ETF 추적오차와 영향요인에 관한 연구:기초자산 복제방식을 중심으로

Exchange Traded Fund(ETF) is an investment product that can be traded in the stock market and follows underlying assets such as KOSPI200. With the volume of the ETF market growing recently, it has become more important to see how well ETF replicates underlying assets. The ETF's method of tracking underlying assets is first, physical replication, which directly buys stocks and track indexes, and secondly, synthetic replication through swap deals with counterparty investors. This study empirically studied the effects of these two replication methods on tracking error through the random effect panel regression model. The main result showed that tracking errors in physical replication method were more significant than those in synthetic replication. The number of constituents belonging to the underlying index, the cash holding ratio, and intraday volatility in the value of net assets had a significant effect on the errors. This research is valuable in that it is the first in Korea to empirically analyze the difference in tracking error of ETF by the replication method. These ETF replication type characteristics suggest that ETF operations may seek to improve price efficiency, which reduces tracking error.