Three essays on contagion and network issues

This dissertation consists of three essays that examine contagion and network-related issues. In the first essay, we document that the contagion effect of earnings management is much broader than documented by Chiu, et al (2013). We find that firms are exposed to the spillover effect of earnings management by contagious firms when directors on their corporate boards also serve as directors on corporate boards of other firms at the same time and even when the contagious firms do not restate their reported earnings. These findings thus confirm that earnings management behavior, measured by absolute discretionary accruals, spills over from contagious firms to other firms with common corporate board directors. The contagion effect is stronger as the number of interlinked directors between contagious and affected firms increases, and it is especially stronger when interlinked directors have accounting expertise. We, however, do not find a stronger contagion effect of directors with finance expertise or any other non-accounting expertise. Furthermore, we do not find any significant difference in the contagion effect when institutional shareholdings in the Affected firms are significant or interlinked directors are independent or executive. The second essay examines the contagion effect of SEC comment letters through auditor offices. Using 7,451 initial comment letters for 10-K filings during the years 2005-2015, we discover that firms are more likely to receive comment letters after the SEC discovered a material deficiency in clarity or explanation in the 10-K filings of their industry peers audited by the same auditor office. Such contagion effect is only observable in auditor office whose clients receive accounting-related comment letters, and is attenuated by auditor office size. We also find that auditor offices develop expertise in resolving comment letters. Particularly, their client firms experience lower remediation costs (i.e., fewer topics in comment letters and fewer rounds of conversations) of addressing 10-K comment letters in the subsequent years. Overall, findings in this paper suggest that auditor offices are an important channel from which the SEC identifies firms in the comment letter review process. In the third essay, we collect data on the career paths of accountants who work at the SEC to examine the effect of revolving doors on their effort while at the SEC. We examine outbound accountants, that is accountants who leave the SEC for jobs at big four accounting firms. We also examine inbound accountants that are hired by the SEC from the big four accounting firms. We find no systematic evidence that the regulatory effort (as captured by the severity of the comment letters they issue) of outbound and inbound accountant is different from others. The exceptions are 1) outbound accountants that join big 4 firms issue less severe comment letters to clients of their prospective employer in the last year of service at the SEC and 2) inbound accountants that are hired from the big four accounting firms issue less severe comment letters to clients of their prior employer in the first year of their service at the SEC. The evidence points to some detrimental effect of revolving doors for accountants.Ph.D.Includes bibliographical referencesby Zhengjie Su