韓國ETF市場研究：合成型ETF之追蹤誤差

Abstract

韓國和臺灣的ETF市場同樣都始於2000年初，但後續發展卻有明顯的差異，在商品數的部分，截至2015年底韓國將近兩百檔ETF商品掛牌上市，而臺灣僅有46檔，在數量差異上相當顯著。因此本研究希望能透過研究韓國ETF市場，除了基本組成特性之差異外，進一步分析韓國投資人較喜好的合成型ETF，檢驗其績效是否真的較佳，並將結果供臺灣ETF市場未來發展之參考。 實證結果發現，研究中所選用的韓國ETF商品，不管是一般型、槓桿型還是反向型的ETF商品，在1%顯著水準下，其追蹤誤差均顯著異於零。而進一步針對可能影響追蹤誤差的變數進行複迴歸分析發現：追蹤誤差和標竿指數波動度、標竿指數成分股數、交易量呈現正向關係；管理資產規模、發行持續時間和商品類型呈現反向關係，雖然ETF商品屬於被動式投資商品，但是投資人在決策時可以考量上述因素來衡量追蹤誤差，有效管理自身的投資組合。

The ETF market in Korea and Taiwan both initiated in early 2000, while there has been obviously different development between the two. As to the amount of products, there were almost two hundred listed ETF products in Korea up to 2015, yet there were merely forty-six in Taiwan, quite a significant difference. Therefore, this paper examines the Korea ETF market to analyze the fundamental differences of characteristics as well as the performance of synthetic ETF, which Korean investors prefer, serving as a reference for future development of ETF market in Taiwan. The empirical evidence indicates that, under 1% significant level, the tracking errors of general, leveraged, and inverse ETF in Korea are significantly different from zero. The potential variables that may affect tracking errors were further investigated using multiple regression analysis. The result suggests that there is positive correlation between tracking error and benchmark index volatility. For the number stocks of benchmark index and trading volume, tracking error has positive correlations with the two. However, AUM, duration, and types of products all have negative correlation with tracking error. Although ETF products are passive investment instruments, investors may consider the factors mentioned to measure tracking errors when making decision and manage their portfolios effectively.