文：	预览  下载(下载需要进行登录)