績優基金績效、持續性、決定因素與排名—台灣股票型基金之實證研究

Performing Persistence, Determinants and Ranking of Superior Performing Mutual Funds–An Empirical Study of Mutual Funds in Taiwan

陳姵穎(Pei-Ying Chen)

指導教授 ： 劉定焜

朝陽科技大學/管理學院/財務金融系/碩士(2011年)

摘要

日益擴增且多樣化的共同基金，操作難度頗高。雖然基金本身具分散風險的功能，卻仍無法避免面臨虧損。投資人該如何由眾多基金挑選適合且績效卓越的商品，實為一大挑戰。因此，投資人欲投資基金時，基金經理人的操作績效、績效是否具有持續性、影響基金表現的重要因素，以及基金排名的適切性，為重要的投資參考指標，亦為值得予以深入探討的議題。 以往研究大多著重於單一類別變數對基金績效的影響，加上基金基本資料取得不易，往往忽略基金經理人的特別屬性，殊不知，此屬性對基金績效亦具有重大的影響。因此，本文根據游佳紅 (1995)、Prather et al. (2004)、Farnsworth and Taylor (2006)、Gottesman and Morey (2006) 及Karagiannidis (2010) 等人文獻，將影響基金績效之變數區分為四類，包括基金經理人變數、基金變數、公司變數以及環境變數，並以一、三與六個月期和一、三、五與十年期，各期績效排名前十名的國內股票型基金為樣本，建構2001年1月至2009年12月的balanced panel data。 首先，探討各項變數與基金績效指標之間的關係，期能找出適合解釋基金績效的指標。其次，以傳統績效指標 (基金淨值報酬率、Sharpe Index、Treynor Index與Jensen Index) 評估基金績效，藉由F檢定、LM檢定和Hausman檢定，選取最適的實證模型，並依據Prather et al. (2004) 納入前期績效變數，以檢驗國內績優共同基金是否具有績效持續性。再者，以資料包絡分析法 (data envelopment analysis, DEA) 對基金績效具顯著影響之變數進行評估，取得各期基金的總技術效率值(overall technology efficiency, OTE)，平均各期基金的總技術效率值，對各期基金重新排名，以彌補傳統績效指標之不足。最後，相較於過去文獻僅採較短期間進行分析，本研究運用較長的研究期間，洞悉績效期間結構之變化。 實證發現無論短期或長期，皆存在績效持續性之效果。基金經理人變數對基金績效指標之影響性，短期較長期顯著。DEA排名與基金原始排名亦存在較大的差異性。藉由績效期間結構得知，基金淨值報酬率可視為基金績效之領先指標。以期提供相關資訊於政府機關、金融機構及投資大眾決策擬定，以及後續研究者具價值之參考。

關鍵字

績優基金績效 ； 績效持續性 ； 基金排名 ； 資料包絡分析法

並列摘要

The diverse of mutual funds and the expansion of the market contribute to the difficulty of their operation. Although fund has the capability of risk dispersion, it still can’t avoid facing the loss. How to select suitable and excellent commodities from numerous funds is a huge challenge for investors. Therefore, when investors want to invest funds, the operating performance of fund manager, performance persistence, major influence factors of fund performance as well as appropriateness of fund ranking are important referencing indicators that are worthy of discussion.
In most former studies, they only emphasize the single category variable of the influence on fund performance and often neglect the special attribute of fund manager. We didn’t realize that this attribute also has significant influence on fund performance. According to Yu (1995), Prather et al. (2004), Farnsworth and Taylor (2006), Gottesman and Morey (2006) and Karagiannidis (2010), we divide affective variables into four categories, including fund manager variable, fund variable, company variable as well as environment variable. We also sample top ten funds of domestic equity fund from each term of one, three, and six-months, as well as one, three, five and ten-years to construct panel data from January 2001 to December 2009.
First, this study explores the relationship between each variable and fund performance indicator, in the hope to find out appropriate indice of fund performance. Second, it evaluates fund performance with traditional performance indice (Mutual funds return on net value, Sharpe Index, Treynor Index and Jensen Index). Beside, the most suitable empirical model is chosen by F test, LM test and Hausman test. It also subsumes pre-performance variable, according to Prather et al. (2004). Overall the purpose of the study is to check whether domestic superior mutual funds have performance persistence. Furthermore, we apply data envelopment analysis (DEA) to evaluate fund performance with significant variables, in order to obtain fund’s overall technology efficiency (OTE) in each term. We can make up deficiencies of traditional performance indice after reranking funds in each term using the average of OTE. Finally, this study makes use of long period to observe the variation of performance term structure contrasting with previous literatures analyzing only short period.
The empirical results of this study found that short-term and long-term possess performing persistence effects. Fund manager variable for influence of fund performance indictor is more significant in short-term than it is in long-term. The difference in ranking between DEA and raw is huge. From performance term structure, we find that Mutual funds return on net value can be regarded as leading indicator of fund performance. We expect to offer valuable references and relevant information for government, financial institutions, investors and researchers of future study.

並列關鍵字

superior performing mutual funds ； performing persistence ； ranking of mutual funds ； data envelopment analysis

參考文獻

邱永和、陳玉涓、陳素緞與陳剴夫 (2008)。國內共同基金之績效評估，會計學報，1 (1)，29-52。

邱顯比與林清珮 (1999)。共同基金分類與基金績效持續性之研究。中國財務學刊，7 (2)，63-88。

高強、黃旭男與Toshiyuki Sueyoshi (2003)。管理績效評估：資料包絡分析法 (初版)。台北市：華泰文化事業股份有限公司。

陳信憲與王南喻 (2005)。美國開放式股票型基金之長短期績效持續性研究，台灣金融財務季刊，6 (4)，21-39。

莊惠菁與管中閔 (2010)。以無資料窺探偏誤的檢定評估共同基金績效。證券市場發展季刊，22 (3)，181-206。