基于主成分分析的上市公司利润操纵识别模型  

A Recognition Model of Aggressive Earnings Management in Chinese Listed Companies Based on the Principal Components Method

作　　者：姚宏[1] 李延喜[1] 高锐[1] 

机构地区：[1]大连理工大学管理学院,辽宁大连116024

出　　处：《管理科学》2007年第5期83-91,共9页Journal of Management Science

基　　金：国家自然科学基金(70302008);霍英东教育基金会高等院校青年教师基金(101084)

摘　　要：从分析上市公司利润操纵的动机、主要手段及其影响的相关会计科目入手，设计26组有一定配比关系的财务指标，选取1995年～2004年中国证监会以利润操纵名义查处的全部上市公司为样本组，选取同行业规模相近且经营规范的公司为对比组，进行横向对比与相关分析；基于主成分分析法构建上市公司利润操纵行为的识别模型，确定利润操纵行为的识别边界，运用该模型对部分上市公司赢利真实性进行检验，为投资者识别上市公司的利润操纵行为提供一套科学可行的方法。On the basis of the target, the main means of the company＇s earning management and the accounting tides effected by the earning management, this paper designs 26 groups accounting index which having the comparability relation with each other. The paper selects the publics＇ accounting statements that were punished for the earning management by China Securities Regulatory Commission since 1995 -2004 as samples, selects the companies which are in the same industry, having similar asset size and steady management as the contrast samples. Based on the results of the study of the samples, the paper constructs the recognition model of aggressive earnings management, and makes it possible to recognize the listed companies applying aggressive earnings management.

关 键 词：信息失真 主成分分析 利润操纵 识别模型 

分 类 号：F830.9[经济管理—金融学]