Thin File

Back to case studies

The client was a credit card provider looking to quickly expand its lending by targeting ‘thin’ files and bureau ‘bad’ files.


Its internal risk modeling process struggle to create an application score capable of ranking risk and segmenting the application population to identify acceptable customers for on-boarding.




The final scorecard was implemented and used to lend to the top 11% of the thin model score and top 9% of the bureau bad score. Whilst these cut offs points are relatively low they best capture the clients’ risk appetite with a cumulative bad rate of 14% and 15% respectively.


Given these two populations account for circa 60% of all applicants, this small lending cut-off enables significant increase to overall lending volumes whilst maintaining the risk profile of the overall portfolio.