Risk Metrics Validation Study

How can you predict the likelihood that a mortgage loan will suffer from delinquency, foreclosure, or other types of loan problems in the first year? Is there a mortgage risk exposure system that can predict nonperforming loans? Interthinx has released its most recent publication statistically validating the precision of its scoring system for mortgage risk exposure.

Challenge
How can you predict the likelihood that a mortgage loan will suffer from delinquency, foreclosure, or other types of loan problems in the first year? Is there a mortgage risk exposure system that can predict nonperforming loans? Interthinx has released its most recent publication statistically validating the precision of its scoring system for mortgage risk exposure.

Results
Interthinx has used mortgage performance data from credit bureaus to allow ongoing comparison of scoring model predictions with actual loan outcomes. Analysis based on this dataset has demonstrated that the Interthinx system can consistently predict the likelihood that a loan will suffer from delinquency, foreclosure, or other types of loan problems in the first year. With remarkably clear separation between high- and low-risk loans, the statistical validation confirms the leading position of Interthinx in predictive performance metrics.

Techniques
The results use a combination of detailed loan information, performance data, and Interthinx FraudNET™ technology. That allows modeling of complex interactions among different data elements and provides historical context on property and participants. Interthinx has aggregated and scored mortgage application data for more than a decade, making it a premier provider of risk intelligence tools capable of this comprehensive analysis.

Download this publication

View more ISO publications