Files in this item
|(no description provided)|
|Title:||Development and Evaluation of a Credit Scoring Model for the Farm Credit System|
|Author(s):||Splett, Nathan S.|
|Doctoral Committee Chair(s):||Barry, Peter J.|
|Department / Program:||Agricultural Economics|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
Business Administration, Banking
|Abstract:||Credit scoring models were developed to identify the credit risk in term and operating loans for the principal purposes of evaluating portfolio quality and monitoring risk at the loan and portfolio levels in the Farm Credit System. Experience based (lender) and statistical (academic) approaches were used jointly. Initially, experience term and operating loan models were developed, utilizing financial ratios from the Farm Financial Standards Task Force (FFSTF) as the explanatory variable measures. Then, logit regression was employed to estimate term and operating loan models, utilizing classifications and variable measures from the experience models as the dependent and independent variables, respectively. The procedure compared the experience model and statistical model on the basis of the variable measures and classifications, and the FFSTF measures were evaluated with respect to their applicability in credit scoring models. Results from the experience and Logit models' applications indicated similarity in the variable measures selected and classifications, and indicated that the FFSTF measures were appropriate in the models.
The credit scoring accuracy of the experience models was evaluated across structural characteristics of farm businesses, recognizing that trade-offs may be necessary between credit scoring accuracy and the range of structural characteristics over which the models apply. The structural characteristics were described by three "size" categories, including total assets, value of farm production, and total acres operated; five farm business "types", including grain, hogs, dairy, beef/feeder cattle, and poultry; tenure; location (state); and loan size. Results of the models' application among the categories indicated that various structural characteristics may influence credit scores.
Term and operating loans were used to develop term and operating loan models, respectively. The term and operating loan models were applied to operating and term loans, respectively, to determine if two models were necessary. The models place different weights on the variable measures, and the loan data were different in terms of their financial structures, advocating the use of separate models. Classifications varied substantially in the cross-application of the models, indicating that separate term loan and operating loan models were appropriate, rather than using a single model.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1993.
|Date Available in IDEALS:||2014-12-17|
This item appears in the following Collection(s)
Dissertations - Agricultural and Consumer Economics
Graduate Dissertations and Theses at Illinois
Graduate Theses and Dissertations at Illinois