1. Trang chủ
2. » Luận Văn - Báo Cáo

# econometrics project the factors affecting the differences in ceo salary 1

40 6 0

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

THÔNG TIN TÀI LIỆU

Cấu trúc

• IV. DATA ANALYSIS AND RESULTS (12)
• V. CHECKING ERRORS IN THE MODEL 1, Multicollinearity (22)
• 4. F —stat = (Riz — Rk)/ (Kur — kr) (40)

Nội dung

In Vietnam, the average monthly salary for a chief executive officer is roughly 38,000,000 VND.. Chief Executive Officer average salary change by experience in Vietnam: The compensation

## DATA ANALYSIS AND RESULTS

1 Descriptive statistics We have a statistics description in the below table:

Mean 1281120 6923793 1718421 6180383 Median 1039000 3705200 1550000 5200000 Maximum 1482200 9764990 5630000 4180000 Minimum 223.0000 1752000 0500000 -5800000 Std Dev 1372345 1063327 8518509 6817705 Skewness 685493 4999125 1580820 2079588 Kurtosis 60.54128 3529968 6678555 9400044 Jarque-Bera 3047010 9955688 2026987 5073386

Sum 2677540 1447073 3591.500 12917.00 Sum Sq Dev 3.92E+08 2.35E+10 15093.52 966807.0 Observations 209 Statistic description 209 209 209

The following graphs show the relationship between salary and other factors which are sales, ROE , ROS respectively.

2.1 Testing on functional form of regression model In this section, we will use Eview10 to conduct the OLS on the lin-lin model, semi-log model (lin-log & log-lin model), and log-log model in order to determine the best linear unbiased estimator (BLUE) regression However, due to non-positive numbers in the ros figures in the excel file, we are unable to run the log(ros) on the "ceosal1" data set In order to determine the most suitable form of the model, we consider R-squared as the conditions because R-Ssquared represents “goodness of fit” of the model, which means how well the data fit the model The decision rule is the higher R2 of the model, the more preferable model In addition, through running four models, we will obtain R-squared (R2) and covariance(CV) of each model and then choose the optimal model with highest R-squared and lowest CV as follows: a Lin-lin model:

Dependent Variable: SALARY Method: Least Squares Date: 04/29/23 Time: 21:42 Sample: 1 209

Variable Coefficient Std Error t-Statistic Prob

R-squared 0.031914 Mean dependent var 1281.120 Adjusted R-squared 0.017747 S.D dependent var 1372.345 S.E of regression 1360.113 Akaike info criterion 17.28748 Sum squared resid 3.79E+08 Schwarz criterion 17.35144 Log likelihood -1802.541 Hannan-Quimn criter 17.31334 F-statistic 2.252699 Durbin-Watson stat 2.118133

Lin-lin model Obtaining = 0.0319, CV = 1.06166 b Log-lin model:

Log(salary)= B1+ B2*sales+ B3*roe+ B4*ros + u Dependent Variable: LOG(SALARY)

Method: Least Squares Date: 04/29/23 Time: 22:00 Sample: 1 209

Variable Coefficient Std Error t-Statistic Prob

R-squared 0.139699 Mean dependent var 6.950386 Adjusted R-squared 0.127109 S.D dependent var 0.566374 S.E of regression 0.529156 Akaike info criterion 1.583885 Sum squared resid 57.40118 Schwarz criterion 1.647853 Log likelihood -161.5160 Hannan-Quimn criter 1.609748 F-statistic 11.09619 Durbin-Watson stat 1.980010

Log-lin model Obtaining = 0.1397, CV = 0.07613 c Lin-log model:

Salary= B1+ B2*log(sales)+ B3*log(roe)+ B4*ros +u

Dependent Variable: SALARY Method: Least Squares Date: 04/29/23 Time: 22:07 Sample: 1 209

Variable Coefficient Std.Error t-Statistic Prob

Cc -1996.902 9918855 -2013239 00454 LOG(SALES) 2905537 = 98.38641 2.953190 00035 LOG(ROE) 312.2340 1749482 1.784722 0.0758 ROS 0.312329 1.484193 0.210437 0.8335

R-squared 0.053717 Mean dependent var 1281.120 Adjusted R-squared 0.039869 S.D dependent var 1372.345 S.E of regression 1344.710 Akaike info criterion 17.26470 Sum squared resid 3.71E+08 Schwarz criterion 17.32867 Log likelihood -1800.161 Hannan-Quimn criter 17.29056 F-statistic 3.879034 Durbin-Watson stat 2.135835 Prob(F-statistic) 0.009994

Lin-log model Obtaining = 0.0537, CV= 1.0496 d Log-log model:

Log(salary)= 81+ B2*log(sales)+ B3*log(roe)+ B4*ros + u

Dependent Variable: LOG(SALARY) Method: Least Squares

Variable Coefficient Std Error t-Statistic Prob

LOG(SALES) 0.282352 0.035851 7.875731 0.0000 LOG(ROE) 0.220398 0063749 3.457277 0.0007 ROS 0000482 0.000541 0.890893 0.3740

R-squared 0.262316 Mean dependent var 6.950386 Adjusted R-squared 0.251521 S_D dependent var 0.566374 S.E of regression 0.489997 Akaike info criterion 1.430118 Sum squared resid 4921987 Schwarz criterion 1.494086 Log likelihood -145.4473 Hannan-Quimn criter 1.455981 F-statistic 2429892 Durbin-Watson stat 2.008253 Prob(F-statistic ) 0.000000

