我國首檔指數股票型基金（ETF）－「寶來台灣卓越50基金（TTT）」於短短半年內即籌劃完成並上市交易，其定價效率有必要審慎檢視，故本研究由追蹤誤差、折溢價、交易現況以及避險功能四個角度加以分析，研究期間為2003年6月30日至2005年4月30日止，總計458個交易日。 追蹤誤差方面，實證顯示TTT之淨值報酬率平均每月低於指數報酬率2.68個基本點，而根據Elton等學者的方法排除股利和費用所得之追蹤誤差，每個月平均約有18.48個基本點，主要可歸咎於市值涵蓋比率、類股配重差異、標的指數之波動度、基金投資組合之 值以及成分股替換等原因。為確認股利和費用之影響，本文再根據Ackert 等學者的方式衡量TTT經股利和費用調整前和調整後之mispricing，實證發現兩種mispricing差距不大，亦即影響追蹤誤差最主要的因素並非股利和費用，但追蹤誤差顯著不為零，表示TTT在複製指數表現上確實仍有些微誤差。另外，歸功於特殊的實物申購/贖回機制，TTT之折溢價幅度並不顯著，但折溢價狀況約可持續兩天。探究其原因，進一步檢視淨值與市價間之套利空間，發現在套利機會少且套利利潤有限下，導致法人無誘因進場套利，因而促使折溢價以極小的幅度在市場上持續兩天之久。成交量方面，TTT之交易情形相當活絡，且不受市場波動和套利機會所影響。最後，在避險功能方面，研究發現投資人確實嘗試以TTT作為避險工具，但短期內仍無法改變以期貨作為主要避險工具的習慣。 This thesis examines pricing efficiency of the first Exchange Traded Fund in Taiwan － the Polaris Taiwan Top 50 Tracker Fund (TTT), an ETF designed to track the performance of The Taiwan 50 index. The main focus is on its tracking error, discounts/premiums, trading conditions and functions of hedging. The study period for the TTT goes from June 30th, 2003 to April 30th, 2005, total 458 trading days. Empirical results first show that the TTT’s NAV return (With Dividends) underperforms the Taiwan 50 index return by 2.68 basis points per month. The TTT’s tracking error is about 18.48 basis points per month. The large tracking error is primarily resulted from the sampling of holding portfolio as well as the volatility of underlying index. Second, the TTT’s net asset value is close to its market price due to the ability to create and redeem TTT via in-kind transactions, but the NAV deviations from market price persist about two days. I found that the persistence is caused by the lack of the arbitrage profit. Third, the TTT is trading frequently, and the volume isn’ t significantly affected by market volatility or arbitrage opportunity. Finally, according to the regression test, I found that the investors have tried to choose TTT to hedge, but Taiwan Stock Index Future is still the main choice for hadging.
