Home / Capital Structure

Testing for Non-Significant Variables in the Model

Chart 4.3: Chart depicting the fluctuations of variables


60


50


40


30


20


10


0


-10


-20

10 20 30 40 50 60 70 80



ROE PURE SKIN

TSCDHUUHINH ROS

ROA



Source: Implemented and extracted from Eviews 6 software


4.2. Implementing the regression model


Use Eviews to run a regression model for the variable Liabilities / Total Assets. We have the following model output table:

Sample: 1 80

Included observations: 80





Variable

coefficient

Std. Error

t-Statistic

Prob.

C

0.460695

0.037976

12.13121

0.0000

ROA

-1.096318

0.451612

-2.427563

0.0176

ROE

0.024242

0.188911

0.128326

0.8982

ROS

0.049059

0.106985

0.458562

0.6479

CONFIDENTIALITY

-0.005001

0.002052

-2.437338

0.0172

TSCDHUUHINH

-10.15646

3.093439

-3.283227

0.0016

R-squared

0.785042

Mean dependent var

0.325284

Adjusted R-squared

0.762418

SD dependent var

0.252787

SE of regression

0.224219

Akaike info criterion

-0.080345

Sum squared residue

3.720301

Schwarz criterion

0.098307

Log likelihood

9.213801

Hannan-Quinn critic.

-0.008718

F-statistic

5.282545

Durbin-Watson statistics

1.116970

Prob(F-statistic)

0.000333



Maybe you are interested!

Testing for Non-Significant Variables in the Model

Dependent Variable: DA Method: Least Squares Date: 06/23/16 Time: 12:02


From the model output table, we have the econometric equation in the following form:

Y = 0.460695 – 1.096318X 1 + 0.024242X 2 + 0.049059X 3 – 0.005001X 4 -

10.15646X 5 + ε

In which: Y: DA

X 1 : ROA X 2 : ROE X 3 : ROS

X 4 : Liquidity

X 5 : Tangible fixed assets

Based on the above equation, we have the regression mean of 0.33 times and the regression standard deviation of 0.25 times.

Model meaning:

- If a business's return on assets (ROA) increases/decreases 1 time, its debt/total assets ratio decreases/increases 1.096318 times.

- If a business's return on equity (ROE) increases/decreases 1 time, the debt/total assets ratio increases/decreases 0.024242 times.

- If a business's return on sales (ROS) increases/decreases 1 time, the debt/total assets ratio increases/decreases 0.049059 times.

- If a business's liquidity increases/decreases once, the debt/total assets ratio decreases/increases 0.005001 times.

- If the ratio of tangible fixed assets to total assets increases/decreases 1 time, then the debt/total assets ratio decreases/increases 10.15646 times.

4.3. Correlation coefficient between variables


Table 4.2: Table describing the correlation between variables



ROA

ROE

ROS

Liquidity

Tangible fixed assets

ROA

1.000000

0.664512

0.738104

-0.039034

-0.072244

ROE

0.664512

1.000000

0.738836

0.037477

-0.038358

ROS

0.738104

0.738836

1.000000

0.064637

-0.167173

Liquidity

-0.039034

0.037477

0.064637

1.000000

0.069428

Tangible fixed assets

-0.072244

-0.038358

-0.167173

0.069428

1.000000

Source: Data analysis results from Eviews 6 software


The correlation coefficient between return on total assets and return on equity is 0.664512, indicating that the correlation between these two variables is positive.

The correlation coefficient between return on total assets and return on sales is 0.738104, showing that the correlation between these two variables is positive.

The correlation coefficient between return on total assets and current debt ratio is -0.039034, indicating that the correlation between these two variables is negative.

The correlation coefficient between the return on total assets and the ratio of tangible fixed assets to total assets is -0.072244, indicating that the correlation between these two variables is a negative correlation.

The correlation coefficient between return on equity and return on sales is 0.738836, indicating that the correlation between these two variables is positive.

The correlation coefficient between return on equity and current debt ratio is 0.037477, indicating that the correlation between these two variables is positive.

The correlation coefficient between return on equity and the ratio of tangible fixed assets to total assets is -0.038358, indicating that the correlation between these two variables is a negative correlation.

The correlation coefficient between the return on sales and the current debt ratio is 0.064637, indicating that the correlation between these two variables is a positive correlation.