Log-log model Obtaining =0.2623, CV= 0.0705

From Eview tables we can see that with the highest R2 means that the highest percentage change in salary is explained by sales, roe, ros jointly, and lowest CV means the most preferred model, we find out the log-log model is the optimal one

The equation is: Log(salary)= B1+ B2*log(sales)+ B3*log(roe)+ B4*ros +u

2.2 Testing the overall significance on all coefficients We have the equation of log-log model:

Log(salary)= 81+ B2*log(sales)+ B3*log(roe)+ B4*ros + u In hypothesis testing, we use significance level of 10%

The number of observations is n = 209 We use the F-test to test the overall significance of all coefficients This test aims to check the effect of all independent variables

H1: 82z0 or B3z0 or B40 ° Compute: F-statistic = 24.2969, = 0.2623 ° Obtain: = 2.08 Compare: F-statistic > => Reject HO Conclusion: Hence, there is enough statistical evidence to conclude that at least one of coefficients is different from zero.

2.3 Testing the model when dropping one variable

Dependent ¥ariable: LOGCSALARY) Method: Least Squares Date: 05/02/23 Time: 22:41

Variable Coefficient Std Error t Statistic Prob

Adjusted R-squared 0.252270 \$.D dependent var 0.566374 5.E of regression 0.489751 Akaike info criterion 1.424413 Sum squared resid 49.41044 Schwarz criterion 1.472389 Log likelihood -146.8511 Hannan-Quinn criter 1.443810

Log(salary) + B2*log(sales) + B3 * log(roe)+ B4 * ros + u (UR)

Log - log model after dropping “ros”

By dropping "ros," we have a new equation:

Log(salary)= B1’ + B2’ * log(sales) + B3’ * log(roe) + u (R) ° Hypothesis testing:

H1: B4 0 ° Test statistic: F-statistic = = 0.7781 ° Critical value: = 2.71 ° Conclusion: F-sta <

— There is enough statistical evidence to conclude that 4 is not significantly different from zero at a = 10%

2.4 Testing on dummy variables Denoting the variables:

Dependent Variable: LOG(SALARY) Method: Least Squares Date: 05/01/23 Time: 15:02 Sample: 1 209

Variable Coefficient Std Error t-Statistic Prob

Cc 4.441963 0.369222 12.03062 0.0000 LOG(SALES) 0.254944 0.034486 7.392757 0.0000 LOG(ROE) 0.128543 0.065945 1.949242 0.0527 ROS 5.05E-05 0.000544 0.092879 0.9261 D2 0.153797 0.090310 1.702998 0.0901 D3 0.206649 0.088710 2329479 0.0208 D4 -0.299626 0.100457 -2 982627 0.0032

R-squared 0.348029 Mean dependent var 6.950386 Adjusted R-squared 0.328663 S_D dependent var 0.566374 S.E of regression 0.464059 Akaike info criterion 1.335311 Sum squared resid 43.50092 Schwarz criterion 1.447255 Log likelihood -132 5400 Hannan-Quimn criter 1.380570 F-statistic 1797162 Durbin-Watson stat 2.225653

Log(salary) = B1+ B1' * D2+ B2 * log(sales) + B3 * log(roe) + B4 * ros

To determine if B1' is significant, we use hypothesis testing ° Hypothesis testing:

Test statistic: t-statistic = 1.703 Critical value: 'Tco.os, ›os = 1.645

= There is enough statistical evidence to conclude that B1'is significant different from zero at a%

Log(salary)= 4.5958 + 0.2549log(sales)+ 0.1285log(roe)+ 5.05- 05ros

Log(salary)= B1 + B1”’ * D3+ B2 * log(sales)+B3 * log(roe)+ B4 * ros +u

To determine if 1" is significant, we use hypothesis testing ° Hypothesis testing:

Test statistic: t-statistic = 2.3295 ° Critical value: Tco.0s, 205 = 1.645 Conclusion: t-sta > tco.0s, 205

= There is enough statistical evidence to conclude that ,’’ is significant different from zero at a = 10%

= D3 = 1, we have the equation: log(salary)= 4.6486+ 0.2549log(sales)+ 0.1285log(roe)+ 5.05E- 05ros

Log(salary)= B1 + B1”?' * D4+ B2 * log(sales)+B3 * log(roe)+ B4 * ros +u

To determine if 1" is significant, we use hypothesis testing ° Hypothesis testing:

H0:B1” =0 H1:B1” 0 Test statistic: t-statistic = 2.9826 Critical value: 'Tco.os, ›os = 1.645 Conclusion: t-sta > tco.0s, 205

= There is enough statistical evidence to conclude that i'” is significant different from zero at a = 10%

= D4 = 1, we have the equation: log(salary) = 4.1424 + 0.2549log(sales)+ 0.1285log(roe)+ 5.05E- 05ros

4) With the others, we have the equation: log(salary)=4.442 + 0.2549log(sales)+ 0.1285log(roe)+ 5.05E-05ros

## CHECKING ERRORS IN THE MODEL 1, Multicollinearity

Ngày đăng: 30/08/2024, 15:41

TỪ KHÓA LIÊN QUAN

TÀI LIỆU CÙNG NGƯỜI DÙNG

• 28 3 0
• 14 0 0
• 8 8 0
• 12 395 0
• 18 3 0
• 14 5 0

TÀI LIỆU LIÊN QUAN

• 95 1 0
• 129 0 0
• 38 2 0
• 77 2 0
• 70 9 0
• 77 5 0
• 85 4 0
• 76 1 0