A credit scoring model is a mathematical model used to estimate the probability of default, which is the probability that customers may trigger a credit event (e.g., bankruptcy, obligation default, failure to pay, and cross-default events). In a credit scoring model, the probability of default is normally presented in the form of a credit score. A higher score refers to a lower probability of default.
Although there are a number of common credit factors in credit scoring models, different types of loans may involve different credit factors specific to the loan characteristics. For example, the credit factors for a credit card loan may include payment history, age, number of accounts, and credit card utilization; the credit factors for a mortgage loan may include the down payment, job history, and loan size, among others.
Accurate and predictive credit scoring models help maximize the risk-adjusted return of a financial institution. However, markets and consumer behavior can change rapidly during economic cycles, such as recessions or expansions. For this reason, risk managers and credit analysts need to be able to create, adjust, and validate models in a nimble manner. Techniques used to create and validate credit scoring models include:
- Logistic regression andlinear regression
- Machine learningandpredictive analytics
- Binning algorithm (e.g., monotone, equal frequency, and equal width)
- Cumulative Accuracy Profile (CAP)
- Receiver operating characteristic (ROC)
- Kolmogorov-Smirnov (K-S) statistic
For more information on credit scoring models, see MATLAB®, Financial Toolbox™, and Risk Management Toolbox™.
