The purpose of this thesis is to compare rank ordering, which are generated by the various performance measures, using Korean equity funds. To do so, we use 996 equity funds, which exist more than 24 months for the sample period of January 2002 to March 2015.
Using hedge fund data, Eling and Schuhmacher(2007) compared the Sharpe Ratio with some of these performance measures, and found virtually identical rank ordering. However, Ornelas et al.(2012) compares the rank ordering generated by 13 performance measures with that of the traditional Sharpe rate and finds that Manipulation-Proof Performance Measure(MPPM), Upside Potential Ratio and Appraisal Ratio generate the different rank ordering with that of the Sharpe ratio, implying that the choice of the performance measure is important. Based on these results, we compare 14 performance measures with the Sharpe Ratio using Spearman’s rank correlation coefficient. Overall, our results show that the performance measures have different rank ordering. This is the case of the Jensen‘s alpha, Rachev ratio, Information ratio and MPPM. This implies that choosing the performance measure for fund evaluation is actually important, as shown in Ornelas et al.(2012).
To examine the robustness of our results, the various robustness tests are carried out in this thesis: (1) to take the possible survivorship bias in our results into account, we separately consider live and defunct funds. (2) because the different fund groups/classes can generate the relatively different rank orderings, we also separate aggressive funds based on average net asset value and fund types, which divided by FnGuide. (3) to consider the effect of increasing market size of Korean equity funds, our sample period are divided by two sub-periods. (4) to take the possible outlier effect in account, we eliminate 1 to 5 of highest (lowest) returns from fund return series and replace 6th highest (lowest) returns. (5) final robustness test to change the risk aversion parameter used in MPPM to examine the possible effect of investors'' risk aversion.
Overall, the result of 5 robustness tests are qualitatively not very different from that of whole sample funds. Especially, from changing the risk aversion parameter we find that as the Relative Risk Aversion(RRA) increase the correlation between MPPM and Sharpe Ratio decreases, meaning that investors more risk-averse would have very different rankings when using one or other measure, which are consistent with Ornelas et al.(2012). Therefore, these results indicate that the choice of the performance measure is actually important for fund ranking and selection.