CHAPTER 4:
4.0 Introduction
In this chapter, data is analyzed using SPSS computer program to determine the relationships between capital structure and company performance for 200 public listed companies in Malaysia. Descriptive statistics, Pearson correlation matrix, and multiple regression statistics are used in the following data analysis. Descriptive statistics explain the basic features of each individual variable in this study separately. The Pearson correlation matrix and multiple regression statistics are used to determine the relationships between independent variable and dependent variables. All the tables below are simplified to facilitate the interpretation of data. The complete sets of tables from SPSS results are available in the appendix.
4.1 Analysis of Descriptive Statistic
Table 1 shows the descriptive statistics of all the variables with 1000 observations each in this study. There are the minimum values, maximum values, means, standard deviations, skewness and kurtosis values for each of the variables in this study. Skewness and kurtosis are used to characterize the location and variability of a data set. Skewness measures the symmetry of data set with respect to its dispersion from the mean. Kurtosis measures whether the data in a data set are peaked or flat relative to a normal distribution. It is also used to predict the probability of extreme values in a data set as extreme values may distort the accurate results. However, the size has wider spread around the mean.
ROA, ROE and TQ are the three dependent variable measuring performance of companies. They are defined as Net income to Total Asset, Net Income to Total Equity and Total Market Value of Firm to Total Asset Value respectively. TD is the independent variables that measure the capital structure of companies. It is defined as the short-term debt plus long-term debt. Size, AG, SG and EFF represents the control variables of company size, company asset growth, company sales growth and company efficiency (asset turnover) respectively. The minimum value of Size, Efficiency, Total Debt and Tobin's Q are zero. This is due to few observations from the total observations are zero.
The skewness of ROE is -2.462 indicating that most values are skewed to the left of normal distribution and there might have extreme values to the left. The skewness of ROA is 19.434 which mean that most values are skewed to the right and there might have extreme values to the right. The kurtosis for SG, AG, EFF, ROA, ROE, TD and TQ are more than 3. This indicates their respective distributions are Leptokurtic distributions. This means that their values are concentrated around the mean and there is high probability of extreme values. However, the kurtosis for SIZE is less than 3. This indicated the values are wider spread around the mean and it is called Platykurtic distribution. The complete table for descriptive statistics of all data is available in appendices Table 1a.
Table 1: Descriptive Statistics of All Data
Minimum
Maximum
Mean
Std. Deviation
Skewness
Kurtosis
SG
-97.8667
2254.7069
13.327528
89.3902858
17.072
403.459
SIZE
0.0000
9.4381
5.771772
1.2868606
-.449
2.581
AG
-97.3013
383.0709
6.048183
24.7620722
5.336
65.758
EFF
0.0000
4.9413
.732523
.5834356
1.538
4.525
ROA
-87.3716
771.4512
4.929755
31.8906032
19.434
429.886
ROE
-215.3386
215.7610
3.919825
28.2462111
-2.462
29.805
TD
0.0000
3819.8220
173.115962
376.7933687
4.992
34.217
TQ
0.0000
10.9200
1.016603
.8996325
7.043
63.632
Valid N (listwise)
4.2 Analysis of Pearson Correlation Matrix
Table 2 shows the bivariate correlations among the AG, SG, SIZE, EFF, ROE, ROA, TQ and TD. The Pearson correlation is used to find a degree of linear relationship or correlations between at least two continuous variables. It can be used to measure the strength of association between the two variables. A correlation values can be ranged between -1.0 to 1.0. A correlation of -1.0 shows perfect negative relationship, whilst +1.0 shows s perfect positive relationship. The dependent variable of the study is the ROA, ROE and TQ. It is represented Net Income to Total Assets, Net Income to Total Equity and Total Market Value of Firm to Total Asset Value respectively. The p-values are represented by significance values in the second row of each variable. The r-values are represented by the first row of each variable.
4.2.1 Return on Asset
Under the measure of ROA, the significance value for AG and EFF are 0.000 and 0.009 respectively. Since the values are less than 0.01, correlations exist between the ROA and the AG and EFF. The r-values of AG and EFF are 0.111 and 0.083 respectively. This shows that AG and EFF are positively correlated with ROA.
4.2.2 Return on Equity
Under the measure of ROE, the significance value for SIZE, AG and EFF are 0.000. This indicates these three variables are significant correlated with ROE at 1% significant level. The r-values of SIZE, AG and EFF are 0.256, 0.227 and 0.328 respectively. It means the SIZE, AG and EFF are positively correlated with ROE. While, the significant values of SG and TD are 0.003 and 0.005 respectively; it shows the SG and TD are significant correlated with ROE at 1% significant level. The r-values for the SG and TD are 0.094 and 0.088 respectively. Since the r-value is positive value, then the ROE has positive relationship with the SG and TD.
4.2.3 Tobin's Q
Under the measure of TQ, the significant value of SIZE and EFF are 0.005 and 0.000 respectively. Since the value is less than 0.01, then SIZE and EFF is significant correlated with TQ. The r-value of SIZE and EFF are 0.089 and 0.31; this indicates both have positive relationship with TQ. On the other hand, the significant values of AG and TD are 0.036 and 0.022 respectively. It shows that AG and TD are significant correlated with TQ at 1% significant level. The r-value of AG and TD are 0.089 and 0.073 respectively. Therefore, both have positive relationship with TQ.
