【作者】 杨芳；

【导师】 陈收；

【作者基本信息】 湖南大学 ， 管理科学与工程， 2006， 硕士

【摘要】 上市公司财务欺诈问题已经日益成为一个不容忽视的问题,但其识别却具有一定的难度。本文借鉴国内外相关的研究成果,选择68家财务欺诈公司作为研究对象,分别从审计角度、财务状况角度、现金流量角度和公司治理角度识别财务欺诈,将最后得到的17个识别变量运用Logit方法建立了一个判断中国上市公司年报是否存在财务欺诈的综合识别模型。实证结果表明,该模型具有较好的识别能力,可以有效的帮助投资人和债权人、政府监管部门、审计部门正确识别上市公司的财务欺诈。文章分为四个章节:首先,本文提出选题背景,总结国内外研究现状,同时阐明研究目的、研究步骤及行文结构。为了说明本文对财务欺诈的界定标准,以及财务欺诈识别变量和识别方法选取的依据,第2章主要介绍财务欺诈的概念及界定、财务欺诈的原因、财务欺诈的识别变量和财务欺诈的识别方法。第3章分别从审计角度、财务状况角度、现金流量角度、公司治理四个角度识别财务欺诈,找到有效区分财务欺诈公司和非财务欺诈公司的识别变量。第4章将筛选出来的识别变量,运用Logit方法建立财务欺诈综合识别模型,并对模型的识别能力进行了检验;最后的结论对财务欺诈识别定量研究结果进行归纳总结,探讨本研究对识别财务欺诈的理论和实践意义,并指出本研究存在的局限及后续研究方向。本文的创新之处就是在实证中加入较为系统全面的现金流量指标,而且从多角度建立了财务欺诈综合识别模型。研究结果表明,财务欺诈综合识别模型回判精度达到80.9%,预测精度达到68.25%,具有较强的识别能力,可以有效地运用到实际工作中去,帮助各相关主体识别财务欺诈。更多还原

【Abstract】 Nowadays, fraudulent financial reporting has become a big problem which can not be ignored anymore. However, it is difficult to be detected. Using overseas and domestic research results for reference, this paper chooses 68 companies which have fraudulent financial reporting as research object, and then uses statistic method to get 17 ratios from four aspects: audit ratios; financial ratios, cash flow ratios, corporate governance ratios. Finally this paper develops a logit model on the basis of these 17 ratios. The prediction result indicates that the model has a good ability to detect the risk of fraudulent financial reporting in annual reports of China’s listed companies. Thus, the model can be applied to the practice effectively. This article is divided into four chapters:The first chapter introduces the background and literature survey of this research, and clarifies the purpose, process and structure of this paper. The second chapter introduces the concept of fraudulent financial reporting, the reasons of fraudulent financial reporting, identifying factors, and identifying models. The third chapter finds identifying factors which can detect the risk of fraudulent reporting effectively. Based on these factors, this paper develops a logit model, and verifies the identifying ability of this model. At last, The conclusion summarizes the result, probes into its academic and practical significance, and points out the limitation and direction of future of this empirical research.The innovations of this paper include two aspects: firstly, this paper adds cash flow ratios to the research; secondly, this paper sets up a model from comprehensive angles. The results of research show that the Logit model can achieve 80.9% accuracy in identifying fraud of the original samples, 68.5% accuracy in identifying fraud of the testing samples. Thus, this model can be used to identify the fraudulent accounting information successfully. The outcome of the research will provide academic and practical support to creditors, investors , government departments.更多还原