Sunteți pe pagina 1din 3

Q12 (3 marks) Formulate a joint-hypothesis test to test whether the Fama-French 3-Factor model explains the

stock returns better than the model given by Equation (1). Perform the hypothesis test by calculating the
homoskedasticity-consistent F-Statistic, using the relevant formula.
Verify your conclusion by performing the Wald test in Eviews and considering p-values. What is your
conclusion?
Ans:
(1)
In contrast, the FamaFrench model uses three variables:
(2)
Dependent Variable:
RBHP_RF
Dependent Variable: RBHP_RF
Method: Least Squares
Date: 10/26/14 Time: 22.33
Sample: 1 46
Included observations: 46
Variable

Coefficient Std. Error

t-Statistic

Prob.

C
RM_RF
SMB
HML

0.009590
1.076912
-0.220345
-1.024346

1.286617
7.427982
-1.496255
-4.024213

0.2053
0.0000
0.1421
0.0002

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.649697
0.624675
0.047636
0.095305
76.85297
25.96535
0.000000

0.007454
0.144980
0.147264
0.254546

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

Wald Test:
Equation: Untitled
Test Statistic
F-statistic
Chi-square

Value

df

Probability

27.80842
55.61685

(2, 42)
2

0.0000
0.0000

Null Hypothesis: C(2)=C(3)=0

-0.008037
0.077755
-3.167521
-3.008508
-3.107954
2.285920

Null Hypothesis Summary:


Normalized Restriction (= 0)
C(2)
C(3)

Value

Std. Err.

1.076912
-0.220345

0.144980
0.147264

Restrictions are linear in coefficients.


H0:
H1:
=0.05
Testing if and ; where B2 and B3 are the intercept/slope coefficients.
From this null hypothesis, none of the regressors explains any of the variations in monthly return on BHP
stock
Unrestricted (H1):

R2 = 0.649697;
F-Stat(act) = 27.80842 (Wald Test)

Restricted (Ho):

R2 = 0
F[cv] = F(q, n-k-1)df = F(2, 42) df = 3.220

Reject H0 if:

F[act] > F[cv]


27.80842 > 3.220

Therefore, Ho is rejected (that is, H0:


) with a 95% confidence. Consequently, SMB and HML
may be significant. As SMB and HML is significant, the Fama-French 3-Factor model (2) explains the
stock returns better than the model given by the equation (1).
Q13 (3 marks) A Financial Analyst believes that the effect of book-to-market values (HML) on stock returns
is twice as great as the effect of market capitalization (SMB). Formulate an appropriate hypothesis test and use
re-parametrisation to convert it to a simple t-test to test the assertion. Perform the required regression and
paste your Eviews output below. State your conclusion at the 5% level.
Ans:
Dependent Variable: RBHP_RF
Method: Least Squares
Date: 10/27/14 Time: 17:12
Sample: 1 46
Included observations: 46
Variable

Coefficient

C
RM_RF
SMB_HML
HML

0.009590
1.076912
-0.220345
-1.244691

Std. Error

t-Statistic

0.007454 1.286617
0.144980 7.427982
0.147264 -1.496255
0.337021 -3.693212

Prob.
0.2053
0.0000
0.1421
0.0006

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.649697
0.624675
0.047636
0.095305
76.85297
25.96535
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

-0.008037
0.077755
-3.167521
-3.008508
-3.107954
2.285920

Dependent Variable: RBHP_RF


H0: 2 2 = 3 => let = 3 2 2
Method: Least Squares
H1: 2 2 < 3 or 2 2> 3
Date: 10/27/14 Time: 17:12
Sample: 1 46
= 3 2 2
Included observations: 46
3 = + 2 2
Variable
Coefficient Std. Error t-Statistic
Prob.
rBHP - rf = 0 + 1(rm-rf) + 2 2 SMB + 3 HML
= 0 + 1(rm-rf)+
+ ( + 2 20.007454
) HML
C 2 2 SMB0.009590
1.286617 0.2053
= 0 + 1(rm-rf)
+2

(SMB+HML)
+

HML
RM_RF 2
1.076912 0.144980 7.427982 0.0000
SMB_HML
-0.220345 0.147264 -1.496255 0.1421
H0: =0
HML
-1.244691 0.337021 -3.693212 0.0006
H1: >0
R-squared
0.649697 Mean dependent var -0.008037
Adjusted
R-squared
0.077755
Tcrit = 1.645 because =0.05 0.624675 S.D. dependent var
0.047636 Akaike info criterion -3.167521
Tstat S.E. of regression
Sum squared resid
0.095305 Schwarz criterion
-3.008508
Log likelihood
76.85297 Hannan-Quinn criter. -3.107954
= -1.49626
F-statistic
25.96535 Durbin-Watson stat
2.285920
Prob(F-statistic)
0.000000
As a result, we do not reject the null hypothesis. HML is not twice as big in explaining stock returns as SMB.

S-ar putea să vă placă și