13 Table 3 Number of smoother and non smoother banks CV 05 05 CD 1 CV 05 05 CD

# 13 table 3 number of smoother and non smoother banks

This preview shows page 13 - 16 out of 18 pages.

13 Table 3: Number of smoother and non-smoother banks CV <= 0.5 0.5 <=CD <=1 CV <= 0.5 & 0.5 <=CD <=1 Smoother Banks 29 (44 %) 49 (75 %) 26 (40 %) Non smoother Banks 37 17 40 TOTAL 66 66 66 Once income smoothing practices are highlighted by Islamic banks, then we investigate if they use loan loss provisions to smooth their results. Table 4 contains the main results of our econometric investigation. It reports four regressions using: - The full sample - The panel that includes only observations for which the determination coefficient is between 0.5 and 1. - The panel that includes only observations for which the variation coefficient is less than 0.5. - The panel that includes only observations for which the determination coefficient is between 0.5 and 1 and the variation coefficient is less than 0.5. For these four regressions, we test two specifications: the first one includes all the explicative variables defined in our model. In the second specification, we replace the GDP variable by its lagged (Salas and Saurina, 2002)
14 Table 4: Results of the regression model Whole sample Sub-Sample 0.5 <=CD <=1 Sub-Sample CV <= 0.5 Sub-Sample CV <= 0.5 & 0.5 <=CD <=1 (1) (2) (1) (2) (1) (2) (1) (2) LLP LLP LLP LLP LLP LLP LLP LLP NPL 0.144**0.138**0.144***0.139**0.0257 0.0558**0.0257 0.0558**(2.26) (1.82) (2.27) (1.82) (0.64) (1.76) (0.64) (1.76) TL 0.00873**0.00528*0.00707 0.00149 0.00645 0.00315 0.00645 0.00315 (1.89) (1.28) (1.42) (0.26) (0.88) (0.47) (0.88) (0.47) EBTP - 0.0168 - 0.00626 -0.0148 -0.00148 - 0.0171 0.00112 - 0.0171 0.00112 (-0.81) (-0.35) (-0.69) (-0.08) (-0.54) (0.02) (-0.54) (0.02) CAR -0.00012***-0.00010**-0.00012***-0.000110***-0.000087**-0.00009***-0.000087**-0.00009***(-3.50) (-3.03) (-3.35) (-2.87) (-2.30) (-2.99) (-2.30) (-2.99) SIZE 0.00188***0.00175***0.00195*0.00188**0.00118**0.00119**0.00118**0.00119**(2.44) (2.37) (2.34) (2.24) (2.02) (2.22) (2.02) (2.22) CONS -0.00373 -0.00444 -0.00383 -0.00468 0.00264 0.000903 0.00264 0.000903 (-1.25) (-1.48) (-1.24) (-1.48) (1.21) (0.56) (1.21) (0.56) GDP 0.000398 0.000391 -0.000313 -0.000313 (1.26) (1.24) (-1.51) (-1.51) LGDP -0.0000551 -0.000103 -0.000191 -0.000191 (-0.20) (-0.39) (-0.94) (-0.94) cons -0.0296**-0.0231*-0.0295*-0.0222*-0.0153 -0.0145 -0.0153 -0.0145 (-2.17) (-2.04) (-2.20) (-2.04) (-1.66) (-1.76) (-1.66) (-1.76) N109 95 101 88 40 36 40 36 R237.51% 34.60% 37.65% 35.08% 44.96% 43.02% 44.96% 43.02% Where LLP it: The value of net specific and general loan loss provisions of bank i in year t normalized by the total assets; NPL it: Non Performing Loans normalized by the total assets; TL it: ratio of total loans normalized by the total assets; EBTP it: Earnings before taxes and provisions normalized by the total assets; CAR it: Capital Adequacy Ratio approximated by total funds to total assets; SIZEit: Logarithm of Total Assets; GDPit: The rate of growth Gross Domestic Product in year t ; CONSit: Dummy variable takes 1 if data comes from accounting statements of consolidated banking groups and 0 otherwise. The regression results for the full sample are available only for 60 countries (Brunei, Egypt, Sudan, Iran, Mauritania and Tunisia were excluded because a lack of availability of data ) which reduced the number of observations to 104. *, **, and *** denote statistical significance at the 10%, 5%, and 1% level, respectively.
15 The results presented in the table 4 lead us some conclusions. i) The income smoothing hypothesis was not supported either in the whole sample or in the sub-samples. The benefits before tax and provisions scaled by total assets is a non significant factor in determining the amount of loss provision in all specifications. This result suggests that Islamic banks do not use loss provision to manage and smooth their results and confirms those found by Ismail et al.

#### You've reached the end of your free preview.

Want to read all 18 pages?

• Spring '20
• Amanullah
• Test, Capital requirement