CMStatistics 2017: Start Registration
View Submission - CFE
A1483
Title: Examination of hedge fund performance persistence over long-term period using a non-parametric approach Authors:  Ekaterina Ipatova - Roehampton Bisiness School (United Kingdom) [presenting]
Kaizad Doctor - MSI (United Kingdom)
Abstract: Persistence in hedge fund performance is hotly contested by the academic community. Evaluating persistence in fund performance is generally carried out using parametric and non-parametric methods. Procedures that identify superior fund performance after mitigating the possibility of lucky fund managers are used in identifying the set of funds that are generally available in the literature. The methodologies have varying advantages with the False Discovery Rate (FDR) method being less conservative than the Family Wise Error Rate (FWER) however parametric methods are also added to the mix. A comparison of the methodologies has not been carried out in literature. Varying horizons, mitigating survivorship bias and out-of-sample analysis have all been incorporated in my study. A variety of portfolios were used to evaluate the FOF with traditional performance evaluation methods ranging from non-parametric Sharpe ratios and Drawdowns. We find that funds constructed using non-parametric methods outperform those using parametric methods and in general exhibit superior outperformance when compared to conventional portfolios.