Managerial Ownership Changes And Mutual Fund Performance

HEC Paris – Finance Department

University of Cologne – Centre for Financial Research (CFR)

August 11, 2015

HEC Paris Research Paper No. FIN-2015-1102

Abstract:

We study the dynamics of fund manager ownership for a sample of U.S. equity mutual funds from 2005 to 2011. We find that ownership changes positively predict changes in future risk-adjusted fund performance. A one-standard-deviation increase in ownership predicts a 1.6 percent increase in alpha in the following year. Fund managers who are required to increase their ownership by fund family policy show the strongest increase in alpha. They do so by increasing their trading activity in line with the view that higher ownership aligns interests of managers with those of shareholders and induces higher effort.

Managerial Ownership Changes And Mutual Fund Performance – Introduction

Since March 2005, mutual funds have to report fund managers’ ownership within the Statement of Additional Information (SAI) using broad ownership ranges. By using one year of ownership data, Khorana, Servaes, and Wedge (2007) and Evans (2008) show that the level of ownership predicts future risk-adjusted performance. Since then, more and more fund investors are paying attention to manager ownership (see, e.g., Ma and Tang (2014)) and some mutual fund families adopted policies which require their managers to hold ownership in the funds they manage.1 This raises the question if the increased attention to managerial ownership and the implementation of ownership requirements by fund families is warranted. Does managerial ownership provide valuable information about future fund performance or is the observed cross-sectional correlation driven by unobserved fund characteristics? Does ownership align incentives and can ownership requirements therefore be used to increase performance? Or do managers have superior information about future fund performance and choose to invest in funds which they know will perform better in the future? Given the nature of their cross-sectional data, these early studies are unable to answer these questions.

We fill this gap in the literature by examining the relationship between ownership changes and changes in future risk-adjusted fund performance using a hand-collected panel data set on mutual fund manager ownership. Examining ownership changes has two advantages in our setting. First, we are able to eliminate any heterogeneity bias stemming from time-invariant unobserved fund characteristics. For instance, funds differ in the degree of managerial discretion in making investment decisions. This managerial discretion may lead to higher fund returns on average, but exposes the fund investor to greater risk of moral hazard (see, e.g., Chen, Goldstein, and Jiang (2008)) and thus increases the optimal level of ownership.2 Therefore, unobservable differences in managerial discretion may bias the ownership coefficient upwards. If managerial discretion is relatively stable within funds over time, we are able to eliminate this bias by using a first-difference approach. Moreover, we control for changes in a host of fund, board and family characteristics as well as predictors of fund performance which have been recently proposed by the literature.

Second, the cross-sectional studies are unable to examine whether the positive relationship between ownership levels and performance reflects fund managers’ superior information about future fund performance or better alignment of fund managers’ and shareholders’ interests. We use changes in ownership mandated by family policy to disentangle the superior information and the incentive alignment hypotheses. The idea is that family mandated changes unlikely reflect a fund manager’s information about future fund performance. If the positive relationship between manager ownership and fund performance reflects fund manager’s superior information about future fund performance, we do not expect that ownership changes which are mandated by the fund family increase fund performance. If on the other hand manager ownership aligns the fund manager’s interests with those of shareholders, we expect ownership changes to have a causal effect on performance even if the change is required by the fund family.