ALTERNATIVE METHOD FOR DETERMINING INDUSTRIAL BOND RATINGS: THEORY AND EMPIRICAL EVIDENCE

The financial-ratio-based credit-scoring model for bond rating system requires the maximization of two conflicting objectives (i.e., the explanatory and discriminatory power, simultaneously), which had not been directly addressed in literature. The main purpose of this study is to develop a hybrid multivariate credit-scoring model that combines the principle component analysis and Fisher's discriminant analysis using the MINIMAX goal programming technique so that the maximization of the two conflicting objectives can be compromised. The performance of alternative credit-scoring models is analyzed and compared using dataset from previous studies. We find that the proposed hybrid credit-scoring model outperforms other alternative models in both explanatory and discriminatory powers.