New option tests value of risk model redevelopment projects
Risk managers make critical lending decisions that depend on accurate scoring models to manage credit losses, identify application fraud and minimize customer attrition. The innovative FastTrack Modeling process brings the lending industry a more time-efficient way to determine if an incumbent scoring model can be improved.
“FastTrack Modeling offers upfront performance insights in just a few weeks, before you embark on a costly risk model redevelopment project,” said
Using historical credit data and attributes, the FastTrack Modeling process quickly creates a “challenger” to the existing scoring model. The two models are tested with actual performance data, using measures such as C values, KS, and percent losses to see which one better correlates with the performance data.
DMS clients can easily quantify the value of developing a new production model relative to their investment in additional time and expense. If the existing model more accurately reflects account performance, no action needs to be taken since the model is likely functioning as expected. If the challenger model offers a stronger relation to performance, then a better model likely exists. The client has the option for DMS to produce a new full production model based on the work already performed with FastTrack Modeling.
The DMS Decision Science team that developed FastTrack Modeling is experienced in full model development activities, including origination, retention, pricing and survival modeling. Team members possess extensive experience working with some of the largest bank and credit card companies in