Content area
Full Text
1.
Introduction
Over the past decade, investments in exchange-traded funds (ETFs) have strongly increased and reached almost US-$2.6 trillion in July 2014 (see iShares, 2014). ETFs facilitate the investment in a market and/or index, i.e. they enable investors to trade broadly diversified portfolios of securities in single transactions (see e.g. Picard and Braun, 2010). (For a review of the empirical literature on ETFs, we refer the interested reader to Boldin and Cici, 2010.) Since ETFs represent an underlying basket of assets (usually stocks), it is generally assumed that ETF prices are highly efficient and (almost) perfectly match the price of the underlying portfolio (see e.g. Engle and Sarkar, 2006).
In this paper, we focus on the relation of ETF prices and the market value of the underlying portfolio of stocks, i.e. the ETF's net-asset-value (NAV). We are able to document significant (temporary) deviations between these two valuations. These represent an inefficiency that implies the existence of investment strategies leading to systematic, risk-adjusted, and transaction-cost adjusted excess returns (Jensen, 1978, p. 96). Following these considerations together with results by Fulkerson and Jordan (2013), we aim at finding such an investment strategy in ETF markets with a particular focus on accounting for transaction costs.
In our analyses, we suggest a daily long-short trading strategy using daily price/NAV ratios as signals. Specifically, we go long the one ETF with the lowest discount (most undervalued ETF compared to the underlying portfolio) and go short the one ETF with the highest premium (most overvalued ETF compared to the underlying portfolio). For this trading strategy, we find positive, non-trivial gross excess returns. When accounting for trading costs, the results show significant positive net excess returns, but only during the 2008-2010 period. Consequently, our paper provides a comparison of gross and net returns and thus documents the impact of trading costs even in highly efficient markets like ETF markets.
Regarding the literature on price/NAV ratios in ETFs, Engle and Sarkar (2006) provide a comparison between domestic (USA) ETFs and international ETFs. In both intraday and end-of-day analysis, they find comparatively smaller and less persistent price/NAV ratios in the USA. ETFs potentially indicating a higher level of efficiency of the US ETF market (at that time). Aber-Li and Can (2009) examine their price...