The correlation coefficient between the return on sales and the ratio of tangible fixed assets to total assets is -0.167173, indicating that the correlation between these two variables is a negative correlation.

The correlation coefficient between the current debt payment ratio and the ratio of tangible fixed assets to total assets is 0.069428, showing that the correlation between these two variables is a positive correlation.

4.4. Testing the research model


4.4.1. Testing for insignificant variables in the model

After running the debt to total assets regression model, some variables were not meaningful, so we re-examined the unnecessary variables in the model to eliminate multicollinearity.

To verify the insignificant variables, we rerun the Eviews model with a 5% significance level.

Verify the significance of the variable ROA . We hypothesize to verify the variable ROA in the debt to total assets model.

Let H 0 : β 1 = 0, ROA is an unnecessary variable in the model.

H 1 : β 1 ≠ 0, ROA is a necessary variable in the model.


Redundant Variables: ROA


F-statistic

5.893061

Prob. F(1,74)


0.0176

Log likelihood ratio

6.129912

Prob. Chi-Square(1)


0.0133


Test Equation:





Dependent Variable: DA





Method: Least Squares





Date: 06/23/16 Time: 14:31





Sample: 1 80





Included observations: 80





Variable

coefficient

Std. Error

t-Statistic

Prob.

C

0.443202

0.038483

11.51678

0.0000

ROE

-0.092249

0.188581

-0.489176

0.6261

ROS

-0.080286

0.095751

-0.838483

0.4044

CONFIDENTIALITY

-0.004335

0.002099

-2.065409

0.0423

TSCDHUUHINH

-10.58307

3.187598

-3.320076

0.0014

R-squared

0.204353

Mean dependent var

0.325284

Adjusted R-squared

0.161919

SD dependent var

0.252787

SE of regression

0.231418

Akaike info criterion

-0.028721

Sum squared residue

4.016571

Schwarz criterion

0.120156

Log likelihood

6.148845

Hannan-Quinn critic.

0.030968

F-statistic

4.815734

Durbin-Watson statistics

1.045681

Prob(F-statistic)

0.001639



From the table of unnecessary variables in the debt to total assets model, we see that Probability = 0.0176 < α = 0.05. Thus, we reject H 0, which means that the ROA variable is a necessary variable in the model.

Verify the significance of the ROE variable . We hypothesize to verify the ROE variable in the debt to total assets model.

Let H 0 : β 2 = 0, ROE is an unnecessary variable in the model.

H 2 : β 2 ≠ 0, ROE is a necessary variable in the model.

Redundant Variables: ROE


0.016468

Prob. F(1,74)

0.8982

Log likelihood ratio

0.017801

Prob. Chi-Square(1)

0.8939

F-statistic


Test Equation: Dependent Variable: DA Method: Least Squares

Date: 06/23/16 Time: 14:32 Sample: 1 80

Included observations: 80



Variable

coefficient

Std. Error

t-Statistic

Prob.


C


0.459803


0.037089


12.39722


0.0000

ROA

-1.081596

0.433926

-2.492584

0.0149

ROS

0.055879

0.092239

0.605809

0.5465

CONFIDENTIALITY

-0.004999

0.002038

-2.452412

0.0165

TSCDHUUHINH

-10.11241

3.054109

-3.311085

0.0014


R-squared


0.262878


Mean dependent var



0.325284

Adjusted R-squared

0.223564

SD dependent var


0.252787

SE of regression

0.222744

Akaike info criterion


-0.105123

Sum squared residue

3.721129

Schwarz criterion


0.043754

Log likelihood

9.204901

Hannan-Quinn critic.


-0.045434

F-statistic

6.686753

Durbin-Watson statistics


1.117649

Prob(F-statistic)

0.000117





From the table verifying unnecessary variables in the debt to total assets model, we see that Probability = 0.8982 > α = 0.05. Thus, we accept H 0, which means that the ROE variable is an unnecessary variable in the model.

Verifying the significance of the ROS variable: We set up a hypothesis for verifying the ROS variable in the debt to total assets model.

Let H 0 : β 3 = 0, ROS is an unnecessary variable in the model.

H 3 : β 3 ≠ 0, ROS is a necessary variable in the model.


Redundant Variables: ROS


F-statistic

0.210279

Prob. F(1,74)

0.6479

Log likelihood ratio

0.227007

Prob. Chi-Square(1)

0.6338

Test Equation: Dependent Variable: DA Method: Least Squares

Date: 06/23/16 Time: 14:32 Sample: 1 80

Included observations: 80




Variable

coefficient

Std. Error t-Statistic

Prob.

