我国A股上市公司财务数据质量检查与评价的实证研究

发布时间：2019-01-19 16:42

【摘要】：统计数据是认识国情、了解民意、管理企业、监督市场的重要依据。随着我国资本市场的发展,监管机构和市场投资者对上市公司财务数据的需求与日俱增,对财务数据质量的要求越来越高,审计工作的繁杂程度不断加大,然而这些数据的真实性、可靠性却受到了更多人的质疑。因此,检查与评价上市公司财务数据质量,向监管机构和市场投资者提供真实可靠的信息,不仅是外部审计机构的责任,也是从事数据工作的统计学者的责任。本文拟根据数据本身的规律,运用统计学的方法对上市公司财务数据进行筛选和评价,从另一个角度分析上市公司财务数据质量的基本情况,为审计机构对上市公司的财务数据进行审计提供参考。本文所做的具有一定新意的工作如下：首先,对数据失真与数据造假两个基本概念进行了分析与解释,指出数据失真是实际数据与真实数据出现了偏离,数据造假则是主观造成数据失真的行为。论文还指出对出现异常的上市公司财务数据作进一步分析的必要性。这是因为,这种异常可能是真实情况的客观反映,对监管机构和市场投资者有着重大的意义,也可能是出现了数据造假。其次,本文以2012年沪深股市全部上市公司为样本,运用财务学和统计学的方法对财务指标进行筛选,建立财务数据检查评价指标体系,并利用因子分析法和Benford法则拟合检验法对样本公司进行分析,进而筛选出研究的指标。第三,选取2012年上市公司中被出具非标准审计意见的样本公司与相同数量的标准审计意见公司进行配对。并采用判别分析和Logistic回归模型对选取的研究样本公司进行判别归类。判别分析结果表明：判别分析对标准审计意见公司的分类正确率达78%,logistic模型法的结果与审计结果高度一致,达到了95%。第四,根据筛选出来的样本池,对其中的样本公司的具体财务指标作进一步检查与评价。根据财务报表性质的不同与可能获得的资料来源,对于资产负债表中的财务指标采用趋势模拟法分析,对于利润表中的财务指标则利用相关指标建立计量模型进行分析,在此基础上再对有关数据质量作出评价。
[Abstract]:Statistical data is an important basis for understanding national conditions, understanding public opinion, managing enterprises and supervising the market. With the development of the capital market in our country, the demand for financial data of listed companies is increasing day by day, and the quality of financial data is becoming more and more demanding, and the complexity of audit work is increasing. However, the authenticity and reliability of these data has been questioned by more people. Therefore, to check and evaluate the financial data quality of listed companies and to provide real and reliable information to regulators and market investors is not only the responsibility of external auditors, but also the responsibility of statisticians engaged in data work. According to the law of the data itself, this paper uses the statistical method to screen and evaluate the financial data of listed companies, and analyzes the basic situation of the financial data quality of listed companies from another angle. To provide a reference for audit institutions to audit the financial data of listed companies. The work of this paper is as follows: firstly, two basic concepts of data distortion and data falsification are analyzed and explained, and it is pointed out that the data distortion is the deviation between the actual data and the real data. Data falsification is subjective behavior that causes data distortion. The paper also points out the necessity of further analyzing the financial data of listed companies. That's because the anomaly could be an objective reflection of the truth, significant to regulators and market investors, or data fraud. Secondly, taking all listed companies in Shanghai and Shenzhen stock markets in 2012 as samples, this paper uses the methods of financial science and statistics to screen financial indicators and establish the evaluation index system of financial data inspection. The factor analysis method and Benford law fitting test method are used to analyze the sample company, and then the research index is screened out. Thirdly, the sample companies that were issued non-standard audit opinions in 2012 were matched with the same number of standard audit opinion companies. Discriminant analysis and Logistic regression model are used to classify the selected sample companies. The results of discriminant analysis show that the classification accuracy of discriminant analysis to the standard audit opinion company is 78% and the result of logistic model is in high agreement with the audit result, reaching 95%. Fourth, according to the selected sample pool, the specific financial indicators of the sample companies are further examined and evaluated. According to the nature of the financial statements and the possible sources of information, the financial indicators in the balance sheet are analyzed by trend simulation method, and the financial indicators in the income statement are analyzed by using the relevant indicators to establish a measurement model. On the basis of this, the quality of relevant data is evaluated again.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F276.6;F275;F832.51

【参考文献】