Style Analysis of Funds of Hedge Funds: Measurement of Asset Allocation and Style Drift

The main purpose of the study is to reveal how accurate Sharpe’s style analysis model is in the context of the analysis of funds of hedge funds. Therefore, we analyse the influence of the three parameters of a style analysis model, which are defined by the user: the number of indices which are in the set of investment indices, the length of the rolling time window and the nature of the return series (smoothed or unsmoothed). As we find that the unexplained Sharpe index volatilities for the hedge fund group indices are much higher and more stable as the ones for the hedge fund strategy indices, we can concluded that the former are more suitable for such an accurate style analysis model. Regarding the choice of the rolling time window, we find that the length of the rolling time window has practically no influence on the accuracy of the model. The influence of the unsmoothed return series depends on the used set of investment strategy indices. Regressions with unsmoothed return series of hedge fund strategy indices exhibit a higher J-statistic, whereas regression with unsmoothed return series of hedge fund group indices have lower J-statistics as the regression with smoothed return series.As we find that in general Sharpe’s method can distinguish only four hedge fund groups reasonably well from each other, its usage for the risk management tools is limited. Given the fact that each hedge fund strategy exhibits different forms or risk , like credit risk or market risk, a fine-grained decomposition is needed to get an accurate evaluation of the fund’s of hedge funds risks. For the purpose of performance evaluation and classification of fund’s of hedge funds a splitting into four broader hedge fund groups can be satisfactory in some cases.