透過您的圖書館登入
IP:52.54.111.228
  • 學位論文

外部審計監督與盈餘管理

External Auditors' Monitoring and Earnings Management

指導教授 : 蔡彥卿

摘要


本文旨在探討外部審計監督功能與企業盈餘管理程度之關聯性,亦即藉由台灣股市特有的自結盈餘資訊,重新檢視外部審計在緩和企業盈餘管理的角色。雖然過去研究發現,審計品質愈高,則企業盈餘管理程度愈低(Francis, Maydew and Sparks 1999; Becker, DeFond, Jiambalvo, and Subramanyam 1998)。然而過去研究僅運用查帳後的財務資料進行分析,未能清楚瞭解外部審計過程是否確實產生抑制盈餘管理的作用,而且過去研究亦未處理自我選擇問題,其實證結果可能因此受到限制。本研究運用台灣股市特有的自結盈餘資訊,將其視為查帳前盈餘資訊,特別另行定義一項新指標,即自結盈餘與查帳後盈餘之差異(名為”Aggregate Audit Adjustments (簡稱AAA)”),並透過自我選擇之計量模式分析,檢驗外部審計人員在監督上市(櫃)公司避免投機性盈餘管理的能力,藉以重新檢視外部審計在緩和企業盈餘管理的角色。 本研究之實證結果發現,在控制自我選擇問題後,審計品質與盈餘管理之關係依然存在,意即審計品質較高的會計師,不僅報導較低程度的裁量性應計數(盈餘管理之代理變數),亦能減少客戶查帳前盈餘中的錯誤或不合宜的應計數。此外,為了檢驗審計品質之差異化,是否與企業經理人和外部審計人員間報導誘因之衝突或一致性有所關聯,本文進一步根據公司當期相對績效(與同業平均相比),劃分經理人誘因為調高或調降報導盈餘兩類誘因。本研究預期,在經理人具有調高盈餘之誘因時,正的AAA愈大則盈餘管理程度愈小(亦即兩者為負向關係);而在經理人具有調降盈餘之誘因時,AAA負的愈多則盈餘管理程度愈小(亦即兩者為正向關係)。實證結果顯示,不論是在調高或調降盈餘誘因時,AAA變數與盈餘管理的代理變數間均存在顯著的負向關係,且AAA變數與盈餘管理代理變數間的迴歸係數值相當大,表示審計人員查帳後的結果大多傾向於降低盈餘數字的調整。此一證據強烈支持一項結論,即審計人員在監督企業經理人之投機性盈餘管理行為上,皆傾向採取保守態度,特別是五大會計師事務所更是如此。 本研究之主要貢獻有三,首先,本文所發現之證據正好可與過去的相關研究形成互補不足之作用,並且可強化過去研究之發現,特別是Kim, Chung, and Firth (2003)、Francis et al. (1999)以及Becker et al. (1998)等研究。其次,本文採用更直接的指標(亦即AAA),衡量外部審計人員的努力程度及審計監督能力,相較於過去研究的衡量方式而言,更能強化研究的建構效度(construct validity)。最後,本研究運用兩種不同的自我選擇模式進行實證分析,相較於過去的OLS迴歸模式,亦增強研究本身在統計或計量模式上的效度。

並列摘要


The purpose of this study is to use a unique dataset about the information of pre-audit earnings (i.e., early self-reported earnings) from Taiwan stock markets to reexamine the external auditors’ role on alleviating corporate earnings management. A new measure is used to test the auditors’ ability on monitoring public companies away from opportunistic earnings management. This new measure is named the “aggregate audit adjustments” (hereafter called AAA) which is defined as the difference between pre-audited earnings and post-audited earnings. Prior studies such as Francis, Maydew and Sparks (1999) and Becker, DeFond, Jiambalvo, and Subramanyam (1998) have documented an association between audit quality and earnings management. By using a self-selection analysis and the AAA data as an intermediate variable to test the relationship between external auditor’s monitoring and earnings management, I find that even in the case of controlling the self-selection problem, the relationship between audit quality and earnings management still holds. The results imply that higher quality auditors can not only report a lower degree of discretionary accruals but also reduce a larger amount/magnitude of errors and unwarranted accruals included in their client’s pre-audit earnings. To test whether the conflict or convergence of reporting incentives between managers and external auditors interacts with audit quality differentiation, I further divide managers’ incentives to increase or decrease reported earnings on the basis of current relative performance. A firm is considered to have an income-increasing incentive or income-decreasing incentive, respectively, if its current relative performance is less or greater than zero. My hypotheses predict that firms with a higher level of positive aggregate audit adjustments (AAA) are associated with less earnings management in income-increasing incentives and firms with a higher level of negative AAA are associated with less earnings management in the case of income-decreasing earnings management. The empirical results show that the coefficients of AAA measures are all significantly negatively associated with the proxies of earnings management (e.g. DA or DCA) in either the cases of firms with income-increasing or income-decreasing incentives. The evidence strongly supports that auditors, especially the Big 5 auditors, tend to take more conservative actions through their aggregate audit adjustments on monitoring mangers’ opportunistic behavior. This study may contribute to the accounting literatures in two-fold. First, this study extends the stream of research like the works of Kim, Chung, and Firth (2003), Francis et al. (1999) and Becker et al. (1998) by providing evidence whether auditors can ameliorate earnings management through their aggregate audit adjustments on clients' pre-audit earnings. If auditors' aggregate adjustments are negatively associated with the measures of earnings management, the findings of the effects of external auditors' monitoring on earnings management can be articulately justified and then concluded. In deed, my evidence complements and reinforces the findings of prior research, in particular those provided in Kim, Chung, and Firth (2003), Francis et al. (1999) and Becker et al. (1998). Thus, my findings corroborate the relationship between audit quality and earnings management. Second, this study will cast a new light on the problems of construct validity and self-selection bias. Using a new indicator variable (i.e., AAA) to measure the auditors’ monitoring ability and two self-selection models instead of traditional OLS regression model to reexamine the research issue, this study not only improves its construct validity but also its statistical validity relative to this stream of research. In summary, this study reinforces its findings and thus corroborates our understanding on the relationship between external auditors' monitoring and earnings management.

參考文獻


37. Lee, J. and J. Chen. 2004. The effect of audit client’s importance on magnitude of earnings management: from the perspective of audit groups with the Big Five. International Journal of Accounting Studies (Taiwan) 38 (January): 59-80. (in Chinese)
8. Chen, Y. C. and C. Y. Huang. 2001. An empirical study of the relationship between earnings forecast and earnings management in Taiwan. Review of Securities and Futures Markets 13 (2): 97-121. (in Chinese)
1. Bartov, E., F.A. Gul, and J.S.L. Tsui. 2001. Discretionary accruals models and audit qualifications. Journal of Accounting and Economics 30 (4): 421-452.
2. Beatty, R. 1989. Auditor reputation and the pricing of initial public offerings. The Accounting Review 64 (October): 693–709.
4. Beneish, M. D. 2001. Earnings management: A perspective. Managerial Finance 27 (12): 3-17.

延伸閱讀