本研究主要探討分析師的推薦對於過度自信CEO是否具有價值攸關，檢視Malmendier and Tate (2005) 區分CEO為過度自信與否的方法，是否能清楚地將CEO分類為是因為過度樂觀而表現出過度自信，又或是因為其持有一些非公開的資訊，以至於讓他表現出過度自信。故由先前研究中區分CEO的方法為出發點，檢視此方法是否能將CEO分類為過度自信與非過度自信，驗證此分類方法是否是正確的。接著探討分析師的推薦是否是可信的，先釐清分析師推薦以及各公司間累積異常報酬的關係，才能進一步地檢視分析師的推薦是否能成為分類過度自信CEO的指標。實證結果顯示，Malmendier and Tate (2005) 的CEO分類方法未能全然地將CEO分為過度自信與非過度自信，也就是說利用此方法而被歸類為過度自信的CEO中，可能存在著一定比例遭誤判的非過度自信CEO。而在檢視分析師的推薦下，結果顯示分析師的推薦對長窗期累積異常報酬具價值攸關，但分析師推薦對過度自信CEO的長窗期累積異常報酬，未具統計顯著性。;The purpose of this research is to examine whether the recommendation of the analysts can classify that CEO showed overconfident because of their over-optimism or because they hold some non-public information and made them be overconfident. This research started from previous research which classified CEO as overconfident and non-overconfident. First, I examined this classification is correct or not. Second, I use the cumulated abnormal return of the companies to examine the recommendation of these analysts is credible or not. With the previous test, and I can prove whether the recommendation of the analysts can classify overconfident CEO or not. The results showed that the previous CEO classification couldn’t completely divide CEO into overconfidence and non-overconfidence. It meant that there would be a certain percentage of non-overconfident CEOs who were be classified as overconfident. And the results showed that recommendation of analysts is reliable, so I would want to use the analysts’ recommendation to examine they could be used to classify overconfident CEO or non-overconfident CEO.
