Analysis of the Effectiveness of Accrual Models and Market Reactions to Earnings Management through Restatement Events

• Main Authors: Ying-Wei Hsieh, 謝穎維
• Other Authors: none
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/31276457004073473049

碩士 === 國立中正大學 === 會計與資訊科技研究所 === 98 === There are three objectives in this research. First, we make a research on that whether the result from modified Jones Model with ROA is a good index of the earnings management amount of a company. Second, we look into the effectiveness of predicting whether a company would have financial report restatement by using the result of modified Jones Model with ROA. Third, we study whether the investors’ sophistication would impact the mechanism of market reaction to a company’s earnings management. According to Jones et al (2008), Estimated Discretionary Accruals (EDA) from Jones Model could be an important indicator of the involvement of a company’s earnings management. However, the result of that study also shows that a gap exists between the EDA and Actual Earnings Management Amount (AEMA, it is deflated by total asset of the beginning period in order to match with the unit of EDA). Besides, in a capital market overwhelmed by financial report restatement, it is a critical issue to help investors identify the earnings management by the company. By making deeper investigation of the related topic, investors can avoid the loss from the company’s fraud. With these incentives, we build up the first and second objective of this research. The study of Balsam et al. (2002) indicates that the mechanism of market reaction to a company’s earnings management would be impacted by the investors’ sophistication. This phenomenon is due to the information advantage of more sophisticated investors. With the above inference, we hypothesize that investors with different sophistication level would also have different degree of accuracy in predicting the AEMA of a company. As a consequence, it is our hypothesis that the relationship between a company’s stock price fluctuation and AEMA (or EDA) would be impacted by the investors’ sophistication level. For example, since more sophisticated investors possess more information advantages, they could predict AEMA before the company makes the restatement announcement. So the stock price tends to react to AEMA or EDA. On the other hand, as for less sophisticated investors, since they have fewer information advantages, they could not predict AEMA precisely unless the company finally makes the restatement announcement. So the stock price tends not to react to AEMA or EDA. Our research result shows that, to those companies who have restatement announcement, there is a significant gap between EDA and AEMA. Furthermore, the relationship between EDA and whether the company indeed announces the restatement in the end isn’t significant. On the other hand, the mechanism of market reaction to a company’s earnings management would be impacted by the investors’ sophistication level. For example, the relationship between a company’s stock price fluctuation and EDA would be impacted by the investors’ sophistication level. However, the relationship between a company’s stock price fluctuation and AEMA would not be impacted by the investors’ sophistication level. Besides, if we categorize the companies into different groups by investors’ sophistication, none of the groups shows the stock price fluctuation have significant relationship with AEMA.