Nationwide and Jaywing explore “explainable” AI driven application scoring
Prompted by the desire to boost the performance of its own unsecured loan application risk models, Nationwide wanted to embark on an analysis of different risk modelling techniques.
Due to industry concerns about the transparency of some AI-based modelling techniques, Nationwide wanted to find out how Jaywing’s Archetype software compared to other approaches.
The proprietary AI-based modelling software was shown to a generate a 6% uplift in predictive power, compared to traditional linear regression. This offers the potential to increase acceptance rates or reduce bad debt.
While the results were comparable with other AI-based modelling approaches, Nationwide was impressed by Archetype’s ability to offer the analyst control over how the input variables are used and to explain how the model behaves.
“Archetype clearly represents a powerful alternative in building credit scores,” says Matthew Jones, head of retail modelling at Nationwide. “We were impressed by the ease of use, the rapidly-realised benefits and ability to enforce intuitive behaviour on key variables in neural networks.”
He adds: “Jaywing’s approach and depth of knowledge has made us confident that the use of neural networks and similar technologies has reached a level of maturity where it can be safely deployed within the credit risk arena.”
Jaywing has recently launched an AI consultancy practice.