Investments, tenth edition

 Figure 11.7 

Mutual fund alphas computed using a four-factor model of expected return, 1993–2007. 

(The best and worst 2.5% of observations are excluded from this distribution.)  

 Source: Professor Richard Evans, University of Virginia, Darden School of Business. 

bod61671_ch11_349-387.indd   377

bod61671_ch11_349-387.indd   377

7/17/13   3:41 PM

7/17/13   3:41 PM

Final PDF to printer

---

378

P A R T   I I I

  Equilibrium in Capital Markets

before  fees, after the fees charged to their customers, alphas were negative. Likewise, 

Wermers,  

56

   who uses both style portfolios as well as the characteristics of the stocks held 

by mutual funds to control for performance, also finds positive gross alphas but negative 

net alphas after controlling for fees and risk.

 

 Carhart  

57

   reexamines the issue of consistency in mutual fund performance and finds 

that, after controlling for these factors, there is only minor persistence in relative perfor-

mance across managers. Moreover, much of that persistence seems due to expenses and 

transactions costs rather than gross investment returns.

 

 However, Bollen and Busse  

58

   do find evidence of performance persistence, at least over 

short horizons. They rank mutual fund performance using the four-factor model over a 

base quarter, assign funds into one of ten deciles according to base-period alpha, and then 

look at performance in the following quarter.  Figure 11.8  illustrates their results. The solid 

line is the average alpha of funds within each of the deciles in the base period (expressed 

on a quarterly basis). The steepness of that curve reflects the considerable dispersion in 

performance in the ranking period. The dashed line is the average performance of the funds 

in each decile in the following quarter. The shallowness of this curve indicates that most 

of the original performance differential disappears. Nevertheless, the plot is still clearly 

downward sloping, so it appears that, at least over a short horizon such as one quarter, 

there is some performance consistency. However, that persistence is probably too small a 

fraction of the original performance differential to justify performance chasing by mutual 

fund customers.

Quarterly Return (%)

−2

−4

−6

0

6

4

2

Performance Decile in Ranking Quarter

Ranking Quarter

Post-Ranking Quarter

1

2

3

4

5

6

7

8

9

10