國內共同基金的結構、績效與行為之成因初探

Abstract

影響共同基金在某一段期間報酬率的主要因素包括市場趨勢、投資組合、基金經理團隊的能力、基金費用比率以及運氣。其中市場趨勢、運氣等無法控制因素的影響力，又往往比可控制因素的影響力來的大，在可控制因素中，以基金投資組合影響最鉅。就基金的投資組合而言其報酬率的來源有三個，其一為資產配置，即基金經理對產業、公司作一深入研究後其決定對各類資產的投資比重策略。另外兩個為經理人的擇時策略、選股策略，然此屬於投機策略，基金經理根據短期的市場預測，運用短線交易以獲取超額報酬。基金顧名思義是經理人所管理的一筆資產，其績效優劣應可部分歸因於經理人的操作行為，早期的學者常研究於何種投資策略具有優異的績效表現、基金經理是否具有擇時與選股能力，本文試圖從另一個角度切入，針對國內35支開放與封閉型基金，觀察基金經理的行為是否會促使基金績效差異化。 本文以民國85年至87年為研究期間，就期間資料所得之結論如下： 基金績效差異化成因中，以週轉率與報酬率之負相關較為顯著，至於風險與投資組合則較不明顯。投資組合無法解釋的可能原因為：一、本研究所採資產分類方法所致，本文的重點在於基金投資組合標的，在此為股票，因此股票的分類標準極為重要，如採不攸關之分類標準，則所得到的結果將差之甚遠。二、由於基金的投資組合集中於某些個股上，容易產生各基金齊漲齊跌的現象，造成彼此間的績效差異不顯著。此外本研究亦發現到基金經理的集群行為（Herding），各年度均呈現基金經理將投注於股票市場資金的66％投資於市場上11％的股票中，然此種行為對績效不足以構成差異化成因。

The main factors, which influence the performance of mutual funds in a period, are trend, portfolio, the ability of manager groups, luck, and fee. These factors above can be divided into two parts, controllable and uncontrollable factors. Among controllable factors, the portfolio is most important. The sources of the return for the portfolio are three. One of them is asset allocation, which is the portfolio strategy decided by fund managers, who research industries and firms deeply. The others are timing and selecting, however, they are speculative strategies because fund managers earn a lot of return by immediately trading according to market prediction of short term. Funds, as implied by the name, are assets managed by managers, so the performance is attributed to the operations of managers. The early scholars in general research what kind of investing strategies have great performance, and whether fund managers have abilities for timing and selecting. This paper researches another aspect, taking 35 open and close funds as examples and observing whether the operations of fund managers make the performance different or not. This paper takes 1996 to 1998 as a period of study, and the conclusion by the available information is as follows: Among the reasons which make the fund performance different, the negative-correlation between turn-over-rate and return is most significant, and the correlations between return and risk and between return and portfolio are not significant like the former. Two suspicious reasons of why portfolio cannot explain the performance are as follows. One reason is the method of asset classification used by this paper. The key point of this paper is funds' investment, so the criterion of stocks' classification are very important. If the classification criterion is not relevant, the result will go wrong. The other reason is because many funds' portfolio are invested in some stocks, it's easy to generate a situation that all funds' return is changed in the same way to make the performance not significantly different. Besides, this paper also discovers the herding of fund managers. It shows every year fund managers invest 66% of the funds for the market in 11% of stocks in the market, but this kind of behavior is not reasonable to make the performance different.