Vintages Not Known by a Sommelier
Altering credit strategies for maximum accuracy has to be the task of credit managers in the current environment. In our analytics leader’s final post, he talked about what are some steps that can be taken. His best point was “Am I doing all that I need to be doing to accurately determine risk management policy, considering the reality of economic changes?” He mentioned doing vintage analysis on younger accounts as a good way to track bad rates. The news has blessed us with a perfect example of when this method might help drive policy change.
Right now, no one has missed the news about the iPhone and it’s big relaunch party in February. Expectations of migrations ranged from 11 to 13 million iPhone users in the first year. Obviously this number was to be heavily front-loaded towards the launch. How many conversations do you think Marketing and Risk had about the acceptable risk of customers given the size of the launch? What about credit operations? What was the approach? Toy with the score bands? Loosen the criteria? Prepare a resume and a letter of resignation? Credit score predictiveness is determined after 24 months of performance. 24 months ago people were just figuring out what “apps” were. Timothy Geithner had just announced the US government was in the business of purchasing toxic assets. Do you think starting from that period in history accurately forecasts the bad rates on 11 to 13 million new phone activations?
The case in point focuses on why there is a need to track and compare delinquency levels month by month. The question though, is how. How does a retail banking risk analyst team understand this? It starts with data visibility. Understand that vintage analysis requires data to be aggregated. Whether directly dumped to a group of analytical experts or fed into a data warehouse capable of driving actionable vintage KPI’s, actionable advanced analytic insights start with centrailized viewing of the data. A good team or tool can track accounts as they activate and provide relevant one, three, and six month performance stats that can indicate if policies are accurate. When combined with adequate decisioning frameworks, analyst predictions can drive policy change.
To get back to our example in question, by my count there are 19 different mobile phones that could serve as voice and data predictors from 2011. Do you think any of those devices map better to future trends than bad rates that are being reported now? Delivering on these results requires a trusted squad of internal analysts with a system capable of vintage reporting and segmentation. If you want to talk to our team of trusted squad of analysts with their systematic vintage reporting and segmentation system, let us know.
If you are interested in a monthly e-mails about blog updates, sign up for our newsletter
This post was contributed by: