UC Berkeley

This dissertation examines the incentives of executives to distort firm earnings and the resulting economic costs of earnings distortion. I first investigate the effect of improvements in the detection of earnings distortion on CEO incentives in the presence of career concerns and endogenous contracting. The theoretical model incorporates both the short-term and career concern incentives of executives to increase earnings. Then I allow the signal of earnings distortion to exogenously change. The model leads to prediction that early in the career, there is less of an incentive to increase earnings through earnings distortion. The model is evaluated empirically using firms' restatements of earnings. I consider how the change in the earnings distortion detection ratio due to the implementation of Sarbanes Oxley Act of 2002 (SOX) affects these short and long-term incentives to manipulate earnings. I unexpectedly find that both executives and boards of directors act as predicted if a restatement is a weaker signal of earnings distortion after SOX.

The next chapter also considers the effect of the Sarbanes Oxley Act. I look at how the additional costs and benefits of being a public company after SOX affect the likelihood of a firm going public or private. I use the fact that firms which have publicly available debt at the start of the implementation of the Sarbanes Oxley Act of 2002 must comply with many aspects of the law. I take a difference in differences approach to see the effect of Sarbanes Oxley on the decision to list or delist from an exchange. I find that fewer firms listed after the implementation of the act. Therefore, the costs of SOX may be larger than the benefits for the marginal public firm. I also find that marginally fewer firms are choosing to delist after SOX. The costs of choosing to delist seem to be higher for the marginal public firm. The difference may be due to additional reputational costs for the firm choosing to delist to avoid complying with SOX.

The final chapter examines if the market use signals from other firms in the same industry to evaluate the information risk of a given firm. In particular, if one firm in the industry has to restate their financial statements, does the information risk at other firms in the industry increase? I consider how the market responds to an earnings announcement for firms who have not had a restatement. Overall, I find that the market reacts less positively to positive earnings announcements if other firms in the industry have restated their accounting statements within the last year. In addition, I find that this change in the reaction is especially true for high growth industries.