Table 2: Pearson Correlation Matrix
SG
SIZE
AG
EFF
ROA
ROE
TD
TQ
SG
Pearson Correlation
1
.032
.205**
.036
.037
.094**
-.004
-.020
Sig. (2-tailed)
.311
.000
.252
.245
.003
.903
.532
SIZE
Pearson Correlation
1
.153**
-.136**
-.041
.256**
.629**
.089**
Sig. (2-tailed)
.000
.000
.194
.000
.000
.005
AG
Pearson Correlation
1
.086**
.111**
.227**
.113**
.066*
Sig. (2-tailed)
.006
.000
.000
.000
.036
EFF
Pearson Correlation
1
.083**
.328**
-.161**
.310**
Sig. (2-tailed)
.009
.000
.000
.000
ROA
Pearson Correlation
1
-.167**
-.018
.183**
Sig. (2-tailed)
.000
.579
.000
ROE
Pearson Correlation
1
.088**
.133**
Sig. (2-tailed)
.005
.000
TD
Pearson Correlation
1
.073*
Sig. (2-tailed)
.022
TQ
Pearson Correlation
1
Sig. (2-tailed)
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
4.3 Analysis of outcome for 200 companies
Table 3.0 summarizes the relationship between ROA, ROE and TQ toward the use of Debt. In this study, the performance of companies is measure by accounting measures (ROA and ROE) and non-accounting measures (TQ) while the capital structure is measure by Total Debt. The capital R refers to a measure of the strength and direction of the linear relationship between the company performance and capital structure while R-Square (also known as coefficient of determination) is uses to indicate how much dependent variable (company performance) can be explained by the independent variable (capital structure). Next, the Anova is uses access whether regression equation is explaining a statistically significant of the variability in the dependent variable from variability in the independent variables.
4.3.1 Analysis outcome of ROA
Table 3.0 explains that 20% of company performance in term of ROA. The other 80% of ROA could be explained by other characteristics of company which is not included in this model. The F-statistic value of this model is 8.529 and it is significant at 1% level. In this model, the variables that significant with ROA are Asset Growth (AG) and Efficiency (EFF).
The result shows that AG has a positive relationship (t=-1.512) with ROA at 1% significant level with a coefficient of 0.109. This indicates any increase in AG would lead to the increases in ROA. Efficiency shows a positive relationship (t=2.113) at 5% significant level with coefficient of 0.068. It shows efficiency is moving along with the movement of ROA. These finding is consistent with Zuraidah et al (2010) that investigated the capital structure effect on firm performance with focusing on consumer and industrial sectors.
Model Summary
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.143a
.020
.015
31.6426762
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY),
ASSET GROWTH, Ln TOTAL ASSET
b. Dependent Variable: RETURN ON ASSET
ANOVA
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
49793.601
6
8298.933
8.529
.000b
Residual
966199.962
993
973.011
Total
1015993.563
999
a. Dependent Variable: RETURN ON ASSET
b. Predictors: (Constant), TOBIN'S Q, SALES GROWTH, TOTAL DEBT, ASSET GROWTH, ASSET TURNOVER (EFFICIENCY), Ln TOTAL ASSET
4.3.2 Analysis outcome of ROE
Data analysis on table 3.0 explains that 22.8 % of company performance in term of ROE. This model is significant at 1% significant level with F-statistic value of 52.469. In this model, the variables that significant with ROE are SIZE, AG, EFF and TD.
The result shows that SIZE has a positive relationship (t=9.075) with ROE at 1% significant level with coefficient of 0.328. This indicates that the larger the firm size, the higher its ROE. This may be true because shareholders are likely to invest more into a firm that makes more profit than those making losses. This result is consistent with Majumdar and Chibber (1999), Abor (2005), Zuraidah et al (2012) and Hammes and Chen (2004).
ROE has a positive relationship (t=5.089) with AG at 1% significant level with coefficient of 0.148. An increase in one unit of ROE would affect an increase in one unit of asset. It means that if asset is increase, the ROE also increase. This result is consistent with Abor (2005) and Zuraidah et al (2012). Efficiency also has a positive relationship (t=12.136) with ROE at 1% significant level with its coefficient of 0.346. This result indicates in any increases of efficiency, the ROE would also increase.
The result indicates TD is significant with ROE at 5% level. The coefficient is -0.0079 with a negative relationship (t=-2.182). It shows that any decrease in total debt would make the firm perform better. This is possible because a firm with less debt would be able to perform and grow better. This result is supported by Abor (2005) and Wang (2010).
