| Summary: | Basel II allows banks to use internal rating model as the basis in
determining the risk of their respective assets. The availability of well rated and
validly qualified model has been a key prerequisite prior to before the banks use
internal models.
This research was conducted to evaluate whether the internal rating system
model applied in BNI, particularly in corporate loans, can be utilized to distinguish
the liquid (good) or default (bad) debtors. The internal rating system performance
model evaluation applies dicriminatory power techniques (Accuracy Ratio,
Receiver Operating Characteristic, Kolmogorof-Smirnov Statistic, Kullback-Leibler
Statistic and Information Statistics). Each variable�s characteristics test applies
Information Statistics, while the Probability of Default estimation test applies the
binomial test.
The results showed that in general the BNI internal model corporate credit
rating system has shown good performance and been able to distinguish the liquid
(good) or default (bad). Each parameter calibration test particularly on non-financial
aspects has indicated that the criteria settings for each parameter are still relevant
and are able to distinguish between good and bad debtors. However in the
parameters� financial ratios, the results indicate that this parameter is not good
enough in distinguishing the debtor qualities which generally marginal or poor
derivations are achieved.
|
|---|