本文的研究目的,在於以分析師預測誤差與股票異常報酬做為資訊不對稱之代理變數,用以衡量股利資訊效果,以探討股利宣告前,分析師預測盈餘是否會受公司內部人傳遞不實資訊而影響其預測的準確度,及股利宣告後是否會造成公司股價有異常變動,另外,並以九十五年與九十六年證券暨期貨市場發展基金會(以下簡稱證基會)(Securities & Futures Institute, SFI)公佈的資訊揭露評鑑結果,深入探究資訊揭露程度的高低及其他的影響因子對股利資訊效果的影響。此外,針對不同的資訊揭露程度對股票累積異常報酬發生機率的影響,則使用次序羅吉斯迴歸模型進行探討。 實證結果發現,股利的宣告的確會造成分析師預測準確度降低的現象,且愈接近股利宣告時間點的預測,產生誤差的比例會愈大,資訊效果愈明顯。在解釋因子方面,當公司盈餘變異性愈大、規模愈大及資訊揭露透明度愈低,分析師盈餘預測的誤差會愈大。 以事件研究法進行檢定股利宣告對公司股價產生異常報酬的資訊效果,所得結果為公司股利的宣告會使公司股價產生異常報酬。在解釋因子方面,宣告期間平均成交量愈大與公司規模愈小者,公司的累積異常報酬會愈大。另一方面,當內部人持股比率愈大,公司宣告期間的累積異常報酬會愈大。最後,透過次序羅吉斯模型檢定的結果,發現當資訊揭露程度愈低、董監事和經理人持股比率愈高,公司股價產生正的累積異常報酬的機率會愈高。
The purposes of this study are to investigate whether there are dividend declaration information effects for Taiwan’s stock listing companies and whether those effects are influenced by the degree of information disclosure transparency proclaimed by SFI in year 2006 and 2007. Used as the proxies of information asymmetry, the analyst earnings forecast errors (AF_ERROR) is utilized to test the cross-sectional variation with respect to managers’ incentive to manage investors’ expectations before dividend announcements. Besides, SCAR is used in the event study to exam whether stock price variation were caused by the dividend declaration information. Moreover, this study use ordered logistic regression model (OLR) to analysis the probabilities of SCAR caused by the different degrees of information disclosure and the other interest factors. The results show that AF_ERROR indeed is influenced by dividend information effect, and the errors will expand when the time approaching the date of dividend declaration. Besides, the higher earnings variation and/or size factors, the higher AF_ERROR. But it is contrarily for the factor of information disclosure transparency to AF_ERROR. As for the result of the event study, this paper finds that the dividend declaration will result in stock abnormal returns. The higher (lower) are average volumes in the event period and/or insider’s shareholding (size), the higher is the SCAR. Using the same factors in the OLR model, this paper finds that the probabilities of positive SCAR are higher associated with lower information disclosure transparency and/or the higher insider’s shareholding.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。