Model Summary
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.477a
.228
.224
24.8873224
Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
Dependent Variable: RETURN ON EQUITY
ANOVA
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
191819.935
6
31969.989
52.469
.000b
Residual
604432.808
992
609.307
Total
796252.743
998
a. Dependent Variable: RETURN ON EQUITY
b. Predictors: (Constant), LONG TERM DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SHORT TERM DEBT, Ln TOTAL ASSET
4.3.3 Analysis outcome of TQ
Data analysis on table 3.0 which used Tobin's Q as a proxy for market based performance shows that the model is fit appropriateness with the F-value of 26.709 is significant at any level of significant and also R-square 11.8%. The R-square indicated the model explains 11.8% of the company performance. In this model only two out of five variables that significant with TQ.
For SIZE, a result show positive relationship (t=2.332) with significant at 5% level. This represent the greater the firm's size, the performance of firm would increase. Other than that, EFF also indicate positive relationship (t=10.939) with coefficient 0.333 at 1% significant level. It means one unit increase of efficiency would increase one unit of TQ. This shows the more efficient a firm manages its resources, the better the firm would perform.
Model Summary
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.344a
.118
.114
.8468007
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
b. Dependent Variable: TOBIN'S Q
ANOVA
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
95.760
5
19.152
26.709
.000b
Residual
712.769
994
.717
Total
808.529
999
a. Dependent Variable: TOBIN'S Q
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
Table 3.0: Regression for Relationship between ROA, ROE, TQ with Total Debt and control variables on 200 companies
ROA
ROE
TQ
Variables
Coefficients
T
sig.
Coefficients
t
sig.
Coefficients
T
sig.
SG
.014
.437
.662
.040
1.414
.158
-.039
-1.296
.195
SIZE
-.062
-1.512
.131
.328
9.075
.000
.090
2.332
.020
AG
.109
3.348
.001
.148
5.089
.000
.024
.786
.432
EFF
.068
2.113
.035
.346
12.136
.000
.333
10.939
.000
TD
.020
.486
.627
-.079
-2.182
.029
.067
1.725
.085
R-square
.020
.228
.118
Adjusted R Square
.015
.224
.114
F-Statistics
8.529**
52.469**
26.709**
** Significant at the 0.01 level
* Significant at the 0.05 level
Table 3.1: Summary of relationship between independent variables and Firm Performance on 200 companies
Independent variables
Hypothesis
Conclusion for ROA
Conclusion for ROE
Conclusion for TQ
Sales Growth
H0: There is no relationship between Sales Growth with ROA, ROE and TQ
Accept H0
Accept H0
Accept H0
Size
H0: There is no relationship between Size with ROA, ROE and TQ
Accept H0
Reject H0
Reject H0
Asset Growth
H0: There is no relationship between Asset Growth with ROA, ROE and TQ
Reject H0
Reject H0
Accept H0
Efficiency
H0: There is no relationship between Efficiency with ROA, ROE and TQ
Reject H0
Reject H0
Reject H0
Total Debt
H0: There is no relationship between Total Debt with ROA, ROE and TQ
Accept H0
Reject H0
Accept H0
4.4 Analysis of Multiple Regressions by sectors
4.4.1 Analysis outcome for properties industry
4.4.1.1 Analysis outcome of ROA
From table 4.0, the R Square and adjusted R square are 0.164 and 0.149 respectively. This means the model explains 16.4% in term on ROA and 14.9% of the movement in ROA is due to the changeability of the independent variables, namely, sales growth, size, asset growth, efficiency and total debt. The correlation of coefficient or R is 40.6%, it means a positive linear relationship between the dependent and independent variables.
Table 4.0 shows that the independent variables such as sales growth, size and total debt did not significant in this study. However, asset growth and efficiency did have significant relationship with ROA. Both have positive relationship (t=5.838 and t=2.838 respectively) with ROA. These indicate the higher value of ROA will increase the AG and EFF. The result is consistent with the result of total 200 companies.
Model Summary
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.405a
.164
.149
5.254
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET
GROWTH, LN TOTAL ASSET
b. Dependent Variable: RETURN ON ASSET
ANOVA
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
1513.633
5
302.727
10.965
.000b
Residual
7702.543
279
27.608
Total
9216.175
284
a. Dependent Variable: RETURN ON ASSET
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET
GROWTH, LN TOTAL ASSET
4.4.1.2 Analysis outcome of ROE
Table 4.0 shows the Adjusted R Square is at 0.208. This indicate the 20.8% of the movement in ROE is expressed by the changeability of all the independent variables which are sales growth, firm's size, asset growth, efficiency and total debt. The correlation of coefficient or R is 45.6%, indicates a positive linear relationship between dependent and independent variables.
From the table, it shows three out of five of the independent variables are significant. However, the other two do not show significant relationship with ROE. Asset growth, efficiency and total debt have a positive relationship (t= 5.503, 4,242 and 1.721 respectively) with ROE. Thus, it indicates the increases in asset growth, efficiency as well as total debt will increase the firm performance. The result (total debt and ROE) is consistent with Abu-Rub (2012). However, it does not consistent with the result of total 200 companies in term of its total debt and size of the company. This means in properties industry, the increases in uses on total debt will increase the company performance. Other than that, the firm's performance did not affected by the firm size of the properties industry.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.456a
.208
.194
9.283
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET
GROWTH, LN TOTAL ASSET
b. Dependent Variable: RETURN ON EQUITY
ANOVAa
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
6308.386
5
1261.677
14.640
.000b
Residual
24044.225
279
86.180
Total
30352.611
284
a. Dependent Variable: RETURN ON EQUITY
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, LN TOTAL ASSET
4.4.1.3 Analysis outcome of TQ
Table 4 indicates that the Adjusted R Square and R square are 0.03 and o.047 respectively. These means that 3% of the movement in Tobin's Q is expressed by the changeability of the independent variables while the R value of 21.6% shows a weak positive linear relationship between the dependent and independent variables.
