Professional Documents
Culture Documents
Name
Student ID.
Dept/Year
Problem:
C3. The file CEOSAL2.RAW contains data on 177 chief executive officers and can be used
to examine the effects of firm performance on CEO salary.
i.
Estimate a model relating annual salary to firm sales and market value. Make the model
of the constant elasticity variety for both independent variables. Write the results out in
equation form.
ii.
Add profits to the model from part (i). Why can this variable not be included in
logarithmic form? Would you say that these firm performance variables explain most of
the variation in CEO salaries?
iii.
Add the variable ceoten to the model in part (ii). What is the estimated percentage return
for another year of CEO tenure, holding other factors fixed?
iv. Find the sample correlation coefficient between the variables log(mktval) and profits. Are
these variables highly correlated? What does this say about the OLS estimators?
Solution:
i. Regression result can be explained in Table 1.
Table 1 Regression result of lsalary toward lsales and lmktval
Variable
C
LSALES
LMKTVAL
Coefficient
4.620918
0.162128
0.106708
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.299114
0.291057
0.510294
45.30966
-130.5594
37.12853
0
Std. Error
0.254408
0.03967
0.050124
t-Statistic
18.16339
4.086898
2.128881
Prob.
0
0.0001
0.0347
6.582848
0.606059
1.509146
1.562979
1.530978
2.092116
Based on Table 1, regression model of constant elasticity of annual salary toward the firm
sales and market value can be expressed by equation below:
^
log (slry )=4.62+ 0.162 log ( sales ) +0.107 log ( mktval) . (1)
This model means:
If sales rise by 1%, the annual CEO salary will rise by about 0.162%, holding the market
value fixed. In addition, if market value rise by 1%, the annual CEO salary will rise by
0.107%, holding sales fixed.
Then, the coefficient of determination (R-squared) of this model is 29.91%. It means that
the proportion of total variation on salary can be explained by sales and market value by
29.91%. Then, the remaining (1-29.91)% can be explained by other explanatory variables that
is not in the model yet (all of this include in error term).
Moreover, we can look at the parameter significance test. In the overall test, p-value of F
statistics is 0. It shows that at least one explanatory variabel influences the explained variable.
Then, in the individual test we get p-value of t statistic test. Both p-value of sales and market
HOMEWORK 3 ECONOMETRICS
value are less than 5%, so we conclude that sales and market value have significant influence
to salary.
ii. In this part, we add the profit variable as explanatory variable in the regression model. Profit
doesnt need to be converted as logarithmic form because its value can be negative that
means the firm lose. Table 2 is the result of this regression.
Table 2 Regression result of lsalary toward lsales, lmktval, and profit
Variable
C
LSALES
LMKTVAL
PROFIT
Coefficient
4.686924
0.161368
0.097529
3.57E-05
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.299337
0.287186
0.511686
45.29524
-130.5312
24.63629
0
Std. Error
0.379729
0.03991
0.063689
0.000152
t-Statistic
12.3428
4.043299
1.531334
0.234667
Prob.
0
0.0001
0.1275
0.8147
6.582848
0.606059
1.520127
1.591904
1.549237
2.096547
According to Table 2, regression model of constant elasticity of annual salary toward the
firm sales, market value, and profit can be expressed by equation below:
^
log (slry )=4.69+ 0.161 log ( sales ) +0.0975 log ( mktval ) +0.000036 profit
..(2)
Coefficient
4.55778
0.162234
0.10176
2.91E-05
0.011685
R-squared
Adjusted R-squared
S.E. of regression
0.318299
0.302445
0.506179
Std. Error
0.380255
0.039483
0.063033
0.00015
0.005342
t-Statistic
11.98612
4.109003
1.61439
0.193231
2.187313
Prob.
0
0.0001
0.1083
0.847
0.0301
6.582848
0.606059
1.50399
HOMEWORK 3 ECONOMETRICS
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
44.06941
-128.1031
20.07748
0
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
1.593712
1.540378
2.047158
Based on Table 3, regression model of constant elasticity of annual salary toward the firm
sales, market value, profit, and CEO tenure can be expressed by equation below:
^
log (slry )=4 .56 +0 .162 log ( sales )+ 0 .102 log ( mktval ) +0 . 000029 profit +0 . 0117 ceoten
. (3)
(x ix )( y i y )
r ( X ,Y )=
i=1
177
i=1
=0.777
177
( x i x )2
i=1
( y i y )2
Correlation between profit and log(market value) is 0.777. We can say that profit and log
(market value) are highly correlated. Actually, profit is a short term measure of how the firm
is doing, while market value is based on past, current, and expected future profitability.
With profit and log (market value) are highly correlated, it will make OLS estimator still
^
^
E( lmktval)=lmktval ), but the variance (
unbiased ( E( profit )= profit
and
^
^
var ( profit ) and var ( lmktval) ) become larger.