Page 57 - SUSTAINABILITY ISSUES & COVID-19
P. 57

dependent variables used is solely the cost of debt (COD) to shed some light on the relationship between COD and ESG
               disclosure. Hence, the test is not extended to YTM and EIR.

               4.  RESULT AND ANALYSIS

               The econometrical test is performed before analyzing the regression result. This is necessary to determine whether the
               results of the regression estimation are normally distributed and are truly free from the presence of multicollinearity,
               heteroscedasticity  and  autocorrelation  problems.  From  the  results  of  the  multicollinearity  test  for  H1,  H2,  H3  the
               tolerance value obtained is above 0.1 and  VIF less than 10. Additionally, from the following Correlation Analysis, there
               is no Pearson’s correlation of more than 0.8 noted between all the variables. From both tests, it can be concluded that
               in this model there are no multicollinearity problems or there is no strong relationship or correlation between the
               independent variables. Heteroscedasticity test is shown to test whether in the regression model there is an inequality
               of variance from the residuals of one observation to another. This test was carried out with the Breusch-Pagan / Cook-
               Weisberg test for heteroscedasticity. The significance value obtained for H2 is more than 0.05 or 5%, which means the
               regression H2 in this study is free from heteroscedasticity problems. From the results of the heteroscedasticity test for
               H1 and H3, the Prob>chi2 obtained is less than 0.05 or 5%, which means the regression H1 and H3 in this study have
               heteroscedasticity problems. The heteroscedasticity problem is treated with robust treatment in STATA. Autocorrelation
               tests whether in the regression model there is a correlation between residuals in period t with residuals in period t-1
               or to see whether there is a relationship between errors of one observation with other observations. The regression
               analysis is conducted to prove the impact of independent variables towards dependent variables, thus it must not
               have any correlation between current observation and the previous one. Testing is done with the Wooldridge Test.
               From the Wooldridge test result, H1 and H3 show Prob>F less than 0.05, therefore there are autocorrelation problems.
               The autocorrelation problem is treated with robust treatment in STATA. H2 shows Prob>F more than 0.05, therefore no
               existence of autocorrelation. Table 3A, 3B, 3C further show the correlation matrixes that are consistent with previous
               research of a negative significant correlation between ESG disclosure index with the cost of debt, including from public
               bonds and banks.
                                      Table 3A  Pearson Correlation H1 (COD measurement)
                               COD        ESG        SIZE       LEV       ROE        CFO        GDP
                COD              1.0000
                ESG            -0.1306**    1.0000
                SIZE           -0.2597**  0.3067**     1.0000
                LEV             -0.0817*    0.059      0.32**    1.0000
                ROE              0.0124     0.0208   -0.2445**   0.178**    1.0000
                CFO             -0.0272    0.0909*   -0.2423**  -0.1229**  0.5811**    1.0000
                GDP            0.2118**   -0.1926**    0.0058    0.0150     -0.0277    -0.0339    1.0000
                **Correlation is significant at the 0.01 level (2-tailed)
                *Correlation is significant at the 0.05 level (2-tailed)

                                      Table 3B Pearson Correlation H2 (YTM measurement)
                               YTM        ESG        SIZE       LEV       ROE        CFO        GDP
                YTM              1.0000
                ESG            -0.2723**    1.0000
                SIZE           -0.5270**  0.3030**    1.0000
                LEV            -0.3028*    -0.428**  -0.0252**   1.0000
                ROE              0.0367     -0.136   -0.0845**   0.1341     1.0000
                CFO             -0.1068     0.0562   0.1442**  -0.2344**   0.1541**    1.0000
                GDP            0.3713**   -0.4359**  -0.2239**  0.2995**    0.0551     -0.0404    1.0000
                **Correlation is significant at the 0.01 level (2-tailed)
                *Correlation is significant at the 0.05 level (2-tailed)






         56     International Conference on Sustainability
                (5  Sustainability Practitioner Conference)
                 Th
   52   53   54   55   56   57   58   59   60   61   62