In this model, the only variable that shows significant with Tobin's Q is total debt. Total debt shows a positive relationship (t=2.000) with TQ. This means the increase of using debt to finance the operations will increase the firm's performance of properties industry. This result is consistent with Abu-Rub (2012), Saeedi and Mahmoodi (2011). However, it is inconsistent with the overall 200 companies' result.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.216a
.047
.030
.316
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET
GROWTH, LN TOTAL ASSET
b. Dependent Variable: TOBIN'S Q
ANOVAa
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
1.370
5
.274
2.743
.019b
Residual
27.879
279
.100
Total
29.249
284
a. Dependent Variable: TOBIN'S Q
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, LN TOTAL ASSET
Table 4.0: Regression for Relationship between ROA, ROE, TQ with Total Debt and control variables on properties industry.
ROA
ROE
TQ
Variables
Coefficients
t
sig.
Coefficients
t
sig.
Coefficients
t
sig.
SG
.021
.367
.714
.030
.542
.588
.029
.482
.630
SIZE
.016
.206
.837
.003
.041
.968
.015
.172
.863
AG
.346
5.838
.000
.317
5.503
.000
.033
.519
.604
EFF
.159
2.838
.005
.231
4.242
.000
.112
1.879
.061
TD
.010
.122
.903
.134
1.721
.086
.171
2.000
.046
R-square
.164
.208
.047
Adjusted R Square
.149
.194
.030
F-Statistics
10.965**
14.640**
2.743*
** Significant at the 0.01 level
* Significant at the 0.05 level
Table 4.1 Summary of relationship between independent variables and Firm Performance on Properties Industry
Independent variables
Hypothesis
Conclusion for ROA
Conclusion for ROE
Conclusion for TQ
Sales Growth
H0: There is no relationship between Sales Growth with ROA, ROE and TQ
Accept H0
Accept H0
Accept H0
Size
H0: There is no relationship between Size with ROA, ROE and TQ
Accept H0
Accept H0
Accept H0
Asset Growth
H0: There is no relationship between Asset Growth with ROA, ROE and TQ
Reject H0
Reject H0
Accept H0
Efficiency
H0: There is no relationship between Efficiency with ROA, ROE and TQ
Reject H0
Reject H0
Accept H0
Total Debt
H0: There is no relationship between Total Debt with ROA, ROE and TQ
Accept H0
Accept H0
Reject H0
4.4.2 Analysis outcome for construction industry
4.4.2.1 Analysis outcome of ROA
From table 5.0, 10.2% of the movement in ROA is expressed by the changeability of the independent variables. The model is explains by 12.4% in term of ROA. The correlation of coefficient or R value is 35.2%. This indicates the relationship between dependent and independent variables in this model is weak positive correlated.
Table 5.0 shows that only two of the independent variables are significant. Those variables are size and total debt. Size is significant at 1% level and total debt is significant at 5% level. From the result, size has negative relationship (t=-4.790) with ROA. This indicates smaller size firm would have better performance than larger firm. However, total debt has positive relationship (t=2.403) with ROA. The higher the total debt of companies, the better the performance of the firm in construction industry. The result (ROA and TD) is consistent with Abu-Rub (2012), and Pratheepkanth (2011). Other variables such as sales growth, asset growth and efficiency are not significant to ROA. This result is inconsistent with the result of total 200 companies.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.352a
.124
.102
63.269
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH,
LN TOTAL ASSET
b. Dependent Variable: RETURN ON ASSET
ANOVAa
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
113789.456
5
22757.891
5.685
.000b
Residual
804590.641
201
4002.939
Total
918380.097
206
a. Dependent Variable: RETURN ON ASSET
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, LN TOTAL ASSET
4.4.2.2 Analysis outcome of ROE
From table 5.0, the Adjusted R Square and R Square are 0.336 and 0.352. These portrays that 33.6% of the movement in ROE is expressed by the changeability of the independent variables comprise of sales growth, size, asset growth, efficiency and total debt. The correlation of coefficient or R shows a relationship between dependent and independent variables obtain 59.3%. It means that a strong positive linear relationship present.
