公司治理视角下基于粗糙集的中国上市公司经营困境评价模型及实证研究

项目名称： 公司治理视角下基于粗糙集的中国上市公司经营困境评价模型及实证研究

项目编号： No.70872033

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 金属学与金属工艺

项目作者： 王宗军

作者单位： 华中科技大学

项目金额： 25万元

中文摘要： 对公司经营困境成因和评价的研究具有重要的理论和现实意义。目前国内外学者在研究中很少深入细化讨论公司治理因素如何对经营困境产生影响。本研究从公司治理视角出发，构建了我国上市公司经营困境的多因素综合评价模型。以1998年我国开始实行特别处理（ST）制度以来至2010年底被特别处理的所有上市公司及其配对公司为研究样本，根据研究样本t-1~t-3年数据，分别检验了股权结构、内部董事、董事会独立性、治理会议有效性、管理层激励及审计机制对公司财务困境的影响，并分别构建基于公司治理各个视角的粗糙集评价模型，然后构建公司治理视角下的经营困境多因素综合评价模型，对中国上市公司的经营困境进行评价。将评价结果与仅包含财务数据的粗糙集评价以及传统统计分析logistic评价方法进行对比的结果表明：加入公司治理因素变量后的粗糙集模型整体的分类识别率高于单纯只包含财务因素变量的粗糙集评价模型，相对于传统方法来说，应用粗糙集方法构建的模型分类识别率要远远高于logistic方法。另外，本研究进一步探索采用变精度粗糙集（VPRS）和变精度粗集神经网（VPRNN）模型对上市公司经营困境进行评价，也取得了较好的结果。

中文关键词： 公司治理；经营困境；经营困境评价；粗糙集

英文摘要： There is great significance on studying the factors and evaluation of financial difficulties both theoretically and practically .The research has been a Hot spot at home and abroad. Many scholars have already studies the effect of corporate governance on financial distress, but most of these researches are too simple, and little have analyzed how corporate governance affects companies financial distress status in detail. This paper establishes the multi-indexes competitive framework of financial distress evolution of Chinese listed companies. With the use of a sample through years 1998-2006, and their data covering the periods t-1~t-3, this paper tests the effects of ownership structure, insider directors, board independence, board effectiveness, managerial incentives and audit system on financial distress. Then we establish the multi-index competitive evaluation model based on each part of corporate governance characteristics and the whole corporate governance characteristics with the use of rough set methodology. And then we make use of this model to carry out financial distress prediction of Chinese listed companies, thus to provide a strong support for Chinese listed companies to understand their financial status correctly. The empirical results in year t-1~t-3 indicates that: there are higher classification ratios after adding corporate governance characteristics in the model. Comparing to the traditional regression model, rough set methodology has great classification power than traditional method. In addition, this research explored the use of variable precision rough set (VPRS) and variable precision rough set neural network (VPRNN) model to evaluate financial difficulties, which achieved good results.

英文关键词： corporate governance；financial distress；financial distress prediction；rough set