台灣股票型共同基金投資集中度與績效持續性之關聯探討

Abstract

本文以台灣股票型共同基金的月資料進行實證分析，使用基金前五大持股的Herfindahl Index作為衡量投資集中度的指標，探討投資集中度對未來績效及其持續性的影響。首先，我們依照投資集中度和過去報酬率兩個維度進行分組並以T檢定測試持有期間分組報酬差異。其次，透過Fama- MacBeth (1973)橫斷面迴歸分析方法，檢視在控制其他變數的情況下，投資集中度和基金績效持續性是否仍存在線性關係。最後我們在兩種研究方法下，都觀察到投資集中度越高的基金績效持續性越差的情形，且投資集中度和前期績效為正相關，和後期績效為負相關。 為了探討集中度和前後期績效相關性不一致的可能因素，本文依基金過去績效區分為贏家和輸家子樣本進行迴歸分析，發現投資相對分散且過去平均表現較好的基金在後期績效持續性較佳。整體而言，我們認為投資集中度和前期績效的正向關係可以歸因於經理人因為運氣好而集中投資到表現優異的個股，但在運氣不容易維持的情況下，集中投資策略難以持續發揮優勢。然而，若經理人過去績效好且採取相對分散的投資策略，則較能維持基金的績效持續性。

In this study, we use Herfindahl index of the top-five fund holdings as an indicator to measure portfolio concentration, and investigate the impact of investment concentration on performance persistence with monthly data of open-ended equity mutual funds in Taiwan. We conduct the empirical analysis through two methods: first, we divide our sample into groups based on portfolio concentration and past performance, and verify the significance of difference in group returns with t-test. In the second part, we use Fama-MacBeth (1973) cross-sectional regression model to check the linear relationship between portfolio concentration and fund performance persistence. The result shows that portfolio concentration is negatively correlated with performance persistence, while portfolio concentration is positively correlated with past performance and negatively correlated with future performance. We conclude that concentrated managers outperform than other funds because they take big bets precisely on the outstanding stocks. However, these overconfidence managers will not always win since good luck will not consist. The evidence suggests that managers with selection skills may enhance their future performance while taking more diversified strategy.