In this construction industry, the independent variables that are significant are size, efficiency and total debt. Size of the firm and Efficiency have positive relationship (t=7.525 and t=4.292 respectively) with ROE. However, total debt shows a negative relationship (t=-3.834) with ROE. This indicates a firm with large size and efficient in managing its resources with using less debt will have a better performance. This result (ROE and TD) is consistent with Onaolaopo and Keloja (2010), and Su and Vo (2010) but it is inconsistent with the 200 companies.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.593a
.352
.336
33.750
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, LN TOTAL ASSET
b. Dependent Variable: RETURN ON EQUITY
ANOVAa
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
123868.859
5
24773.772
21.750
.000b
Residual
227808.558
200
1139.043
Total
351677.417
205
a. Dependent Variable: RETURN ON EQUITY
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER(EFFICIENCY), ASSET GROWTH, LN TOTAL ASSET
4.4.2.3 Analysis outcome of TQ
From table 5.0, it was shown that 13.8% of the movement in Tobin's Q is expressed by the changeability of the sales growth, firm size, asset growth, efficiency and total debt. The value of correlation of coefficient is 39.8%, which indicate a positive linear relationship between Tobin's Q and independent variables.
In the construction industry, only the size and total debt are significant. The size does has negative relationship (t=-5.083) with the firm performance. However, the total debt indicates a positive relationship (t=2.900) with Tobin's Q which is consistent with Saeedi and Mahmoodi (2011) findings. This means in Construction Company, firm with small size and use more debt will have a better performance. This result is inconsistent with the 200 companies result.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.398a
.159
.138
.963
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET
GROWTH, LN TOTAL ASSET
b. Dependent Variable: TOBIN'S Q
ANOVAa
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
35.182
5
7.036
7.581
.000b
Residual
186.557
201
.928
Total
221.739
206
a. Dependent Variable: TOBIN'S Q
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, LN TOTAL ASSET
Table 5.0: Regression for Relationship between ROA, ROE, TQ with Total Debt and control variables on construction industry.
ROA
ROE
TQ
Variables
Coefficients
t
sig.
Coefficients
t
sig.
Coefficients
t
sig.
SG
.114
1.494
.137
.056
.857
.392
-.134
-1.798
.074
SIZE
-.454
-4.790
.000
.615
7.525
.000
-.472
-5.083
.000
AG
.106
1.429
.155
.087
1.345
.180
.138
1.887
.061
EFF
-.112
-1.567
.119
.265
4.292
.000
-.084
-1.193
.234
TD
.229
2.403
.017
-.316
-3.834
.000
.271
2.900
.004
R-square
.124
.352
.159
Adjusted R Square
.102
.336
.138
F-Statistics
5.685**
21.750**
7.581**
** Significant at the 0.01 level
* Significant at the 0.05 level
Table 5.1: Summary of relationship between independent variables and Firm Performance on Construction Industry
Independent variables
Hypothesis
Conclusion for ROA
Conclusion for ROE
Conclusion for TQ
Sales Growth
H0: There is no relationship between Sales Growth with ROA, ROE and TQ
Accept H0
Accept H0
Accept H0
Size
H0: There is no relationship between Size with ROA, ROE and TQ
Reject H0
Reject H0
Reject H0
Asset Growth
H0: There is no relationship between Asset Growth with ROA, ROE and TQ
Accept H0
Accept H0
Accept H0
Efficiency
H0: There is no relationship between Efficiency with ROA, ROE and TQ
Accept H0
Reject H0
Accept H0
Total Debt
H0: There is no relationship between Total Debt with ROA, ROE and TQ
Reject H0
Reject H0
Reject H0
4.4.3 Analysis outcome for industrial product industry
4.4.3.1 Analysis outcome of ROA
Table 6.0 indicates that 20.5% of the movement in ROA is due to the changeability of the sales growth, size, asset growth, efficiency and total debt while the models is explained 22.1% in term on ROA. In this model, the R value is 47.1%, which indicate a linear positive relationship present between the dependent and independent variables.
In this industry, sales growth, size, asset growth, and total debt are significant at 1% level. From the result, sales growth, size and asset growth have positive relationship (t=3.325, 4.361 and 3.675 respectively) with ROA. It means any one unit increase in ROA, also will increase one unit in sales growth, size and asset growth. The result (SIZE and ROA) is consistent with Majumdar and Chibber (1999) and Hammes and Chen (2004). However, total debt indicates a negative relationship (t=-3.203) with ROA. Thus, it means in the industrial product industry, a firm shall have lower debt to improve their performance. Zuraidah et al (2012), Onaolapo and Keloja (2010) and Gleason et al (2000) also suggest debt is negatively related with ROA. The result is inconsistent with the 200 companies result.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.471a
.221
.205
12.126
a. Predictors: (Constant), TOTAL DEBT, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SALES GROWTH, Ln TOTAL ASSET
b. Dependent Variable: RETURN ON ASSET
ANOVAa
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
10200.346
5
2040.069
13.874
.000b
Residual
35877.690
244
147.040
Total
46078.036
249
a. Dependent Variable: RETURN ON ASSET
b. Predictors: (Constant), TOTAL DEBT, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SALES GROWTH, Ln TOTAL ASSET
4.4.3.2 Analysis outcome of ROE
Table 6.0 shows the Adjusted R Square and R Square are 0.224 and 0.228 respectively. These means 22.4% of the movement in ROE is explained by the changeability in the independent variables. From the R values (47.9%), there is a positive linear relationship exist between dependent and independent variables.
