基于案例推理的财务报告舞弊识别研究

李清; 任朝阳

基于案例推理的财务报告舞弊识别研究

引用本文：李清，任朝阳.基于案例推理的财务报告舞弊识别研究[J].财经理论与实践,2015,(3):84-89

摘要点击次数: 939

全文下载次数: 0

作者	单位
李清，任朝阳	（吉林大学商学院，吉林长春130012）

中文摘要:上市公司财务报告舞弊对资本市场建设和投资者利益保护掣肘已久。以2003~2012年我国沪深A股财务报告舞弊公司及其配对公司为样本，依据“舞弊钻石”理论和现有文献提取舞弊识别指标，基于案例推理思想构建财务报告舞弊识别模型。研究发现，较之于非舞弊公司，舞弊公司通常面临更大的盈利压力和偿债压力，成长能力更弱，事务所变更现象更为普遍，被出具非标准审计意见的概率更高。基于仿兰氏距离的案例推理模型对测试集样本的识别总正确率为66.7%。

中文关键词:财务报告舞弊案例推理熵值法 K近邻法

Using Case Based Reasoning For Financial Reporting Fraud Detection

Abstract:The financial reporting fraud committed by a public company has brought heavy impact on the construction of capital market and the protection of investors' interest. Using a sample of 2003~2012 domestic A-share listed companies, extracting factors about financial reporting fraud detection with the guidance of "The Fraud Diamond" theory and existing literature, this paper constructed the case based reasoning model for financial reporting fraud detection. This paper finds that fraudulent firms are more likely to be faced with greater earnings and debt pressure, weaker growth ability than non-fraudulent firms. This paper also finds that fraudulent firms have higher frequency of auditor changes and are more likely to get unqualified audit appraisals.The optimal case based reasoning model based on the Imitation Lance distance gets the hit ratio of 66.7% on the test set.

keywords:Financial reporting fraud Case based reasoning Entropy right method K-nearest-neighbor method

查看全文 查看/发表评论 下载pdf阅读器