To identify policy enhancement opportunities, we designed a framework capable of mapping the application pool to the existing underwriting criteria – LTV, Credit Score and DTI. This framework was built for every product, loan amount and other criteria combination. The result was a Mortgage Policy Cube’.
We further overlaid the historical application pool into the cube and identiﬁed areas for assessment. Expected loss models were applied on each area of the cube to ensure that policy recommendations were within the bank’s implied risk tolerance levels. Policy change recommendations were developed within regulatory and business constraints. New exception thresholds were determined in a manner that was well within the policy for credit score and LTV combinations, where the expected loss levels are within the current expected NCL limit.
We also performed analysis on decline swap to compare accepted applications vs. applicants evaluating the credit score threshold, providing evidence that these applicants could be accepted as ‘within policy’.
- Our framework highlighted new criteria, updated addendums and provided evidence and recommendations to the credit committee
- The solution was capable of operationalizing recommendations to assess risk proﬁles and updating tools for NSI and High Touch, High Speed (HTHS) underwriters