會計裁量與實質交易之權衡與時點：台灣上市櫃公司之盈餘管理研究

Tradeoff and Timing Decisions between Accounting Discretions and Real Activities Manipulations: A Study on Earnings Management for Listed Firms in Taiwan

由於盈餘管理在瀕臨虧損之企業最易發生，本研究探討經理人為避免年度虧損，如何權衡「會計裁量」與「實質交易操作」之採用順序。本研究以年資料分析之結果發現：經理人為避免年度虧損，在年度中係先決定「會計裁量」、後決定「實質交易操作」，而且使用「會計裁量」相對於「實質交易操作」之可能性亦較高，此與Zang（2012）之結論有異。本研究進一步以季資料分析之結果發現，經理人為避免年度虧損，相對於前三季，第四季利用「實質交易操作」較「會計裁量」顯著為高，而且第四季使用「實質交易操作」，相對於「會計裁量」，可能性亦較高。Zang之結論僅限於第四季之盈餘管理，而不及於全年度各季，本研究之發現為目前文獻所未探究者。 Archived studies on earnings management in Taiwan generally investigate single real activity or accounting accruals. Unlike prior studies, this study examines multiple real activities and accounting accruals simultaneously for Taiwan's listed firms. We examine the sequence of exercising accounting discretions and real activities manipulations. Our empirical results show that managers determine the level of accounting discretions before that of real activities manipulations. In addition, in order to avoid loss, the likelihood of employing accounting accruals is significantly higher than that of real activities manipulations such as channel stuffing, cutting discretionary expenditures and overproducing inventories. Our empirical results are inconsistent with Zang (2012). Further investigation with quarterly data shows Zang's conclusions are valid only in the fourth quarter. That is, to avoid annual losses, managers are apt to exercise real activities manipulation than accounting accruals in the fourth quarter, compared to previous three quarters. This finding has not been documented in literature.

Archived studies on earnings management in Taiwan generally investigate single real activity or accounting accruals. Unlike prior studies, this study examines multiple real activities and accounting accruals simultaneously for Taiwan's listed firms. We examine the sequence of exercising accounting discretions and real activities manipulations. Our empirical results show that managers determine the level of accounting discretions before that of real activities manipulations. In addition, in order to avoid loss, the likelihood of employing accounting accruals is significantly higher than that of real activities manipulations such as channel stuffing, cutting discretionary expenditures and overproducing inventories. Our empirical results are inconsistent with Zang (2012). Further investigation with quarterly data shows Zang's conclusions are valid only in the fourth quarter. That is, to avoid annual losses, managers are apt to exercise real activities manipulation than accounting accruals in the fourth quarter, compared to previous three quarters. This finding has not been documented in literature.