If the behavioral model has been built to last a few years, how is the regulator going to trust that the model is fit for purpose over time? Considerable time and other resources are spent on building these models. A large bank may have hundreds of such models built across various time periods. Some of these models might not even be representative of the current business scenario. As new data comes in, the model fit may not be accurate. To mitigate this, a validation exercise is usually carried out after a certain time period, to calibrate the model on the latest data and assess if the original model parameters are observed at the time of validation.
Validation
Get SAS for Finance now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
