The Use of Unsigned Measures in Tests of Earnings Management- An example of Taiwan

• Main Authors: Ya-Wen Chang, 張雅雯
• Other Authors: 王泰昌
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/06504347889264743032

碩士 === 國立臺灣大學 === 會計學研究所 === 95 === According to the results of Hribar and Nichols (2006), this study uses the data in Taiwan to test the hypotheses that the volatility measures are correlated with unsigned measures of earning management. To reduce the risk of incorrectly rejecting the null hypothesis of no earning management, this study tends to identify firm characteristics that are correlated with earnings management estimates and empirically estimate the effect of different levels of correlation between the partitioning variables and the firm characteristic on rejection rate frequencies. The empirical analysis results of this study show that unsigned measures have high correlations with volatility of sales, volatility of earnings, and volatility of cash flows, despite the fact that prior work often ignores these characteristics. When operating volatility variables are correlated with the partitioning variable, the results show that the risk of incorrectly rejecting the null hypothesis of no earnings management is inflated. Finally, this study replicates a previous study measured by unsigned discretionary accruals. The results show that the volatility variables don’t affect the results from that previous study. The possible reason is that there is no or little correlation between volatility variables and the partitioning variable in that previous study.