我國銀行業盈餘管理行為、工具與還原之分析

Analyses of the Behavior, Tools and Recovery of Earnings Management in Banking of Taiwan

吳淳渼(Tsun-Mei Wu)

指導教授 ： 戚務君

中原大學/商學院/會計研究所/碩士(2001年)

https://doi.org/10.6840/cycu200100001

摘要

論文摘要

本研究係以避免負盈餘、維持前期盈餘績效與符合強制性財務預測誤差為三個門檻，探討銀行業盈餘管理之行為，並且檢視應計項目是否為管理當局所採用之重要盈餘管理工具，在此假設成立下，藉由應計項目還原盈餘之本質。此外，本研究亦探討審計品質對盈餘管理行為之影響。

實證結果顯示盈餘數樣本分佈在實驗組銀行(即小額正盈餘銀行)與對照組銀行(即小額負盈餘銀行)產生不連續情形，亦即銀行業管理當局會為了避免負盈餘而從事盈餘管理，且在應計項目的行為上，裁決性應計項目與壞帳費用為重要的盈餘管理工具，經由扣除裁決性應計項目變動數與壞帳費用變動數後之還原盈餘數樣本分佈上，發現小額正盈餘銀行區間與小額負盈餘銀行區間並未產生不連續情形，故還原後之盈餘數樣本分佈呈現較平滑趨勢；財務預測誤差率樣本分佈在實驗組銀行(即財務預測誤差小額低於20%絕對值限制銀行)與對照組銀行(即財務預測誤差小額高於20%絕對值限制銀行)產生不連續情形，亦即銀行業管理當局會為了符合財務預測誤差率而從事盈餘管理，且在應計項目的行為上，裁決性應計項目變動數與買賣票券損益變動數亦為盈餘管理工具，故經由扣除裁決性應計項目變動數與買賣票券損益變動數後之還原財務預測誤差率樣本分佈，發現財務預測誤差小額低於20%絕對值限制銀行區間與財務預測誤差小額高於20%絕對值限制銀行區間並未產生不連續情形，故還原後之財務預測誤差率樣本分佈呈現較平滑趨勢；但在盈餘變動數樣本則無上述現象產生。此外，審計品質對於銀行業為避免負盈餘、維持前期盈餘績效與符合財務預測誤差門檻之盈餘管理行為並無影響。

關鍵字

盈餘管理 ； 審計品質 ； 強制性財務預測誤差 ； 應計項目 ； 門檻

並列摘要

Abstract

This study identified earnings management to meet each of three thresholds: avoiding earnings losses, sustaining last year’s earnings performance, and exceeding mandatory management forecast errors. We explored the behavior of earnings management in banking and examined whether the accrual items were used to be the important managed tools. Under this hypothesis established, we recovered the nature of earnings by accrual items. Furthermore, we also explored what the audit quality impacted on the behavior of earnings management.

Our empirical result was that the density of the distribution of earnings discontinued between the treatment group (the banks had small positive earnings) and the controlled group (the banks had small negative earnings). Therefore, the banking managed earnings to avoid earnings losses. In characteristics of accrual items, discretionary accruals or bad debts were the important tools. By reducing the changes of discretionary accruals or bad debts to manage earnings, the distribution of recovery earnings continued between the treatment group and the controlled group. That is the distribution of recovery earnings became smoother. Second, we found that the density of the distribution of mandatory management forecast errors discontinued at the treatment group (the banks’ forecast errors had below absolute value of 20%) and the controlled group (the banks’ forecast errors had under absolute value of 20%). In other words, the banking managed earnings not to exceed the threshold of mandatory management forecast errors. In characteristics of accrual items, discretionary accruals or security exchanging gains (losses) were the important tools to manage earnings. By reducing the changes of discretionary accruals or security exchanging gains (losses), the distribution of recovery mandatory management forecast errors continued between the treatment group and the controlled group. Therefore, the distribution of recovery mandatory management forecast errors became smoother. However, the earnings changes samples didn’t have the above-mentioned phenomena. Besides, audit quality didn’t affect the behavior of earnings management in the banking of Taiwan, either.

並列關鍵字

accrual items ； thresholds ； earnings management ； mandatory management forecast errors ； audit quality

參考文獻

金成隆、林修葳與張永芳，強制性財務預測誤差與盈餘管理關係：20﹪門檻限制影響之研究，中國財務學刊，第七卷，第一期，民88，4月： 59-96。

夏侯欣榮，台灣增資新股上市後長期績效之整體研究，管理評論，第19期，民89，5月：1-33。

劉啟群，我國金融業會計損益項目季別分析：盈餘管理間接測試法，會計評論，第31期，民88，7月： 63-79。

李建然，影響台灣上市公司自願性盈餘預測行為因素之研究，政治大學會計學系博士班博士論文，民86。

Bartov, E. Timing of asset sales and earnings manipulations. The Accounting Review . June 1992：840-855.