Portfolio Selection and Hedge Funds: The Effects of Autocorrelation and Tail Risk
MetadataShow full item record
This paper examines the sensitivities in portfolio selection due to the effects of serial correlation in asset returns. Traditional mean-variance analysis (MVA) and mean-Conditional Value at Risk (M-CVaR) portfolio frameworks are employed on an investment universe of global stocks, world bonds and global hedge funds. The findings reveal that autocorrelation in asset returns tends to induce a downward bias in the second sample moment making hedge funds artificially desirable. This effect is found in both MVA and M-CVaR frameworks. The MVA results show that hedge funds are an attractive asset class for mean-variance investors as they lower portfolio volatility at the cost of undesirable third and fourth portfolio moments. Conversely, M-CVaR investors who prefer to minimise the left tail of portfolio returns tend to allocate little to hedge fund investments. Furthermore, M-CVaR investors with a heightened aversion to tail-risk will hold a zero portfolio weighting to hedge funds. These findings are consistent with the notion that the inherent risk in hedge funds is located in the tail of their distribution of returns.
The 7th Global Conference on Business and Economics