As evident from the tables above, the model is significant at 1% level of significance. The p-value for the F-statistic in the above model is very low which implies that the model is significant. The explanatory power of the model is given by the R-squared value. Thus, in this model the variation in the dependent variable is explained to the extent of 40.3%. The model appears to be free from multicollinearity despite high correlations because 50% the variables are statistically significant and the explanatory power of the model is low.The variable MS corresponding to market share has a coefficient of 0.026. The corresponding p-value of the estimate is 0.044. This implies that the corresponding null hypothesis that the coefficient is statistically not different from zero can be rejected at 5% level of significance.
However, the coefficient of CR3 i.e. the market concentration variable is not statistically significant. Even though the coefficient is negative, it is not statically different from zero which means that it’s effect on the dependent variable ROA is not statistically significant because the associated p-value for the t-test hypothesis is 0.671 which is higher than even 0.1. So, even at 10% level of significance, the null hypothesis cannot be rejected. The coefficients of both GDP and total deposits in the country are very small but are statistically significant. This implies that even though they are small, they are statistically different from zero.
The p-value for the coefficient of GDP is 0.017. This implies that the null hypothesis can be rejected with more than 98% surety. Similarly the p-value for total deposits is nearly 0.000 which means the null hypothesis that the coefficient is statistically equal to zero can be rejected with almost 100% confidence. The exceedingly low value of the coefficients of GDP and total deposits arise because the dependent variable is very low in value while the variable in itself is very high in magnitude. To compensate for that, the coefficients are small in magnitude. The variable CPI corresponding to the inflation of the economy however is not statistically significant as the p-value (0.912) is greater than 0.1. So, the causality between CPI and ROA does not has any practical implication.