台灣共同基金之分項投入效率

Abstract

本文利用隨機邊界法(SFA)來計算出台灣共同基金的分項投入效率表現，而根據Zhou (2002)重新推導的距離函數去估計單一投入效率，以及Battese and Coelli (1995)提出的縱橫面隨機邊界模型同時估算隨機邊界以及無效率方程式。資料期間為2007~2012 共六年，合計310檔共同基金的資料。我們將費用比率、總資產以及淨值作為三個投入變項，風險調整後報酬則作為一個產出變項，並利用2006年作為基期將所有名目變數平減為實質變數。本文的環境變數有五個，分別是基金的風險等級、基金的成立年限、基金經理人的性別、基金經理人是否擁有國外學歷以及基金經理人是否擁有商管碩士學位。較低的風險、較長的成立年限、女性基金經理人、國內學歷以及有管理碩士學位的基金經理人在費用的掌控上較有效率；較高的風險、較短的成立年限、擁有國外學歷以及擁有管理碩士學位的基金經理人在總資產的運用上較有效率；而較低的風險、較短的成立年限、男性基金經理人、擁有國外學歷以及擁有管理碩士學位的基金經理人在淨值的操作上較有效率。

This thesis applies the stochastic frontier approach (SFA) to compute disaggregate input efficiency of mutual funds in Taiwan. Following Zhou et al. (2002), a re-arranged distance function is derived to estimate an individual input efficiency. The panel data SFA model proposed by Battese and Coelli (1995) is applied to simultaneously estimate the stochastic frontier and the equation of inefficiency. A panel dataset contains 310 mutual funds in Taiwan from 2007 to 2012. There are three inputs (expense ratio, total fund asset, and net assets value) and one output (risk-adjusted return) in the SFA model. All nominal variables are deflated into real variables in the base year of 2006. The Risk Rank, age of fund, gender, foreign academic degree, and MBA degree are the six environmental variables. Lower risk, longer operation, female, domestic academic degree, and MBA degree help improve efficiency for expense ratio. Higher risk, shorter operation, foreign academic degree, and MBA degree help improve efficiency for total fund asset. Lower risk, shorter operation, male, foreign academic degree, and MBA degree help improve efficiency for net assets value.