From the result, three out of five of the independent variables are significant at 1% significant level. The variables are size, asset growth and efficiency. These three variables have positive relationship (t=3.724, 3.773, and 3.941 respectively) with ROE. This shown to increase the shareholders' return, a firm must increase the firm's size, asset and efficiency.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.479a
.230
.214
19.626
a. Predictors: (Constant), TOTAL DEBT, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SALES GROWTH, Ln TOTAL ASSET
b. Dependent Variable: RETURN ON EQUITY
ANOVAa
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
28042.418
5
5608.484
14.561
.000b
Residual
93980.498
244
385.166
Total
122022.916
249
a. Dependent Variable: RETURN ON EQUITY
b. Predictors: (Constant), TOTAL DEBT, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SALES GROWTH, Ln TOTAL ASSET
4.4.3.3 Analysis outcome of TQ
From table 6.0, the value of Adjusted R Square and R Square are 0.052 and 0.071 respectively. It indicates 5.2% of the movement in the Tobin's Q is expressed by the changeability of the independent variables such as sales growth, size, asset growth, efficiency and total debt. The correlation of coefficient or R shows a relationship between dependent and independent variables obtains a value of 26.7%. It means that there is a weak positive relationship exists.
Table 6.0, shows that independent variables such as sales growth, asset growth and total debt did not significant in this model. However, size and efficiency did have a significant relationship with TQ. Size has a positive relationship (t=2.168) with TQ, which means that the larger the firm size, the higher the value of TQ. Conversely, efficiency has a negative relationship (t=-2.291) with TQ. It shows that a firm with lower efficient will have higher TQ. This result is inconsistent with the total 200 companies result.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.267a
.071
.052
.474
a. Predictors: (Constant), TOTAL DEBT, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SALES GROWTH, Ln TOTAL ASSET
b. Dependent Variable: TOBIN'S Q
ANOVAa
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
4.187
5
.837
3.735
.003b
Residual
54.713
244
.224
Total
58.900
249
a. Dependent Variable: TOBIN'S Q
b. Predictors: (Constant), TOTAL DEBT, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, SALES GROWTH, Ln TOTAL ASSET
Table 6.0: Regression for Relationship between ROA, ROE, TQ with Total Debt and control variables on industrial product industry.
ROA
ROE
TQ
Variables
Coefficients
t
sig.
Coefficients
t
sig.
Coefficients
t
sig.
SG
.200
3.325
.001
.117
1.963
.051
.122
1.859
.064
SIZE
.325
4.361
.000
.276
3.724
.000
.177
2.168
.031
AG
.220
3.675
.000
.225
3.773
.000
.111
1.696
.091
EFF
.068
1.174
.242
.226
3.941
.000
-.145
-2.291
.023
TD
-.235
-3.202
.002
-.122
-1.677
.095
-.101
-1.257
.210
R-square
.221
.230
.071
Adjusted R Square
.205
.214
.052
F-Statistics
13.874**
14.561**
3.735**
** Significant at the 0.01 level
* Significant at the 0.05 level
Table 6.1: Summary of relationship between independent variables and Firm Performance on Industrial Product Industry
Independent variables
Hypothesis
Conclusion for ROA
Conclusion for ROE
Conclusion for TQ
Sales Growth
H0: There is no relationship between Sales Growth with ROA, ROE and TQ
Reject H0
Accept H0
Accept H0
Size
H0: There is no relationship between Size with ROA, ROE and TQ
Reject H0
Reject H0
Reject H0
Asset Growth
H0: There is no relationship between Asset Growth with ROA, ROE and TQ
Reject H0
Reject H0
Accept H0
Efficiency
H0: There is no relationship between Efficiency with ROA, ROE and TQ
Accept H0
Reject H0
Reject H0
Total Debt
H0: There is no relationship between Total Debt with ROA, ROE and TQ
Reject H0
Accept H0
Accept H0
4.4.4 Analysis outcome for consumer product industry
4.4.4.1 Analysis outcome of ROA
From table 7.0, the Adjusted R Square and R Square value are 0.356 and 0.369 respectively. These indicate 35.6% of the movement in ROA is explained by the changeability of independent variables. The correlation of coefficient or R value in this model is 60.7%. This means there exist a strong positive linear relationship between the dependent and independent variables.
In this industry, three of the independent variables are significant at 1% level. Size, asset growth and efficiency have positive relationship (t=3.501, 3.600, and 8.246 respectively) with ROA. This indicates larger firms with high asset growth and efficiency will have better return on investment in the consumer product industry. This result is inconsistent with the total 200 companies.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.607a
.369
.356
9.577
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
b. Dependent Variable: RETURN ON ASSET
ANOVAa
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
13061.721
5
2612.344
28.480
.000b
Residual
22380.743
244
91.724
Total
35442.464
249
a. Dependent Variable: RETURN ON ASSET
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
4.4.4.2 Analysis outcome of ROE
Table 7.0 shows the Adjusted R Square value is 0.356. This indicates 35.6% of the movement in ROE is expressed by the changeability of sales growth, size, asset growth, efficiency and total debt. From the R value (60.7%), shows a strong positive linear relationship exist between ROE and independent variables.
