Basu(1997)測試會計盈餘對於好壞消息之反應是否具不同程度的及時性,並定義此為穩健性。本研究根據該定義為基礎,在盈餘之不對稱反應中加入盈餘落後反應的概念。首先檢視好壞消息對於會計盈餘的影響是否具有持續性,再進一步檢視盈餘反應壞消息之多期及時性,以及盈餘對好壞消息反應持續性之差異。結果發現,1994到2007年間,台灣上市櫃公司之盈餘對於好壞消息的反應均具有持續性,對於壞消息的反應則具有多期及時性,但盈餘對壞消息的反應時間比預期長,顯示我國盈餘在及時性方面雖存有穩健性,但在持續性上尚不夠穩健。研究結果說明了利用盈餘之多期不對稱反應模型較單期模型能描述台灣上市櫃公司之盈餘穩健性。
Basu (1997) uses a reverse regression model to examine the conservatism concept. With the positive incremental responsiveness to contemporaneous bad news, he interprets conservatism as accountants' tendency to require more prudent verification of good news than bad news. We extend his model to a multi-period one and examine how conservatism manifests under the multi-period concept. Our empirical results show that during 1994 to 2007, earnings of all TSE-listed and OTC-listed firms in Taiwan response stock return with multi-period lags. Under the concept of the multi-period conservatism, earnings respond both to good and bad news with persistency and respond to bad news with timeliness. However, the impact of bad news persists longer than our expectation. This indicates that earnings of Taiwanese firms are not thoroughly conservative under the multi-period concept.