The Bayesian Additive Classification Tree Applied to Credit Risk Modelling

We propose a new nonlinear classification method based on a Bayesian "sum-of-trees" model, the Bayesian Additive Classification Tree (BACT), whichextends the Bayesian Additive Regression Tree (BART) method into the classification context. Like BART, the BACT is a Bayesian nonparametric additivemodel specified by a prior and a likelihood in which the additive components aretrees, and it is fitted by an iterative MCMC algorithm. Each of the trees learnsa different part of the underlying function relating the dependent variable tothe input variables, but the sum of the trees offers a flexible and robust model.Through several benchmark examples, we show that the BACT has excellentperformance. We apply the BACT technique to classify whether firms would beinsolvent. This practical example is very important for banks to construct theirrisk profile and operate successfully. We use the German Creditreform databaseand classify the solvency status of German firms based on financial statementinformation. We show that the BACT outperforms the logit model, CART and the Support Vector Machine in identifying insolvent firms.