With this theoretical background, if now the two profitability models be analysed, it is very clear that both the models follow the same behaviour. The variables that are statistically significant are same; the sign of the coefficients of the variables are also same. Hence, essentially it points out that both the measures of profitability exhibit same behaviour. Note that the two models cannot be compared because their error terms follow different distributions. However, the behaviour of the variables can be compared. The analysis of regression is widely used to point out the causality and the directionality of the independent variables as against using the exact estimates.
In that capacity, the comparison of directionality and causality is reasonable, but the comparison of magnitudes is not reasonable. From the above two models, it is evident that the concentration of the banks in the market plays a negative but statistically insignificant role in the determination of the profitability of the banks. Interestingly, in both the models, the coefficient of the variable is negative. Hence, it can be concluded from the model that the effect of market concentration on the banks is neither positive nor statistically significant. This effect has been observed in several of the recent researches conducted with the banks as the firms.
The model clearly points out that the variable of market share is significant. It affects both the ROE and ROA positively. This clearly implies that market share significantly affects the profitability of the bank. The market concentration plays no role in the determination of profitability, but the market share has a positive bearing it. This is in line with the findings of Smirlock (1985) who posited that when the variables account for the market share, the concentration of the firms in the market brings no new information in the model. Nevertheless, in the absence of concentration, market share bears a positive effect on the profitability of the firm.