为了提高审计报告的信息含量、满足资本市场的信息诉求，我国财政部于 2016年年底颁布了新的审计报告准则，新准则的核心内容是要求审计师在审计 报告中披露关键审计事项，即审计师应当选择对当期财务报告审计影响最大的事 项报告为关键审计事项，并就所选原因及审计应对在审计报告中进行说明。为了 检验关键审计事项信息是否具有增量信息价值，本文首先借助我国新审计报告准 则先行在A+H股上市公司的2016年年报审计中试点所创造的准自然实验环境， 采用倾向得分匹配和双重差分模型相结合（PSM-DID）的方法，检验关键审计事 项信息披露对分析师盈余预测质量的影响。接着，揭示关键审计事项的风险信息 本质，由此出发对上市公司2016年和2017年关键审计事项信息中的风险信息量 进行统计，探究关键审计事项信息中的风险信息量与分析师盈余预测质量之间的 关系，并考虑公司信息透明度对上述关系的调节作用。 本文的研究结果表明，首先，关键审计事项信息披露显著提高了分析师盈余 预测准确度、降低了预测分歧度，即关键审计事项信息披露提高了分析师盈余预 测质量，证明了关键审计事项信息具有增量信息价值。其次，关键审计事项信息 中的风险信息越多，分析师盈余预测准确度越高、分歧度越低，说明审计师对关 键审计事项中隐含的风险解释得越充分，分析师在做出盈余预测时面临的不确定 性越低，盈余预测质量越高。进一步地，本文还发现关键审计事项信息中的风险 信息量与分析师盈余预测质量之间的关系主要体现在信息透明度较低的公司。 本文的研究不仅深化了对关键审计事项的增量信息价值的检验，还揭示了关 键审计事项的风险信息本质，由此分析了关键审计事项信息中的风险信息量差异 及其影响，拓展了关于关键审计事项信息内容分析的相关研究，进而为关注新准 则执行效果及问题的监管者及其他相关人士提供参考，并对关键审计事项的相关 研究有着借鉴意义。 关键词：关键审计事项，风险信息，预测准确度，预测分歧度 1本文受到教育部人文社会科学基金青年项目（项目编号：18YJC790145）的资助 II THE EFFECTS OF CRITICAL AUDIT MATTERS ON THE QUALITY OF ANALYSTS EARNINGS FORECASTS Abstract In order to enhance the information content of audit reports and meet the information demand of the capital market, the Ministry of Finance of China promulgated new audit reporting standards at the end of 2016. The core of new standards is to require auditors to disclose critical audit matters in audit reports, that is, auditors should select the matters that have the greatest impact on current financial report audit as critical audit matters, and explain the reasons and the audit countermeasures in the audit reports. In order to test whether critical audit matters has incremental information value, this paper firstly uses quasi-natural experimental environment created by the implementation of new auditing reporting standards in A+H-share companies in 2016, and adopts the method of propensity score matching and difference-in-differences model (PSM-DID) to test the impact of disclosing critical audit matters on the quality of analyst earnings forecasts. Secondly, this paper reveals the essence of risk information of critical audit matters, and examines the relationship between the content of risk information on critical audit matters and the quality of analysts earnings forecasts, and then consider the moderating role of company transparency of information on the above relationship. The results of this paper show that, firstly, the disclosure of critical audit matters significantly improves the accuracy of analyst forecasts and reduces the dispersion, indicating that the disclosure of critical audit matters improves the quality of analysts earnings forecasts, and proving incremental information value. Secondly, the more risk information contained in the critical audit matters, the higher the accuracy of analyst forecasts and the lower the dispersion of analyst forecasts, which indicates that the more auditors fully explain the risks behind critical audit matters, the lower the uncertainty faced by analysts in making earnings forecasts, the higher the accuracy of earnings forecasts and the lower the dispersion. Further, the research in this paper finds that the relationship between the risk information content of critical audit matters and the analysts earnings forecasts accuracy and dispersion is mainly reflected in companies III with low information transparency. The research in this paper not only deepens the test of incremental information value of critical audit matters, but also reveals the essence of the risk information of critical audit matters. Based on this, it analyzes the impact of the risk information content of critical audit matters, which expands the related research on the analysis of the information content of critical audit matters. It will provide a reference for regulators and other relevant persons who are concerned about the effectiveness and problems of the implementation of the new standard, and also provide a reference for the in-depth study of critical audit matters. Keywords: Critical Audit Matters, Risk Information, Forecast Accuracy, Forecast Dispersion IV 目 录 中文摘要 .............................................. I Abstract ............................................... II 第一章 绪论 .......................................... 1 1.1 研究背景.................................................... 1 1.2 研究意义.................................................... 3 1.2.1 理论意义................................................ 3 1.2.1 实践意义................................................ 3 1.3 研究问题与方法.............................................. 4 1.3.1 研究问题................................................ 4 1.3.2 研究方法................................................ 4 1.4 研究内容与创新.............................................. 5 1.4.1 研究内容................................................ 5 1.4.2 研究框架................................................ 6 1.4.3 研究创新................................................ 7 第二章 文献回顾 ....................................... 8 2.1 关键审计事项信息的相关研究.................................. 8 2.1.1 关键审计事项信息的定义.................................. 8 2.1.2关键审计事项信息披露对审计师的影响 ...................... 8 2.1.3关键审计事项信息披露对管理层及治理层的影响 .............. 9 2.1.4关键审计事项信息披露对投资者的影响 ..................... 10 2.1.5关键审计事项信息披露对分析师的影响 ..................... 12 2.1.6关键审计事项信息内容分析的研究 ......................... 13 2.1.7关键审计事项的相关研究述评 ............................. 13 2.2 风险信息披露的相关研究..................................... 14 V 2.2.1 风险信息量的计算....................................... 14 2.2.1 风险信息量与信息使用者决策............................. 14 2.3 分析师盈余预测质量的相关研究............................... 15 2.3.1 分析师盈余预测质量的评价依据........................... 15 2.3.2 影响分析师盈余预测质量的因素........................... 16 2.4 已有研究述评............................................... 17 第三章 研究假设 ...................................... 19 3.1关键审计事项信息披露对分析师盈余预测质量的影响 ............. 19 3.2 关键审计事项信息中的风险信息量与分析师盈余预测质量......... 20 第四章 研究设计 ...................................... 22 4.1 样本选择与数据来源......................................... 22 4.2 变量测量................................................... 24 4.2.1 解释变量............................................... 24 4.2.2 被解释变量............................................. 24 4.2.3 调节变量............................................... 25 4.2.4 控制变量............................................... 25 4.3 模型构建................................................... 26 第五章 实证结果与分析 ................................ 28 5.1 描述性统计分析............................................. 28 5.2 相关型分析................................................. 29 5.3 回归分析................................................... 32 5.3.1 关键审计事项信息披露对分析师盈余预测质量的影响：双重差分检 验........................................................... 32 5.3.2 关键审计事项信息中的风险信息量与分析师盈余预测质量：多元回 归分析....................................................... 33 5.4 稳健性检验...........