国际“四大”审计师与审计质量的再审视

发布时间：2019-01-11 09:46

【摘要】：本文利用2007-2015年A股上市公司的数据,选取了可操纵应计利润、出具非标准无保留意见的概率和审计费用作为审计质量的代理变量,运用Heckman的两阶段回归模型和倾向得分匹配模型的研究方法对审计师与审计质量中存在的自选择偏差问题进行再审视。运用Heckman的两阶段回归模型的结果表明审计师与审计质量中存在显著的自选择偏差问题;并通过倾向得分匹配方法来控制实验组和对照组中的客户的公司特征,比较匹配前和匹配后不同会计师事务所对应的两组公司的审计质量差异,从而得出实验结果。研究结果发现:(1)通过Heckman的两阶段回归模型,表明审计师与审计质量中存在显著的自选择相关问题。(2)通过倾向得分匹配模型,我们发现,以可操纵应计利润,出具非标准无保留意见的概率为审计质量代理变量时,匹配前,四大会计师事务所的审计质量显著高于非四大会计师事务所,但经过倾向得分匹配后,四大会计师事务所高于非四大会计师事务所审计质量的差异并不显著;但以审计费用作为审计质量代理变量时,匹配前,四大会计师事务所有显著高于非四大会计师事务所的审计质量,经过倾向得分匹配后,四大会计师事务所仍显著高于非四大会计师事务所的审计质量。(3)此外,为了消除倾向得分匹配模型设计的敏感性对实验结果的影响,本文按照相同的迭代过程分别对匹配比率、匹配比率和我非线性协变量的选取进行随机调整,在每一个审计质量代理变量下设计出3000种不同的模型来比较匹配前和匹配后审计质量差异,实验结果支持上述结论。
[Abstract]:Based on the data of A share listed companies from 2007 to 2015, this paper selects the operable accrual profits, the probability of issuing non-standard unqualified opinions and audit expenses as the proxy variables of audit quality. This paper uses Heckman's two-stage regression model and tendency score matching model to re-examine the problem of self-selection deviation between auditor and audit quality. The results of Heckman's two-stage regression model show that there is a significant self-selection deviation between auditor and audit quality. And through the tendency score matching method to control the company characteristics of the clients in the experimental group and the control group, and compare the audit quality difference between the two groups before and after matching, so as to get the experimental results. The results show that: (1) through Heckman's two-stage regression model, there are significant self-selection correlation problems between auditor and audit quality. (2) through the tendency score matching model, we find that, in order to manipulate the accrual profit, When the probability of issuing non-standard unqualified opinion is a proxy variable of audit quality, before matching, the audit quality of the four major accounting firms is significantly higher than that of the non-Big four accounting firms, but after the tendency score matches, the audit quality of the four major accounting firms is significantly higher than that of the non-big accounting firms. The difference of audit quality between the four major accounting firms and non-big accounting firms is not significant. However, when the audit cost is used as the proxy variable of audit quality, before matching, the audit quality of the four major accounting firms is significantly higher than that of the non-big accounting firms. The audit quality of the four major accounting firms is still significantly higher than that of the non-major accounting firms. (3) in order to eliminate the influence of the sensitivity of the design of the propensity score matching model on the experimental results, In this paper, the selection of matching ratio, matching ratio and my nonlinear covariable are randomly adjusted according to the same iterative process. Under each audit quality agent variable, 3000 different models are designed to compare the difference of audit quality before and after matching. The experimental results support the above conclusions.
【学位授予单位】：武汉大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F239.4

【相似文献】