**4. Analysis of the comparative impacts between conventional and Islamic banks**

### **4.1 Individual impacts analysis**

To bridge the problem of evaluation and comparison between the Islamic and conventional FP levels via the EAC effect, it is necessary to first solve the comparison ambiguity of individual effects on each FP measure to see if the EAC individual effects are the same or different among FP measures. Since EAC can have a positive impact or negatively affect the banks' FP, it all depends on the bank type and their financial situation. First, we checked each model's EAC effects significance. Second, we established a comparative study of the impacts resulting from similar models. Finally, we compared the predicted signs with the actual signs. In what follows, **Tables 15**-**22** in the appendices include the findings of equivalent regressions for two bank types.

Returning to the list of tables in the indexes, we noticed that for each bank type, the estimates gave different individual coefficients that varied from one model to another according to the sign and significance of each FP measure. This implies that the individual effects are heterogeneous. In addition, with regard to the similar separate impacts of CBs and IBs on the same FP measures, the same models also recorded variable impacts according to sign and significance. Therefore, we cannot rely on these preliminary results to make a definitive comparative decision. We must advance the decision rule even further by bringing the divergent signs together. The solution requires the grouping of the separate EAC impacts given by a similar modeling technique.

### **4.2 Combined impacts analysis**

To avoid the detailed EAC ambiguity effects on the FP of each bank type, we followed a more developed analysis process. We started by grouping the individual EAC impacts and each control variable on each FP measure. Then, we compared the

### *New Topics in Emerging Markets*

close effects of the independent variable in each IB model with the similar close effects of the independent variable in each CB model.
