This paper seeks to ascertain whether shocks to hedge fund returns are permanent or temporary by using M1 and M2 unit root procedures advanced by Narayan and Popp. In addition, the paper implements the GARCH-based unit root test developed by Liu and Narayan. These procedures allow for two structural breaks in the data. The results from M1 and M2 models indicate that the various hedge fund returns under study are stationary processes with two structural breaks. Similarly, the results from the GARCH-based unit root model confirm those obtained from both the M1 and M2 techniques in that the hedge fund return series were found to be stationary. Taken together, the results suggest that shocks to the various hedge fund returns are temporary. This finding implies that hedge funds can be included in portfolios with traditional assets such as stocks and bonds to reduce risk and enhance returns.

Over the past two decades, the hedge fund industry has grown substantially. Hedge Fund Research (HFR) reported on January 20, 2010 that there were about 530 hedge funds in the industry in 1990 with roughly $50 bn of assets under management. By 2009, that number rose to approximately 8,000 hedge funds with about $1.6 tn of assets under management. This impressive growth of the hedge fund industry has attracted the attention of investors, practitioners and academicians alike. It has been documented by Liang (1999), Agarwal and Naik (2000), Edwards and Caglayan (2000), Cerrato and Iannelli (2006) that hedge funds tend to outperform the conventional market indexes such as the S&P 500 during bearish markets, while the traditional indexes tend to outperform hedge funds during bullish markets. This feature, among others, like low risk and low correlation with traditional asset returns make hedge funds attractive hedging instruments.

Given their importance in financial markets, a number of studies have examined the ability of hedge funds to act as diversification tools. For instance, Gregoriou et al. (2001) examined the random walk behavior of 10 hedge fund classes, including event-driven, global international, global established, global emerging, global macro, market neutral, fund of hedge funds, short sellers, long only, and sector, for the time period January 1991 through December 2000. They applied the standard ADF unit root testing procedure and found that the median returns of all the hedge fund classes with the exception of the market neutral are random walk processes. They therefore concluded that investors should be able to reduce risk by including hedge funds in their portfolios within the market neutral class.

Applied Finance Journal, Are Shocks, Hedge, Fund Returns Permanent, Temporary.