摘要
本文选择2004—2012年间被证监会处罚的48家处于财务困境中的财务报告舞弊上市公司和48家正常公司为样本，以舞弊前三年到舞弊年度为研究区间，利用配对样本T检验和Wilcoxon符号秩检验在选取的31个财务指标中筛选存在显著差异的指标，通过Logistic回归技术构建基于财务困境的上市公司财务报告舞弊预警模型。模型对舞弊公司预警的有效性高达91．7％，综合有效性达到88．5％，取得了显著的预警效果。
This paper selects 48 listed companies in financial distress which received enforcement actions from the Chinese Security Regularity Commission due to financial fraud and 48 normal listed companies between 2004- 2012 as sample, and selects the third year before the fraud detection to the fraud year as research period. Using Paired-Samples T Test and Wilcoxon Signed-Rank Test, the paper filters the index of significant differences from the selected 31 financial in dexes. Through the Logistic regression technique, the paper builds a financial reporting fraud early warning model of listed companies based on financial distress. The effectiveness of the early warning model of fraud company reaches as high as 91.7% ,and comprehensive efficiency up to 88.5% ,significantly warning effect is obtained.
出处
《经济与管理研究》
CSSCI
北大核心
2013年第10期116-121,共6页
Research on Economics and Management
基金
国家自然科学基金项目"基于数据挖掘的上市公司会计舞弊识别问题研究"(70872082)
国家自然科学基金项目"产权视角下的审计师声誉机制及其经济后果研究"(71272189)
教育部人文社会科学研究项目"内部控制与独立审计的耦合效应研究:财务重述视角"(12YJA790193)
关键词
上市公司
财务困境
财务报告舞弊
预警模型
Listed Company
Financial Distress
Financial Reporting Fraud
Early Warning Model
