Credit Granting: A Comparative Analysis of Classification Procedures

Financial classification issues, and particularly the financial distress problem, continue to be subject to vigorous investigation. The corporate credit granting process has not received as much attention in the literature. This paper examines the relative effectiveness of parametric, nonparametric and judgemental classification procedures on a sample of corporate credit data. The judgemental model is based on the Analytic Hierarchy Process. Evidence indicates that (nonparametric) recursive partitioning methods provide greater information than simultaneous partitioning procedures. The judgemental model is found to perform as well as statistical models. A complementary relationship is proposed between the statistical and the judgemental models as an effective paradigm for granting credit. THE CREDIT GRANTING process involves a tradeoff between the perceived default risk of the credit applicant and potential returns from granting requested credit. The main objective in credit granting is to determine the optimal amount of credit to grant. The amount of credit requested along with other financial and nonfinancial factors influences the assessment of default risk. While default risk assessment can be appropriately modeled using classificatory models, integration of such risk assessment with potential returns can be accomplished using a dynamic expected value framework.' Academic research on credit granting can be grouped into two basic categories:2 (i) attempts to apply classification procedures to customer attribute data and develop classification models to assign group membership in the future; and (ii) attempts to explicitly recognize the need to integrate such risk assessment with potential return as well as allow for differing degrees of credit investigation depending on the costs of such investigation. To our knowledge, this study represents the first attempt at examining the relative performance of classification procedures for corporate credit granting. Relative performance is measured in terms of how well the models replicate expert judgement. We examine four statistical models: multiple discriminant analysis (MDA), logistic regression * Northeastern University and University of Cincinnati, respectively. We are grateful to Robert Eisenbeis and Sangit Chatterjee for their insightful comments on an earlier draft. We also thank an anonymous Fortune 500 corporation as well as the Credit Research Foundation for their support. The usual disclaimer applies. 1 This paper is primarily concerned with the credit granting decision in an industrial (nonfinancial) setting. A more detailed version of this paper is available from the authors. Further, for a review of the various motives for corporations to extend credit, refer Emery [10]. 2 We do not make any attempts to present a comprehensive review of the literature. Interested readers are referred to an excellent review by Altman et al. [1].