Retail Lending Loan Forecasting Model (RLLF)

This is a desktop application Prescio developed with extensive data analysis, portfolio segmentation, and reporting capabilities.

Prescio was approached by a western U.S. bank to help model their residential mortgage, consumer loan, small business, and cash reserve portfolio loss forecasts. The model utilizes a Markov chain approach to the portfolio loss forecasting that incorporates both assumed (forecasted) new business for each portfolio on a monthly basis as well as expert judgment adjustments to account for macro and microeconomic variables.

The application is currently deployed at the bank and is under further development and enhancement. In future versions of the model, regression analysis, time series forecasting, and a Bayesian updating technique for the Markov chain are anticipated to be incorporated. Prescio also anticipates a vintage analysis approach to loss forecasting to be incorporated in future versions of the application.

Latest posts by Prescio Consulting (see all)