2OS used existing bureau data to build a custom machine learning risk model and credit policy for a small business lender to decrease portfolio exposure and risk while maintaining profitability.
The custom model developed by 2OS outperformed the industry benchmark. The credit policy was designed to meet the lender's target return while maintaining loss rate constraints.
The model decreased charge off rates on the client's portfolio by 11%, compared to the industry benchmark, while maintaining target approval rates.
Credit Policy Construction
2OS developed a credit policy for acquisitions and line assignments (approve/decline and receivable purchase size assignment) to achieve the client's objectives.