Listed Companies’ Networks and Firms’ Auditing Behaviors: Evidence from Textual Analysis on Annual Report

Wu Kaiwen

Journal of Shanghai University of Finance and Economics Vol. 21, Issue 03, pp. 74 - 90 (2019)  DOI:10.16538/j.cnki.jsufe.2019.03.006

Summary

China is a highly relationship-based economy. The widely-used relationship-based transactions form the structure of the capital market and the behaviors of information intermediaries. Theoretically, the company’s networks have an impact on auditing behaviors in three ways. First, the company’s networks increase the difficulty of evaluating the company’s specific investments. Second, the members within the same network repeat certain transactions indefinitely. Third, the company’s networks also make a broader scope of its stakeholders, which may increases the externality of its transactions. However, the company’s networks also provide more information sources for the auditors, which increase the informativeness of the financial report. Thus, this paper predicts that the company’s networks would increase the audit fee and earnings quality. Empirically, this paper investigates how the company’ networks affect auditing behaviors by using textual analysis on annual reports of all the Shanghai and Shenzhen A share listed companies from 2001 to 2012. We measure the networks of each company by using the number of network nodes from each company’s financial report with textual analysis. The empirical results show that the company’s networks increase the audit fee. The fee premiums are even larger for the company’s local networks and cross-industry networks. However, we don’t find any results of the network coefficient on earnings quality. In addition, further analysis shows that earnings quality is higher when the company has larger local networks and larger intra-industry networks. This finding suggests that it is difficult for auditors to understand the cross-province and cross-industry networks of the company. Further cross-sectional analysis shows that the effects of the company’s networks on fee premium and earnings quality are larger with more transparent information environment including more open local market, larger size of auditing firms, and more analyst following. Overall, this paper provides some empirical evidence to the relational contract theory from an auditing perspective, by using the data of the company’s networks to investigate how relational contracts affect auditing behaviors. This paper also has potential implications for the network study, as it offers some new mythologies to measure the company’s networks by using textual analysis.

Keywords

networks; textual analysis; audit fee; earnings quality