One of the largest banks in the US had a problem in their mortgage division: sales were steadily declining, delinquency rates were on the rise, and pre-payments were at an all time high. This spelled disaster for the profitability of the division.  One of Peak Data Solutions’ consultants was called in to recommend a solution and revitalize the division. The bank’s goals were to:

• Increase sales by increasing the acceptance rates of customers

who call into the division inquiring about a mortgage

• Reduce losses by decreasing mortgage delinquency rates

• Reduce losses by decreasing pre-payment rates

After careful study of the division, the mortgage sales process, and customer risk profiles, several solutions were recommended and subsequently implemented:

Segmentation: Current practices in the bank segmented the population based on grade levels (A, B, C). These grades were based on business rules deduced from past experience. Even though this segmentation did an accurate job of rank ordering risk, further segmentation would provide additional value.

The following is an example of the grading system. Customers with a high grade had lower charge-off rates and were offered a better price than the rest of the population.  However, the bank was only able to book a small percentage of these customers and most of those who signed on pre-paid their mortgage within two years.

Custom Risk Score: A custom risk score was developed using variables that are independent of the grading system. This helped discover new segments within the population.

Changing of Pricing Strategy: We proposed changing the pricing system to take advantage of newly discovered segments.

Solution: lower prices for this segment
Results: Acceptance rates increased, overall charge-off rates decreased (more customers in this segment) and pre-payment rates decreased (since a lower initial price motivated customers to remain with the bank)

To test this new revenue management system, software was developed to automatically segment customers calling in and provide customer service representatives with automatic price recommendations.  The results were compared to a similar division working with the current prices.

After enough time had passed to measure the results of pre-payments and delinquencies, results showed a 50% increase in revenues, a 20% decline in charge-offs and a 30% decline in pre-payments.  This led to a 25% increase in profitability.