Two essays on empirical accounting research.

摘要

Essay I. Investment banking and earnings forecast in prospectus of Hong Kong initial public offerings. About 75 percent of firms going public on Hong Kong Stock Exchange voluntarily include earnings forecasts in their prospectuses for the period from 1997 to 2006. Almost all forecasts underestimate actual earnings, with an average of 7%. A unique listing requirement in Hong Kong is that the investment bank (underwriter) issue an opinion report on the earnings forecast in prospectus. We investigate the impact of the underwriter's commission rate, reputation capital, and conflict of interest on the forecast bias in prospectus, after controlling for the issuer's incentives. We find that the commission rate and reputation capital are negatively associated with the forecast bias. Using IPO size as proxy for future trading commissions generated from institutional investors, we find that IPO size is positively associated with forecast bias. We further find that underpricing is positively associated with forecast bias due to underwriter's influence. Overall, these results are broadly consistent with the predictions based on the agency theories of the underwriter by Baron (1982) and Loughran and Ritter (2002).;Key Words: Earnings Forecast, IPO, Underwriter, Conflict of Interest;Essay II: Are Outside Directors Privately Informed about Management Fraud? Recent accounting scandals raise concerns about outside directors, who in principle monitor management on behalf of shareholders. However, they may fail to effectively monitor management if they are uninformed about management behavior. Outside directors rely largely on information provided by management and managers who engage in fraud have especially strong incentives to hide their activities from outside directors. It is therefore an open question as to whether outside directors know about, or suspect, the existence of fraudulent activities before the frauds are made public. To address this question, we examine the turnover of outside directors vii during the fraud period. To avoid the potential labor market penalties or reputation costs, some outside directors who get prima facie evidence of fraudulent activities may select to voluntarily leave fraud firms. Another possibility is that those directors who probe management are forced out of the board, since the management in reality can determine who will be placed on the slate for election. Both cases would result in a positive association between the occurrence of fraud and the departure of outside directors. On the other hand, there would be no association between fraud and departure if the directors are completely unaware of fraudulent activities. The empirical results show that there is no association between fraud and turnover for outside directors as a whole. However, after we distinguish between outside directors who are either "independent" or "linked", we find a significantly positive association between fraud and turnover for independent directors. Thus, the results indicate that independent directors may have some information about the on-going fraudulent activities. However, the magnitude of the association is small, indicating that independent directors are not particularly well-informed about management fraud.;Key Words: Turnover, Independent Director, Accounting Fraud