By: Amanda Roth
As discussed earlier, the validation of a risk based-pricing program can mean several different things. Let’s break these options down.
The first option is to complete a validation of the scoring model being used to set the pricing for your program. This is the most basic validation of the program, and does not guarantee any insight on loan profitability expectations. A validation of this nature will help you to determine if the score being used is actually helping to determine the risk level of an applicant.
This analysis is completed by using a snapshot of new booked loans received during a period of time usually 18–24 months prior to the current period. It is extremely important to view only the new booked loans taken during the time period and the score they received at the time of application. By maintaining this specific population only, you will ensure the analysis is truly indicative of the predictive nature of your score at the time you make the decision and apply the recommended risk-base pricing.
By analyzing the distribution of good accounts vs. the delinquent accounts, you can determine if the score being used is truly able to separate these groups. Without acceptable separation, it would be difficult to make any decisions based on the score models, especially risk-based pricing.
Although beneficial in determining whether you are using the appropriate scoring models for pricing, this analysis does not provide insight into whether your risk-based pricing program is set up correctly or not. Please join me next time to take a look at another option for this analysis.