Classification of commercial bank loans through policy capturing
Academic Article
Overview
Identity
Additional Document Info
Other
View All
Overview
abstract
The judgmental policy of a group of national bank examiners in classifying commercial bank loans was identified using stepwise multiple discriminant analyses and was expressed in the form of a linear model. Credit scenarios (which were composed of both financial and non-financial cues) were presented to 24 national bank examiners at the Office of the Comptroller of the Currency (OCC) in Washington D.C. Each examiner received 20 scenarios and was asked to classify the scenarios in one of the five classifications used by the OCC. Based upon 75% of these observations, a linear discriminant model was developed using stepwise MDA analyses. The remaining observations were used to validate the model developed. Non-financial variables appeared to be important discriminators; the financial variable of industry comparison of the interest coverage ratio using cash flow was an excellent discriminatory variable; and industry comparisons and trends of financial variables appeared to be more important than the variables themselves. The model developed contained two of these variables: (a) an industry comparison of the interest coverage ratio using cash flow and (b) the trend of the debt/equity ratio. 1980.
