Implications of manager replacement: evidence from the Spanish mutual fund industry

Abstract

We analyse the financial consequences of manager replacement in a sample that includes all domestic and European equity funds in Spain. Specifically, we examine a total of 104 funds of the sample that experience manager turnover over the period 1999–2009. We find that underperforming funds in the pre-replacement period experience a significant improvement in the excess returns and performance after the manager change, an improvement that lasts over time for domestic equity funds. The analysis of the risk profile indicates that funds experiencing a manager change do not show significantly different levels of risk before the replacement dates although they tend to show an increase in the level of total risk after the change. Finally, the pool-regression analysis of the investment flows confirms that manager changes tend to impact negatively on subsequent flows of those funds with manager turnover.

Keywords:

• investor flows
• manager replacement
• money flows
• mutual funds
• performance
• risk

JEL Classification:

• G11
• G14

Acknowledgements

The authors are grateful to participants of XXI Finance Forum held in November 2013, to participants of 2013 ASEPELT Annual Conference held in July 2013 as well as to participants of the seminar of the Master in Accounting and Finance held in October 2012 in Zaragoza. Any possible errors contained in this paper are the exclusive responsibility of the authors.

Notes

¹ Agrawal et al. (1999) indicate that studies about executive and manager replacement have emerged due to scandals of fraud in some companies.

² The assets under management in Spanish mutual funds have grown more than 25% in the last two decades.

³ In December 2009, management companies altogether managed 2593 mutual funds with a market value of €171 billion.

⁴ See, e. g. Cambon and Losada (2014) for a detailed description of the Spanish mutual fund industry.

⁵ Objective-adjusted performance measures are computed as the difference between a fund’s performance and the average performance of all funds in that investment objective. It is interesting to make this distinction, since a fund may improve after the change, but the industry average do better than the fund, and therefore the concrete mutual fund has not improved as much as previously thought.

⁶ This is the case in the Spanish mutual fund market, where around 98% of investors are retail investors and the remaining investors manage approximately a 25% of the industry in terms of money (see, e.g., Vicente et al., 2011).

⁷ We compare the SEs obtained using different panel data techniques to identify the potential correlations between the residuals across funds and/or time. Specifically, we obtain White consistent SEs and OLS SEs clustered by fund, by time and by both fund and time. The comparison of these SEs leads us to reject the existence of a fund effect or time effect in our data – that is, a time-series dependence of the residuals for a given fund and a cross-sectional dependence of the residuals for a given year across the different funds – since the SEs clustered by time or fund are not larger than twice or three times the White consistent SEs (see, e.g., Petersen, 2009 for more details about the identification of fund-fixed effects or time-fixed effects).