More Powerful Default Forecasts: Incorporating Signal Strength and Dependency
45 Pages Posted: 3 Jan 2008 Last revised: 16 Mar 2014
Date Written: March 16, 2014
The financial market crisis has exposed a weakness in predicting defaults. A new wave of models must be more powerful and should be more successful in predicting multiple defaults. I argue that such models incorporate signal strength, cross and serial dependencies. The conditional default probability is the most powerful predictor and I demonstrate that heteroscedastic probits with random effects represent a flexible framework to estimate the conditional probability. The lower the variance the higher is the signal strength of the underlying credit score. I derive an approximate likelihood procedure, including asymptotic standard errors, and apply it successfully to a dataset.
Keywords: Generalized Linear Mixed Models (GLMM), Heteroscedastic Probit, Receiver Operating Characteristic (ROC), Cumulative Accuracy Profile (CAP), Basel Committee on Banking Supervision, Credit Scoring, Credit Risk Modeling
JEL Classification: C13, C21, G21
Suggested Citation: Suggested Citation