Challenge
A leading private issuer of mortgage-backed securities and mortgage-related, loan asset-based securities needed to employ quantitative models for predicting outcomes when a borrower's mortgage payment is 60 days past due. Such quantitative models were a prerequisite to underwriting by the Federal Home Loan Mortgage Corporation (Freddie Mac).
Actions
An advanced, data-mining solution -that generated real-time, predictive models on a case-by-case basis- was developed and loaded with historical data. This solution indicated whether an individual mortgage would result in a foreclosure, a cure, or some intermediate resolution, once payment became 60 days past due.
Result
The solution outperformed the best statistical models developed by our customer's in-house statistical staff, as well as the state-of-the-art modeling technology offered by Freddie Mac.