From table 7.0, it shows asset growth, efficiency and total debt have significant relationship with ROE. These three variables having a positive relationship (t=3.145, 7.556 and 4.067 respectively) with ROE. This shows one unit increase in ROE will also increase the value of asset growth, efficiency and total debt. The result is inconsistent with the 200 companies in term of firm size.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.607a
.369
.356
26.652
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
b. Dependent Variable: RETURN ON EQUITY
ANOVAa
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
101187.561
5
20237.512
28.491
.000b
Residual
173318.539
244
710.322
Total
274506.100
249
a. Dependent Variable: RETURN ON EQUITY
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
4.4.4.3 Analysis outcome of TQ
Table 7.0 indicates the Adjusted R Square and R Square value are 0.367 and 0.380 respectively. These indicate 36.7% of the movement in Tobin's Q is expressed by the changeability of the independent variables. The R value (61.6%) indicates a strong positive linear relationship present between the dependent and independent variable.
In this consumer product industry, there are only two variables that are significant, which are efficiency and total debt. Both are significant at 1% level and having positive relationship (t=8.820 and 5.477 respectively) with TQ. This means the higher the value of efficiency can total debt, the higher the value of TQ. It is supported by the Saeedi and Mahmoodi (2011) findings. However, the result is inconsistent with the total 200 companies.
Model Summaryb
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.616a
.380
.367
1.051
a. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
b. Dependent Variable: TOBIN'S Q
ANOVAa
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
164.894
5
32.979
29.860
.000b
Residual
269.490
244
1.104
Total
434.384
249
a. Dependent Variable: TOBIN'S Q
b. Predictors: (Constant), TOTAL DEBT, SALES GROWTH, ASSET TURNOVER (EFFICIENCY), ASSET GROWTH, Ln TOTAL ASSET
Table 7.0: Regression for Relationship between ROA, ROE, TQ with Total Debt and control variables on consumer product industry.
ROA
ROE
TQ
Variables
Coefficients
t
sig.
Coefficients
t
sig.
Coefficients
t
sig.
SG
.078
1.502
.134
.014
.266
.790
-.043
-.828
.409
SIZE
.249
3.501
.001
.111
1.560
.120
-.007
-.103
.918
AG
.188
3.600
.000
.165
3.145
.002
.008
.152
.879
EFF
.436
8.246
.000
.399
7.556
.000
.462
8.820
.000
TD
.005
.077
.938
.281
4.067
.000
.375
5.477
.000
R-square
.369
.369
.380
Adjusted R Square
.356
.356
.367
F-Statistics
28.480**
28.491**
29.860**
** Significant at the 0.01 level
* Significant at the 0.05 level
Table 7.1: Summary of relationship between independent variables and Firm Performance on Consumer Product Industry
Independent variables
Hypothesis
Conclusion for ROA
Conclusion for ROE
Conclusion for TQ
Sales Growth
H0: There is no relationship between Sales Growth with ROA, ROE and TQ
Accept H0
Accept H0
Accept H0
Size
H0: There is no relationship between Size with ROA, ROE and TQ
Reject H0
Accept H0
Accept H0
Asset Growth
H0: There is no relationship between Asset Growth with ROA, ROE and TQ
Reject H0
Reject H0
Accept H0
Efficiency
H0: There is no relationship between Efficiency with ROA, ROE and TQ
Reject H0
Reject H0
Reject H0
Total Debt
H0: There is no relationship between Total Debt with ROA, ROE and TQ
Accept H0
Reject H0
Reject H0
CHAPTER 5
CONCLUSION AND RECOMMENDATIONS
5.0 Introduction
This paper studies on the relationship of the capital structure and companies performance by observing 200 listed companies in Malaysia and also divided it into four different sectors, namely, properties industry, construction industry, industrial product industry and consumer product industry. This chapter summarizes the findings of the study, outlines research limitation, and proposes recommendation for future research.
5.1 Summary of Findings
This paper examines on 57 properties companies, 43 construction companies, 50 industrial product companies and 50 consumer product companies over 5 years periods, from 2007 to 2011. These totals of 200 companies are listed in Bursa Malaysia during the selected analysis period. ROA and ROE had been used as the accounting measure while Tobin's Q used as the market measure. In this study, five independent variables had been examined, namely, Sales Growth, Firm Size, Asset Growth, Efficiency and Total Debt.
This study found that efficiency is significant to all the firm performance measures for the total 200 companies. It means the firm listed in bursa Malaysia must have higher total asset turnover to perform better. Asset growth is significant to ROA and ROE. This is contradicting with ROA, where it does not show significant relationship with ROA. SIZE is significant to ROE and Tobin's Q. Sales growth does not show significant relationship either with accounting measure nor market measure. Total debt indicates a negative relationship only with ROE, which shown that to increase the return of the investors, the firms must use less debt. This variable is not significant with ROA as well as Tobin's Q.
From the result, the properties industry shows a consistent result with the total 200 companies in term of ROA. The variables that are significant with ROE are same as the variables that are significant with ROA in properties companies. Total debt only significant with market measure but not accounting measure. It has a positive relationship with market measure.
