0 wse">ity is not a good thing for investment since this is after all a
kind of activity that one can upgrade through long-term exer-
cise. Thus, the interaction between fund manager age, securities
working time and fund performance will mitigate the inflection
of fund performance to manager turnover, likewise, when the
working time on one fund is longer, the fund manager has built
long-term cooperating relationship with the fund company, and
the fund company is more likely to indulge the fund manager
whose securities working time is longer especially facing the
current condition of serious lack of fund manager talents, thus
the interaction between the working time on one fund and fund
performance decreases the inflection of fund performance to
manager turnover.
2) Assumption 2: Fund manager age, securities working time
and the working on one fund will mitigate the negative correla-
tion between fund performance and fund manager turnover.
Data Statistics and Analysis
In this study, we estimate the probability of management turn-
over using a logistic regression framework using SPSS17.0.
The regression equation is:
3
0
1
ln αβ β
1i
i
p
X
XZ
p
 
P is the probability of fund manager turnover, x is fund per-
formance indicator, Zi represents the character of fund manager
which consists of fund manager age, securities working time
and the working time on one fund.
In order to avoid the impact of crucial independent variable
(performance) and high correlation between interactive vari-
ables on the result of regression analysis, this paper makes cen-
tralization process for interactive variables, and then we have

X
ZXXZZ

1) The analysis of the effect of fund returns on fund manager
turnover
Firstly, based on fund returns and fund excess returns, this
paper makes an empirical research on the effects of fund per-
formance on fund manager turnover by SPSS17.0 software. The
results are shown in Tables 1 and 2.
Table 1.
Fund manager turnover Logistic Regression Analysis based on fund
returns rate.
Variable Regression
Coefficient Regression C oefficient
Standard Error Wald Sig.
Constant 0.273 0.605 0.2040.652
Returns rate –4.768 2.783 2.9340.087
Returns rate * Age3.002 1.241 5.852.016
Returns rate *
Securities working
time –0.187 0.829 0.0510.822
Returns rate * The
working time on
one fund –0.071 0.202 0.1230.726
Model degree of
fitting Chi-square = 11.259 Sig = 0.024
Table 2.
Fund manager turnover Logistic Regression Analysis based on fund
excess returns rate.
Variable Regression
Coefficient Regression Coefficient
Standard Error WaldSig.
Constant –0.945 0.194 23.818 0.000
Excess returns rate–4.759 2.783 2.924 0.087
Excess returns rate *
Age 3.005 1.242 5.852 0.016
Excess returns rate *
Securities working
time –0.177 0.829 0.046 0.831
Excess returns rate *
The working time on
one fund –0.067 0.202 0.111 0.739
Model degree of
fitting Chi-square = 11.251 Sig = 0.044
Copyright © 2012 SciRes. 53
T. Y. HUI, Z. YUAN
The Tables 1 and 2 show both fund returns rate and fund
excess returns rate have negative impact on fund manager turn-
over, which demonstrates assumption 1. And this means there
is a determinate and reasonable fund manager adjusting mecha-
nism in our country, which can make adjustment of the fund
manager who has poor performance according to fund perfor-
mance to protect the benefit of investors. In terms of the inter-
action, only the coefficient of the interaction between fund
manager age and fund performance is lower than 0.05, which
means the impact is significantly positive. This shows that
when the older the fund manager is, due to the more plentiful
social experiences and the greater social influence, the prob-
ability of the incapable fund manager to be replaced decreases,
which partly verifies assumption 2. And the interaction between
fund returns rate, fund excess returns rate, and fund manager
securities working time, the working time on one fund has in-
significantly negative impact on fund manager turnover. It is
against with preceding assumption, the reason is likely that lack
of fund manager talents currently, which makes it easier for the
fund manager whose securities working time and fund working
time are longer to make job-hopping. Then when the fund
which is charged of this kind of fund manager performs poorly,
they will resign on their own in case of the demotion or the
negative influence of public opinion, which results in the nega-
tive interactive impact effect. This means the longer securities
working time or the working time on one fund is, the more
likely fund manager turnover is when the fund performance is
poor.
2) The analysis of the effect of fund risk-adjusted returns on
fund manager turnover
Then, based on fund risk-adjusted returns, this paper makes
an empirical research on the effects of fund performance on
fund manager turnover. Through Logistic Regression Analysis
on the impact of fund risk-adjusted returns and its interaction
with the character of fund manager on fund manager turnover,
it shows Chi-square = 5.727, Sig = 0.221 in Sharpe index
analysis model; Chi-square = 7.481, Sig = 0.113 in Jensen in-
dex analysis model; Chi-square = 7.481, Sig = 0.113 in Treynor
index analysis model, which means that fund risk-adjusted
returns has no explanatory power on fund manager turnover.
Due to the current condition that fund ranking generally bases
on fund returns rate, additionally the comparison of fund re-
turns rate is intuitionist and simply operated, the fund compa-
nies are largely dependent on the adjustment of fund manager
according to fund returns performance.
3) The analysis of the effect of fund manager market-timing
and stock-selection capability on fund manager turnover
This paper adopts CL model to assess fund manager’s capa-
bility of market-timing and stock-selection, and Logistic Re-
