基于时间序列的虚构利润类舞弊特征识别与分析

发布时间：2018-01-01 00:23

本文关键词：基于时间序列的虚构利润类舞弊特征识别与分析　出处：《北京邮电大学》2015年硕士论文　论文类型：学位论文

【摘要】：21世纪伊始,舞弊问题似乎已成为全球性的焦点问题,会计丑闻日益摧毁公众信心。2013年10月,证监会执行了有史以来规模最大的财务核查,引起业内不小轰动,导致268家拟上市公司退出上市申请,占比超过百分之三十。这证实了即使是所谓的发展势头良好的准上市公司,仍然存在不容忽视的财务问题,甚至舞弊行为。愈发严重的舞弊趋势,不仅给国家和各类组织带来严重损失,更使投资者和公众对资本市场的信心蒙受打击。要对上市公司财务舞弊行为实施有效的监管,必须从根源上剖析其舞弊动因和手段,探究舞弊公司与正常公司相比的显著异常特征,从而有的放矢,提高监管效率。本文以舞弊方式最为典型、舞弊性质最为恶劣的虚构利润类舞弊为切入点,探究舞弊公司与非舞弊公司之间的异常特征。在现有研究成果的基础上,本文引入时间序列的概念,对连续四年的舞弊数据进行分析,并将被众多学者忽略的财务报表科目纳入研究范围,力求从根源上识别舞弊特征。本文选择了自有违规记录以来到2013年12月31日,在我国沪深两市A股上市的舞弊公司,首先从总体上分析舞弊的行业特征、年份特征和类型特征,然后筛选出虚构利润类舞弊公司,并选取同等数量的正常公司,通过独立样本t检验和Mann-Whitney U检验两种方法,检验全部样本在舞弊当年和舞弊前三年的数据是否在财务报表科目和财务指标两个层面上存在显著差异,筛选出存在显著差异的科目和指标并进行相关分析和解释,并探究了两者的相关性,得出了有意义的结论。最后,利用科目和指标的检验结果建立了基于时间序列的Logistic回归模型,为监管机构和组织提供了新的舞弊识别参考依据。
[Abstract]:Since the beginning of the 21st century , fraud seems to have become a global focus , accounting scandals have increasingly destroyed public confidence . In October 2013 , the SFC had implemented the largest financial verification of the largest scale , accounting for more than 30 percent . This confirmed that even the so - called quasi - listed companies with good development momentum still have a financial problem that cannot be ignored , even fraud . On the basis of the present research results , this paper introduces the concept of time series , analyzes the fraud data for four consecutive years , and puts forward the financial statements subject to be ignored by many scholars into the research scope , aiming at identifying the characteristics of fraud from the root causes . On December 31 , 2013 , the paper selects the companies listed on Dec . 31 , 2013 in Shanghai , Shenzhen and Shenzhen . First of all , we analyze the industry characteristics , year characteristics and type characteristics of the fraud , then select the fictitious profit - based fraud companies , and select the same number of normal companies .

【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F275

【参考文献】