In the construction industry, SIZE and Total Debt are significant with accounting measure and market measure. SIZE has negative relationship with ROA and Tobin's Q whiles the opposite with ROE. However, Total Debt shows contradict relations with SIZE. It has positive relationship with ROA and Tobin's Q, and negative relationship with ROE. The Efficiency is only significant with ROE but not the other variables.
For industrial products industry, SIZE is significant to accounting as well as market measures while the Asset Growth only significant with accounting measures. Next, the efficiency in this industry is not significant with ROA, but is significant with ROE and Tobin's Q. The efficiency has different relationship with the ROE and Tobin's Q. It is positively related with ROE whilst negatively related with Tobin's Q. In this industry, the Total Debt is only significant with ROA and with a negative relationship. The Sales Growth has a significant relationship with ROA.
In the consumer product industry, it show a contradict result with industrial product industry in term of its Total Debt. The Total Debt is significant with ROE and Tobin's Q but not ROA. The efficiency is significant with the accounting as well as market measures. However, the Asset Growth is only significant with accounting measures. In this industry, the SIZE is only significant with ROA but not the ROE and Tobin's Q.
In summary, the industrial product industry seems perform better in term of accounting and market measure compared to the other three industries. It can be proved in the table 4, 5, 6 and 7 where more variables in industrial product companies are significant to the variables that being measured compared to others. From the comparison of table 4, 5, 6 and 7, the Total Debt is significant with accounting and market measure only in construction industry. This clearly shows that the increase of use in capital structure in construction industry will improve the performance of the companies.
5.2 Implication of the study
I hope this study could contribute to the other researchers and give indication to the managers regarding the use of capital structure to improve their companies' performance. From the 1000 observation, we know that for a company to increase their performance they must use the least debt to finance their operation. This might be true because company need to pay for the interest payment of the debt, and reduce the earnings that available for the shareholders. The other reasons may be because the marginal tax shelter benefits less than the marginal bankruptcy-related costs. Therefore, reduce the performance of the firm because the increase in debt after the optimal point of capital structure will reduce the value of stock.
Further, this study also found a positive relationship of Total Debt with the ROE in the consumer product industry. This finding is utterly contradicted from the total observation. It indicates for the company operates in consumer product industry, they must increase the use of total debt to improve their performance (ROE). This might be true because the increases in debt at this point have yet exceed the optimal capital structure. Thus, the level of debt can bring the improvement of the companies.
This study does reveal a positive and a negative relationship between the capital structure and the company performance. A positive relationship shows the level of debt a company employed can improve the performance of the company whilst a negative relationship shows the opposite. Hence, a manager can identify the level of debt that can improve the performance of the company or reduce the company performance.
5.3 Limitation of the study
The limitation of this study is the total sample size is 200 companies of public listed companies in Malaysia. So, the result of the sample size may not generalize the whole population of the companies in Malaysia. Other than that, the industry that I have examined in this study is only four, which are, the 57 properties companies, 43 construction companies, 50 industrial companies and 50 consumer product companies. With these industries, it is difficult to determine the exact result all the industries in Malaysia.
Next, in determine the capital structure; the only proxy that used is Total Debt. This is the limitation because with the only use of Total Debt. With total debt alone is not enough to show the significant of capital structure in determine the company performance. Beside, this study is just depends on the historical data and could not accurately predict for future performance. This reason is because the uncertainty of future economic condition, such as war and natural disaster, that will bring to downturn of economic, which will effect to the business operation.
Other than that, time and cost constraint will also be a limitation for this study. Due to time limitation, it is unable to learn and have deeper understanding for the topic of capital structure. This constraint may affect to the accuracy and precision of the findings. The cost constraint has reduced the size of sample data, and this may influence the result for generalization.
5.4 Recommendations for future research
The recommendation for this study is to increase the total sample size in order to have better generalization for the findings. The future researchers also is recommended to have a deeper understanding of capital structure by doing more reading on this particular topic before start the methodology chapter. This is because when the methodology is incorrect or inaccurate, then the findings also will be influence and could not use for forecast future performance.
Researchers are suggested to include more proxies for the capital structure and company performance. The proxy of capital structure that can be used are Long-term debt to capital, Debt to capital, Debt to Asset, Debt to equity Market Value, Debt to Common Equity and Long-term Debt to Common Equity. For company performance, researchers can use are Return on Capital, Earnings per Share, Operating Margin and Net Margin. This is to increase the reliability and generalizability of the findings.
Researchers are advice to pay high effort in term of time and cost. Last but not least, the researchers shall ensure themselves are motivated and interested to the particular topic so that they would have high determination to complete the study. The continuously effort and allocation of time for the study is important to ensure no last minutes and non-quality study for the contribution purposes in this particular topic.
5.5 Conclusion
This chapter discusses the overall conclusion based on the findings and analysis. In term of capital structure, the construction companies seem to be performed better than other industries. However, when the control variables and capital structure are included as the independent variables, industrial product companies have better performance than the other three industries. Therefore, in determine the capital structure that are beneficial for a company, managers can use construction companies as a guide to increase the performance.
