in the total sample and each subsample of banks have a
significant effect on bank profitability only for a limited
number of bank characteristics. More specifically, Income
Diversity is significantly and negatively correlated with
EARTA in the total sample and the sample of IBs but
insignificant for the sample of CBs. Additionally, we
observe that EARTA is positively correlated with “Listed”
and “Crisis” dummy variables only in the sample of IBs.
Other significant variables are Loans/Assets (in the sample
of CBs) and Income Diversity (in the sample of IBs).
Table 8 presents the outcomes of the regression analysis
for EARGL as another measure of bank profitability. We
find that bank characteristics effect on EARTA is negative
and significant for Loans/Assets in the whole sample and the
sample of CBs but not for IBs; the only significant variable
in this group of banks is Other Earning Assets. As in all
previous models, ownership structure has no impact on bank
profitability on either type of banks while risk-taking (as
measured by distance to default, Log(Z)) is an important
determinant of IBs profitability. In conclusion, the effect of
bank characteristics on EARGL is strongly significant only
in the group of CBs, which is not in line with the results
reported in Table 7. Based on the results of both the mean ttest, we may conclude that there is no evidence of a significant difference in bank profitability (as measured by
EARTA and EARGL) between the two types of banking
systems. Hypothesis HA2 is thus rejected, and H02 is
accepted. However, the effect of bank characteristics on
profitability is more pronounced in the group of CBs than in
the group of IBs.
H02: There is no significant difference between the
conventional and Islamic banking systems in
profitability.
The effect of capital adequacy measures (control variable) on
cost efficiency (CIR) is presented in Table 5. The influence
of a bank’s capital adequacy on cost efficiency is significant
and negative for Tangible Equity and Liquid Assets (see
Model 3), and marginally significant for Tier 1 ratio. If we
consider CBs and IBs separately, we will find that only
Liquid Assets have a strongly negative effect on cost efficiency, which is more pronounced in the sample of CBs. We
may conclude that there is a significant effect of a bank’s
capital adequacy on cost efficiency for both types of banks,
but this effect is limited only to Liquid Assets (banks with
more liquid assets are more efficient). Hypothesis HA3 is
thus partially confirmed.
HA3: There is a significantly different effect of capital
adequacy on the banks’ cost efficiency between the
conventional and Islamic banking systems.
The capital adequacy effect on cost efficiency as measured
by NIM is presented in Table 6. The effect of a bank’s
capital adequacy on its cost efficiency is strongly significant
in the whole sample (all the estimated coefficients are statistically significant). We observe similar (significant) relationship in the sample of CBs except for Funding fragility.
However, this effect is relatively weak in the sample of IBs
where Tier 1 ratio and Liquid Assets are marginally significant while Funding fragility is statistically significant at the
5% level of significance. In conclusion, the effect of capital
adequacy on a bank’s cost efficiency is found to be strong
and significant in the sample of CBs but relatively weak for
IBs. The capital adequacy effect on bank profitability as
measured by EARTA is reported in Table 7. We observe
that the effect of a bank’s capital adequacy on bank profitability is insignificant in the whole sample as well as in the
sample of IBs. For the sample of CBs, we find that Liquid
Assets have a strongly negative influence on bank profitability; that is, banks with less Liquid Assets can achieve
higher profitability. Hypothesis HA4 is thus partially
confirmed.
HA4: There is a significantly different effect of capital
adequacy on the banks’ profitability between the
conventional and Islamic banking systems.
The alternative tests for capital adequacy effect on the bank
performance (using EARGL as another measure of bank
profitability) are reported in Table 8. The capital adequacy’s
influence on EARGL is significant yet negative only for
Liquid Assets (see Model 3 and 4 for the whole sample of
UAE banks). Again, we observe a significant difference in
the capital adequacy effect between CBs and IBs. While in
the group of CBs both Tier 1 ratio and Liquid Assets have a
strong influence on bank profitability, no significant effect is
observed in the group of IBs (except for Liquid Assets in
Model 10).
The effect of risk-taking on cost efficiency as measured
by CIR is presented in Table 5. We find that the influence of
bank risk on CIR is strongly significant both in the whole
sample of UAE banks and in the sample of CBs; however, it
is insignificant for IBs. Moreover, this effect is significant
only when bank-level characteristics and capital adequacy
measures are added to the model together with the risk
measures (Equity Volatility and Log(Z)). The results for Log
(Z) show that the banks with high distance to default (low
risk) are more cost-efficient. However, the evidence indicates
that risk-taking does not influence the cost efficiency of IBs
in the UAE.
