Asian Review of Financial Research Vol.32 No.1 pp.1-31
The Effect of Fund Manager Replacement on Fund Performance, Risk, and Money Flows
Key Words : Fund,Fund Manager Replacement,Performance,Risk,New Money Flow
We examine the effects of a fund manager replacement on fund performance, risk, and money flows using data from January 2004 to March 2015. We define a fund as either a positive performance (PP) fund or a negative performance (NP) fund depending on whether it belongs to the upper 50% or lower 50% based on its performance in the previous year. We then compare the effects of manager replacement between the two groups. First, we find that the performance of an NP fund improves after the fund manager replacement, while that of a PP fund decreases. This phenomenon becomes more pronounced for the +2 year period compared with the +1 year period. This result implies that while it is feasible to find a qualified manager to replace a manager whose performance is below average, it is difficult to find a better manager to replace one whose performance is above average. Therefore, manager replacement can be a positive event for investors in an NP fund and a negative event for investors in a PP fund. However, investors should understand that it takes some time for fund performance to change, as the fund performance of a PP fund remains better than that of an NP fund for +1 year after the manager replacement. The improved performance of the NP fund and decreased performance of the PP fund become significant from the +2 year period after the fund manager replacement. Second, the risk level of the NP fund decreases after the manager replacement, whereas that of the PP fund remains virtually the same. We find that both the standard deviation of excess return and the one factor beta of the NP fund are higher than those of the PP fund for the -1 year period. We conjecture that this is because poorly performing fund managers take more risks to recover their inferior performance. This result is consistent with the U.S. market findings of Brown et al. (1996). The risk level of the NP fund decreases significantly after the fund manager replacement, whereas no significant change occurs for the PP fund. Third, we find a significant change in the style of the fund that survives more than two years after the manager replacement, with the fund manager replacement slowly contributing to the fund style change. Comparing the styles before and after the manager replacement, we find that the PP fund has a higher coefficient for the SMB and PR1YR factors and a lower coefficient for the HML factor. This implies that PP fund managers prefer small and medium-sized stocks, growth, and momentum. The meaningful style change is not found for the +1 year period for either the PP fund or NP fund, as it may require some time to change the fund portfolios and style. However, for the +2 year period, we find a significant style change for the NP fund; its negative SMB coefficient increases and its negative HML coefficient decreases, which means the proportion of large and value stocks increases in its portfolio. Fourth, after controlling for the various factors that may affect the fund money flows, we find no significant change in the new money flows before and after the fund manager replacement. We find that the money flows for both the NP and PP funds are negative for the +1 year period, although statistically significant so only for the NP fund. We use a number of control variables to check whether the inferior fund performance or the fund manager replacement causes the negative money flow for the NP fund. We use past performance, the excess return standard deviation, turnover, fund size, and management fee as control variables and run the regression on the new money flow for the fund manager replacement. Our regression results show that the fund manager replacement has no significant effect on the fund money flows. This indicates that important information such as the fund manager replacement is not delivered efficiently to fund investors in the Korean market. Therefore, we recommend that fund management companies and regulatory agencies establish practical methods to promptly deliver crucial information such as fund manager replacements to fund investors to protect the investors' rights. This study is the first to examine the effects of fund manager replacement on fund performance and risk in the Korean market. However, fund manager research in Korea carries some limitations. We collect the fund manager replacement data from fund rating companies since 2004; therefore, the study period is rather short and can be period dependent. Furthermore, many fund management companies do not specify the name or responsible fund manager, but instead use the chief investment officer's name or multiple manager names to identify the head manager of a fund. We hope future research on this topic will resolve the limits of our study and deliver further findings on the effect of fund manager replacement on fund performance.