摘要
一些上市公司财务报告舞弊现象层出不穷,严重侵害了投资者的利益。如何高效识别财务报告中的舞弊行为已成为目前研究的热点。文章在对已有的财务报告舞弊识别模型分析的基础上,提出了一种基于lasso-二元选择分位数回归的识别模型,并通过选取2010-2017年间240家上市公司年报数据作为样本,设计了16个财务指标进行了实证研究。结果证明,与传统的Logistic回归模型相比,lasso-二元选择分位数回归识别模型不但具备良好的变量选择能力,而且可以获得更好的识别效果,并能反映在不同的舞弊风险条件下各指标对于舞弊风险的影响,具有较高的应用价值。
The phenomenon of financial report fraud of listed companies emerges in endlessly,which seriously infringes the interests of investors.How to effectively identify fraud in financial reports had become a hot research topic.Based on the analysis of the existing financial report fraud identification model,this paper proposed a lasso-binary quantile regression identification model,and designed 16 financial indicators by selecting 240 listed financial Reporting data from 2010 to 2017 as samples.The results showed that compared with the traditional Logistic regression model,lasso-binary quantile regression recognition model not only had good variable selection ability,but also could get better recognition effect,and can reflect the impact of various indicators on fraud risk under different fraud risk conditions,and had higher application value.
作者
王威
杨朋之
WANGwei;YAN Pengzhi(Guilin Tourism University International Business School,GuiLin 541006;Guangxi Lixin Accounting Firm,GuiLin 541001)
出处
《上海市经济管理干部学院学报》
2019年第4期40-49,共10页
Journal of Shanghai Economic Management College