Similarly, the risk effect on cost efficiency when measured by NIM is presented in Table 6. The bank risk influence on NIM is significant and negative for the whole
78
F. Mrad and M. Mateev
significant effect on bank profitability only for a limited
number of bank characteristics. More specifically, Income
Diversity is significantly and negatively correlated with
EARTA in the total sample and the sample of IBs but
insignificant for the sample of CBs. Additionally, we
observe that EARTA is positively correlated with “Listed”
and “Crisis” dummy variables only in the sample of IBs.
Other significant variables are Loans/Assets (in the sample
of CBs) and Income Diversity (in the sample of IBs).
Table 8 presents the outcomes of the regression analysis
for EARGL as another measure of bank profitability. We
find that bank characteristics effect on EARTA is negative
and significant for Loans/Assets in the whole sample and the
sample of CBs but not for IBs; the only significant variable
in this group of banks is Other Earning Assets. As in all
previous models, ownership structure has no impact on bank
profitability on either type of banks while risk-taking (as
measured by distance to default, Log(Z)) is an important
determinant of IBs profitability. In conclusion, the effect of
bank characteristics on EARGL is strongly significant only
in the group of CBs, which is not in line with the results
reported in Table 7. Based on the results of both the mean ttest, we may conclude that there is no evidence of a significant difference in bank profitability (as measured by
EARTA and EARGL) between the two types of banking
systems. Hypothesis HA2 is thus rejected, and H02 is
accepted. However, the effect of bank characteristics on
profitability is more pronounced in the group of CBs than in
the group of IBs.
H02: There is no significant difference between the
conventional and Islamic banking systems in
profitability.
The effect of capital adequacy measures (control variable) on
cost efficiency (CIR) is presented in Table 5. The influence
of a bank’s capital adequacy on cost efficiency is significant
and negative for Tangible Equity and Liquid Assets (see
Model 3), and marginally significant for Tier 1 ratio. If we
consider CBs and IBs separately, we will find that only
Liquid Assets have a strongly negative effect on cost efficiency, which is more pronounced in the sample of CBs. We
may conclude that there is a significant effect of a bank’s
capital adequacy on cost efficiency for both types of banks,
but this effect is limited only to Liquid Assets (banks with
more liquid assets are more efficient). Hypothesis HA3 is
thus partially confirmed.
HA3: There is a significantly different effect of capital
adequacy on the banks’ cost efficiency between the
conventional and Islamic banking systems.
The capital adequacy effect on cost efficiency as measured
by NIM is presented in Table 6. The effect of a bank’s
capital adequacy on its cost efficiency is strongly significant
in the whole sample (all the estimated coefficients are statistically significant). We observe similar (significant) relationship in the sample of CBs except for Funding fragility.
However, this effect is relatively weak in the sample of IBs
where Tier 1 ratio and Liquid Assets are marginally significant while Funding fragility is statistically significant at the
5% level of significance. In conclusion, the effect of capital
adequacy on a bank’s cost efficiency is found to be strong
and significant in the sample of CBs but relatively weak for
IBs. The capital adequacy effect on bank profitability as
measured by EARTA is reported in Table 7. We observe
that the effect of a bank’s capital adequacy on bank profitability is insignificant in the whole sample as well as in the
sample of IBs. For the sample of CBs, we find that Liquid
Assets have a strongly negative influence on bank profitability; that is, banks with less Liquid Assets can achieve
higher profitability. Hypothesis HA4 is thus partially
confirmed.
HA4: There is a significantly different effect of capital
adequacy on the banks’ profitability between the
conventional and Islamic banking systems.
The alternative tests for capital adequacy effect on the bank
performance (using EARGL as another measure of bank
profitability) are reported in Table 8. The capital adequacy’s
influence on EARGL is significant yet negative only for
Liquid Assets (see Model 3 and 4 for the whole sample of
UAE banks). Again, we observe a significant difference in
the capital adequacy effect between CBs and IBs. While in
the group of CBs both Tier 1 ratio and Liquid Assets have a
strong influence on bank profitability, no significant effect is
observed in the group of IBs (except for Liquid Assets in
Model 10).
The effect of risk-taking on cost efficiency as measured
by CIR is presented in Table 5. We find that the influence of
bank risk on CIR is strongly significant both in the whole
sample of UAE banks and in the sample of CBs; however, it
is insignificant for IBs. Moreover, this effect is significant
only when bank-level characteristics and capital adequacy
measures are added to the model together with the risk
measures (Equity Volatility and Log(Z)). The results for Log
(Z) show that the banks with high distance to default (low
risk) are more cost-efficient. However, the evidence indicates
that risk-taking does not influence the cost efficiency of IBs
in the UAE.
Similarly, the risk effect on cost efficiency when measured by NIM is presented in Table 6. The bank risk influence on NIM is significant and negative for the whole
78
F. Mrad and M. Mateev