Factors Determine Islamic Banking Performance in Malaysia: a multiple Regression Approach

 

 

LLPTA NLTA TOTA  SER  BS  GDP  INF 

Pearson 

Correlation 

.326

*

 .312

*

 -.896

*

 -.645

*

 .178

.190  -.041 

Sig. (2-tailed) 

.004 

.006 

.000 .000 .126

.103 .724 

Return on 

Assets  

N 75 

75 

75 

75 

75 

75 

75 

Pearson 

Correlation 

-.921

*

 .066 -.058 -.238

*

 .121

-.162  -.020 

Sig. (2-tailed) 

.000 

.575 

.624 .040 .301

.166 .863 

Return on 

Equity  

N 75 

75 

75 

75 

75 

75 

75 

* Correlation is significant at the 0.01 level (2-tailed). 

 

 

Regression Analysis 

Table 3 summarizes ROA and ROE models for Islamic banks. R-square is the 

percentage of the variance from the dependent variable explained by the independent 

variables. In this case, the predictors explain the variances of ROA at 94.2% and ROE at 

93.2%. From these findings it is clear that there are significant relationship between 

dependent and independent variables for Islamic banks. The adjusted R square at 91.9% 

(ROA) and 90.5% (ROE) attempt to correct R square to more closely reflect the goodness 

of suitability of the model in the population. The Durbin-Watson statistics are 2.039 for 

ROA and 2.036 for ROE; these mean that there is no autocorrelation problem in the 

residuals from the regression analysis. 

---

52 

Journal of Islamic Banking and Finance Jan.- March. 2014 

 

Table 3. Model Summary 

Model R 

R 

Square 

Adjusted R 

Square 

Std. Error of the 

Estimate 

F Sig 

Durbin-

Watson 

ROA .971

a

 .942 

.919 

.99052 

40.902 

.000  2.039 

ROE .965

a

 .932 

.905 

18.07184  34.401 

.000  2.036 

From Table 4, it can be seen that variables LLPTA and TOTA are significant at 5% 

alpha for ROA while LLPTA and INF are significant at 5% alpha for ROE. Other 

variables are not significant in predicting both models. On the one hand, for ROA model, 

LLPTA positively influence bank performance and TOTA negatively influence bank 

performance. The former is against hypothesis 1 while the latter is supporting hypothesis 

2. On the other hand, for the ROE model, both LLPTA and inflation relationship towards 

banking performance are supporting the hypothesis 1 and 7 respectively. This shows that 

high credit risk and inflation will lower the probability of Table 4: Regression Analysis