gression Analysis on the impact of it on fund manager turnover
is shown in Tables 3 and 4
Tables 3 and 4 show that fund manager’s capability of stock-
selection and market-timing has insignificantly negative impact
on fund manager turnover. And the interaction between the ca-
pability of stock-selection and fund manager age has signifi-
cantly positive impact on fund manager turnover. While the
fund manager’s capability of market-timing and securities work-
ing time have significantly positive impact on fund manager
turnover under 0.1 level, which partly verifies assumption 2.
This means the longer the securities working time is, the greater
the social influence and recognition. Since individuals consider
that fund industry requires plentiful practical experiences, and
Table 3.
Fund manager turnover Logistic Regression Analysis based on the cap-
ability of stock-selection.
Variable Regression
coefficient Regression coefficient
standard error WaldSig.
Constant –0.155 0.741 0.044 0.834
The capability of
stock-selection –0.459 0.553 0.689 0.406
The capability of
stock-selection *
Age 0.845 0.305 7.678 0.006
The capability of
stock-selection *
Securities work-
ing time
–0.086 0.145 0.351 0.554
The capability of
stock-selection *
The working time
on one fund
–0.018 0.040 0.196 0.658
Model degree of
fitting Chi-square = 12.455 Sig = 0.014
Table 4.
Fund manager turnover Logistic Regression Analysis based on the
capability of market-timing.
Variable Regression
coefficient Regression coefficient
standard error WaldSig.
Constant –0.864 0.260 11.028 0.001
The capability of
market-timing –1.025 1.754 0.342 0.559
The capability of
market-timing *
Age 0.772 0.605 1.629 0.202
The capability of
market-timing *
Securities working
time
0.767 0.468 2.687 0.100
The capability of
market-timing *
The working time
on one fund
0.066 0.125 0.282 0.595
Model degree of
fitting Chi-square = 7.278 Sig = 0 .122
fund manager can acquire high investing capability only if
through long-term practice, fund manager’s securities working
time becomes a crucial standard for selecting fund manager.
Therefore, for the fund manager whose securities working time
is longer but fund performance is poorer, the fund companies
will consider help them reverse the performance if only possi-
ble instead of making them be replaced. Namely, securities
working time can mitigate the impact of fund performance on
turnover.
Conclusions and Suggestions
The empirical research shows that fund performance has
negative impact on fund manager turnover, however, this im-
pact is merely significant for the index of fund returns rate and
excess returns rate to fund manager turnover with a significance
level of 0.1. This means that under the pressure of fund returns
rate ranking, the fund companies generally make adjustment of
fund manager according to the performance on returns rate so
as to form a determinate incentive and restraint mechanism;
Risk-adjusted fund performance and the capability of stock-
selection and market-timing both have no explanatory power on
fund manager turnover, which illustrates that the fund compa-
Copyright © 2012 SciRes.
54
T. Y. HUI, Z. YUAN
Copyright © 2012 SciRes. 55
nies in our country pay immoderate attention to fund manager’s
earning capability but overlook their capability of coping with
risk and deep-seated trading. This kind of elimination mecha-
nism makes many fund managers adopt high risk and high re-
turns investing strategy, which significantly decreases the prac-
tical level of returns. Therefore, we should introduce foreign
advanced fund performance evaluation system into our country
in order to make scientific, reasonable and objective evaluation
for fund practical performance. In addition, on this basis, we
also need to make adjustment of fund manager to form a sur-
vival of the fittest incentive and restraint mechanism so as to
ignite the work passion of fund manager. Finally, for adjusting
variables, the interaction between fund manager age, securities
working time and fund performance has positive impact on
fund manager turnover, which means that when the operation is
poor, the turnover of older fund managers is lower than that of
younger ones and the turnover rate of the fund managers whose
securities working time are longer is lower than that of the ones
whose working time is shorter. The fund companies should try
their best to avoid the evaluation on the character of fund man-
ager, well embody fairness, and ensure the benefits of investors,
so as to inspire the fund managers’ working enthusiasm and ac-
hieve the coordination of the fund companies, fund manager
and investors.
REFERENCES
Bao, R. (2008). Empirical analysis of the replacement of fund manager
and the character of fund behavior in China. Shanghai: Shanghai Jiao-
tong University Press.
Hu, F. A, Hall, R., & Harvey, C. R. (2000). Pro mot io n or d em oti on ? An
empircal investigation of the determinants of top mutual manager
change. Working paper, Durham: Duke university.
Khorana, A. (1996). Top management turnover: An empirical investi-
gation of mutual fund managers. Journal of Financial Economics, 40,
pp. 403-427. doi:10.1016/0304-405X(95)00852-6
Lu, J. C., & Wang, M. B. (2007). What decide the replacement of fund
manager. Securities Market Guide, 3 , 68-77.
Morck, R., Shleifer, A., & Vishny, R. (1989). Management ownership
and market valuation: An empirical analysis. Journal of Financial
Economics, 20, 293-315. doi:10.1016/0304-405X(88)90048-7
Warmers, R., Wu, Y. C., & Zechner, J. (2005). Portfolio performance,
discount dynamics and the turnover of closed-end fund managers.
EFA 2006 Zurich Meetings Paper; AFA 2008 New Orleans Meetings
Paper.
Zhang, L. Q., & Deng, F. (2011). Research on the effect of performance
on the replacement of fund manager. Knowledge Economics, 8, 135-
136.