In general, the
results from the z-score fixed-effects OLS regressions are qualitatively similar to those of
the logit model. Note that, since the z-score increases with the soundness of banks, a
positive coefficient indicates a risk-reducing effect - in contrast to the logit model where a
negative coefficient indicates lower risk. A few differences are worth noting: First, neither
the level of reserves nor the dummy for hidden liabilities exerts a significant effect on risk.
The insignificance of bank reserves and the hidden liabilities dummy is very counterintuitive
given the importance of these variable in the German banking landscape. However,
given the drawbacks of the z-score, we do not give too much weight to these results.
In general, theresults from the z-score fixed-effects OLS regressions are qualitatively similar to those ofthe logit model. Note that, since the z-score increases with the soundness of banks, apositive coefficient indicates a risk-reducing effect - in contrast to the logit model where anegative coefficient indicates lower risk. A few differences are worth noting: First, neitherthe level of reserves nor the dummy for hidden liabilities exerts a significant effect on risk.The insignificance of bank reserves and the hidden liabilities dummy is very counterintuitivegiven the importance of these variable in the German banking landscape. However,given the drawbacks of the z-score, we do not give too much weight to these results.
การแปล กรุณารอสักครู่..
