From a sample of 241 firms in the Brazilian manufacturing sector, we conclude that parameters that define firms' capital structure are good signals of the risk of default on bank loans. In order to derive this result, we use a Logit regression, in which the explanatory variables come from factor analysis and stepwise regression. We also combine this methodology with a model of credit granting that incorporates integer linear programming as developed by Gehrlein & Wagner (1997).
factor analysis; linear programming; credit risk