随着我国资本市场制度体系和基础设施的不断完善，上市公司可被视作外部性的受体暴露在由机构投资者、个人投资者、金融中介、监管者等各种主体参与的市场环境中，这些市场参与者的行为通过改变市场整体的信息质量等方式作用于上市公司，带来公司经营、治理、创新活动等方面的改变，进而影响上市公司质量。本文通过三个实证为主的工作，分别展示近年来中国市场的投资者、金融中介和监管者如何影响上市公司的决策和行为，并对这一过程给企业带来的成本进行测算和探讨。首先，在投资者方面，本文主要聚焦互联网背景下个人投资者关注对企业行为的影响。本文构建了衡量中国机构投资者关注度的“分心”指标并以此为控制变量，利用股吧阅读量和投资者关系维护平台发帖量两种不同的个人投资者关注度代理变量以及基于媒体报道构建的工具变量，建立了个人投资者关注与企业行为变化的因果关系，发现与“帝国建立”假说相一致，更多的个人投资者关注使得企业外部投资减少、内部激励派发增加、实际盈余管理和涉及诉讼也减少，本文还通过时间趋势分析发现，以上效应在投资者关系维护平台上线后得到进一步放大。其次，在金融中介方面，本文主要聚焦上市公司在证券分析师预测的每股盈余指标附近的集聚效应，并以此估算企业操纵财务指标的成本。本文充分讨论了上市公司向分析师一致预测水平集聚的动机和采用的具体手段，并估算了中国上市公司的边际债务成本作为参数准备，本文发现，以2010-2022年全市场样本的集聚效应估算，中国上市公司每操纵一单位每股盈余，需要付出31.15万元的成本，本文还对该现象的异质性、时间趋势以及中美市场的对比情况进行了分析。最后，在监管者方面，本文主要聚焦中国渐进式的退市制度改革，研究其对企业创新活动的影响。通过梳理2012年和2020年两次退市制度改革的新增退市指标，在识别具有退市隐患上市公司的基础上，利用倾向得分匹配和双重差分的方法，本文发现相比没有退市隐患的公司，具有退市隐患的公司在退市制度改革发布后显著减少了发明专利，特别是实用新型专利的申请数量。通过对两次回归结果的比较及基于营业收入退市指标集聚效应的成本测算，本文认为2020年退市制度改革推出组合型财务退市指标后，降低了给企业带来的监管成本。

With the continuous improvement of China’s capital market system and infrastructure, listed firms can be regarded as externality receptors exposed to the market environment with the participation of institutional investors, individual investors, financial intermediaries, regulators and other entities. The behavior of these market participants affects listed firms by changing the overall market information quality and other ways. Bring about changes in corporate actions, governance, and innovation activities, and then affect the quality of listed firms. Through three empirical studies, this paper shows how investors, financial intermediaries and regulators in the Chinese market affect the corporate actions of listed firms in recent years, and estimates the costs brought by this process to firms.Firstly, in terms of investors, this paper mainly focuses on the impact of individual investors’ attention on corporate actions under the background of the Internet. This paper constructs the “distraction” index to measure the attention of Chinese institutional investors and takes it as a key control variable. Using two different proxy variables of individual investors’ attention, namely the amount of Guba reading and the amount of investor relations maintenance (IRM) platform posting, and the instrumental variable constructed based on media reports, the causal relationship between individual investors’ attention and the change of corporate actions is established. Consistent with the “empire building” hypothesis, more attention from individual investors leads to a decrease in external investment, an increase in internal incentive payouts, and a decrease in real earnings management and litigation. Through time trend analysis, this paper also finds that the above effects are further amplified after the launch of the IRM platform.Secondly, in terms of financial intermediation, this paper mainly focuses on the bunching effect near the earnings per share (EPS) predicted by securities analysts, and estimates the cost of manipulating financial indicators. This paper fully discusses the motivation and specific means adopted by firms to bunch to the analyst earnings consensus. This paper estimates the Chinese firms’ marginal cost of debt as a parameter preparation. It is found that based on the bunching estimation of the whole market sample from 2010 to 2022, each unit of earnings per share manipulation by Chinese listed firms leads to a cost of 311,500 yuan. This paper also analyzes the heterogeneity of this phenomenon, the time trend and the comparison between the China and U.S. markets.Finally, in terms of regulators, this paper focuses on China’s gradual delisting system reform and studies its impact on corporate innovation activities. By combing the new delisting indicators of the delisting system reform in 2012 and 2020, on the basis of identifying listed firms with delisting risks, and using the method of propensity score matching (PSM) and difference-in-difference (DID) analysis, this paper finds that compared with firms without delisting risks, firms with delisting risks significantly reduce their invention patents after the delisting system reform, especially the number of applications for patents for utility model. Through the comparison of the two regression results and the cost calculation based on bunching estimation of business income delisting indicators, this paper believes that the combined financial delisting indicators introduced by the delisting system reform in 2020 will reduce the regulatory costs brought to firms.