C

0.462717

0.037520 12.33254

0.0000

ROA

-0.993179

0.389552 -2.549539

0.0128

ROE

0.067276

0.163087 0.412520

0.6811

CONFIDENTIALITY

-0.004881

0.002024 -2.411125

0.0184

TSCDHUUHINH

-10.45562

3.007904 -3.476049

0.0008

R-squared

0.260947

Mean dependent var

0.325284

Adjusted R-squared

0.221531

SD dependent var

0.252787

SE of regression

0.223036

Akaike info criterion

-0.102507

Sum squared residue

3.730873

Schwarz criterion

0.046369

Log likelihood

9.100298

Hannan-Quinn critic.

-0.042819

F-statistic

6.620321

Durbin-Watson statistics

1.129577

Prob(F-statistic)

0.000128




From the table verifying unnecessary variables in the debt to total assets model, we see that Probability = 0.6479 > α = 0.05. Thus, we accept H 0, which means that the ROS variable is an unnecessary variable in the model.

Testing the significance of liquidity variable: We hypothesize for testing liquidity variable in the debt to total assets model.

Set H 0 : β 4 = 0, TINHTHANHKHOAN is an unnecessary variable in the model.

H 4 : β 4 ≠ 0, TINHTHANHKHOAN is a necessary variable in the model.

Redundant Variables: TINHTHANHDRILL



F-statistic

5.940615

Prob. F(1,74)

0.0172

Log likelihood ratio

6.177516

Prob. Chi-Square(1)

0.0129


Test Equation:


Dependent Variable: DA

Method: Least Squares

Date: 06/23/16 Time: 14:33

Sample: 1 80

Included observations: 80

coefficient

Std. Error t-Statistic

Prob.

C

0.421528

0.035524 11.86607

0.0000

ROA

-0.949051

0.462059 -2.053963

0.0435

ROE

0.020086

0.195026 0.102990

0.9182

ROS

0.015729

0.109546 0.143582

0.8862

TSCDHUUHINH

-10.84199

3.180476 -3.408919

0.0011

R-squared

0.203880

Mean dependent var

0.325284

Adjusted R-squared

0.161420

SD dependent var

0.252787

SE of regression

0.231487

Akaike info criterion

-0.028126

Sum squared residue

4.018961

Schwarz criterion

0.120751

Log likelihood

6.125043

Hannan-Quinn critic.

0.031563

F-statistic

4.801715

Durbin-Watson statistics

1.004489

Prob(F-statistic)

0.001673



Variable


From the table of unnecessary variables in the debt to total assets model, we see that Probability = 0.0172 < α = 0.05. Thus, we reject H 0, which means that liquidity is a necessary variable in the model.

Verifying the significance of the variable ratio of tangible fixed assets to total assets: We set up a hypothesis for verifying the variable of tangible fixed assets in the model of liabilities to total assets.

Set H 0 : β 5 = 0, TSCDHUUHINH is an unnecessary variable in the model.

H 5 : β 5 ≠ 0, TSCDHUUHINH is a necessary variable in the model.

Redundant Variables: TSCDHUUHINH


F-statistic

10.77958

Prob. F(1,74)

0.0016

Log likelihood ratio

10.87917

Prob. Chi-Square(1)

0.0010


Test Equation: Dependent Variable: DA Method: Least Squares

Date: 06/23/16 Time: 14:33 Sample: 1 80

Included observations: 80


Variable

coefficient

Std. Error

t-Statistic

Prob.

C

0.393955

0.034105

11.55131

0.0000

-1.180550

0.479379

-2.462666

0.0161

ROE

-0.044586

0.199610

-0.223364

0.8239

ROS

0.123136

0.111188

1.107458

0.2716

CONFIDENTIALITY

-0.005614

0.002173

-2.583926

0.0117


R-squared


0.155689


Mean dependent var


0.325284

Adjusted R-squared

0.110659

SD dependent var

0.252787

SE of regression

0.238390

Akaike info criterion

0.030645

Sum squared residue

4.262237

Schwarz criterion

0.179521

Log likelihood

3.774215

Hannan-Quinn critic.

0.090334

F-statistic

3.457453

Durbin-Watson statistics

0.893262

Prob(F-statistic)

0.011989



Comment


Agree Privacy Policy *