Advances in Investment Analysis and Portfolio Management, Volume 9

strategy that uses the characteristics model. The strategy that uses the FF model

has the highest Sharpe performance across the four strategies in panel A of

Table 5. This is more than three times larger than the corresponding Sharpe

performance of the FTA market index. This stands in sharp contrast to the

Table 5. Performance of Asset Allocation Strategies: Conditional Models.

Panel A

Mean Std Deviation Sharpe Minimum Maximum

CAPM 1.132 3.27 0.346 –6.89 24.56

FF 2.938 5.94 0.495 –9.49 28.16

EGB 2.791 6.39 0.437 –13.56 28.86

APT 2.965 9.85 0.301 –29.82 92.09

CAPM FF EGB APT

Jensen 1.155

(4.15)*

2.957

(7.15)*

2.840

(6.28)*

3.132

(4.62)*

FS 0.831

(3.34)*

2.798

(5.89)*

2.664

(5.32)*

2.695

(3.99)*

Panel B

Mean Std Deviation Sharpe Minimum Maximum

CAPM 0.845 2.27 0.372 –6.89 16.41

FF 1.052 2.70 0.389 –6.13 15.83

EGB 1.063 2.73 0.389 –7.28 16.25

APT 0.920 2.51 0.366 –5.61 12.01

CAPM FF EGB APT

Jensen 0.687

(5.34)*

0.852

(5.53)*

0.864

(5.48)*

0.743

(5.19)*

FS 0.505

(4.31)*

0.673

(5.26)*

0.675

(5.55)*

0.578

(4.36)*

* Significant at 5%.

The out of sample performance of monthly asset allocation strategies is evaluated between

February 1981 and April 2000. Expected returns are estimated from the conditional versions of

linear asset pricing models of the CAPM, Fama and French (1993) (FF), Elton, Gruber and Blake

(1996) (EGB) and APT models. Each panel includes summary statistics of performance and two

performance measures with t-statistics in parentheses. The two performance measures are the

Jensen (1968) measure and the conditional measure of Ferson and Schadt (1996) (FS). The t-

statistics are corrected for heteroskedasticity using White (1980). Panel A refers to the case where

there are no investment constraints and panel B refers to the case where there are short selling and

20% upper bound constraints in the risky assets. All performance numbers are monthly %.

266 JONATHAN FLETCHER