現代的投資理論主張將資產配置於不同的產業來降低風險,以獲得合理報酬。不過近年來,有些學者卻認為持股集中於某些產業,反而會獲得較好的績效。不過在這議題上卻是很少被探討。本研究利用資料包絡法與分位數迴歸分析,來衡量在不同效率水準上,產業集中程度對基金績效的影響與變化。結果發現在效率水準較高的基金,基金持股之產業集中程度與基金績效呈現顯著性正向關係,但是在較差效率水準的基金,則無顯著差異。此意謂著基金經理人在資訊取得上擁有相對優勢,因而投資於前景佳、獲利性高、財務透明的公司股票來獲取高於市場平均報酬的利潤。本研究也從基金投資於產品類別的股票、投資於所屬集團之集團股的集中程度來探討。亦發現在高效率水準時產品集中度與基金績效有正向關係,但是在較低率水準時集團集中度與基金績效呈現負向關係。
The investment textbook suggests that diversifications of stockholdings across industries can reduce portfolios’ idiosyncratic risk. But some researcher found that there exist positive relationships between industry concentration and mutual funds performance. They argued that fund managers with superior investment ability or information advantages may hold portfolios concentrated in a few industries. Our study argues that for those mutual funds with concentrated portfolios, their performance should be on the two extremes: the best and the worst. Therefore, this study applies the quantile regression techniques to evaluate the relation between the industry concentration and the mutual fund performance while the mutual funds are grouped into different performance efficiency quantile level. The result shows that the industrial concentration is positive correlated significantly with efficient score at higher quantle levels, but not significant at lower quantile levels.