The client had very little, if any, information about their recovery portfolio and had stopped originating new loans during the financial crisis. Resources and business experience were limited.
Performed intensive analysis on 36 unique variables (investigated a total of 99) and identified 14 predictive of payment performance. Provided four payer categories and defined what borrowers in each looked like.
The client, a non-profit student loan servicer was focused strictly on servicing and collecting previous loans that had been made. They had limited resources to conduct analytics on the recovery portfolio. Most decisions were made by “gut feel” and there was little insight of variables driving payment in recovery.
Bridgeforce analyzed a dataset containing all active recovery accounts, with variables captured at originations, performance variables in recovery, and credit bureau variables. We analyzed performance trends to identify the most predictive variables of repayment. Payers were categorized as: No Payers, Light Payers, Moderate Payers and Significant Payers. We defined what a borrower in each category looks like.
Ultimately, we analyzed 36 unique variables (although investigated 99) and identified 14 predictive of payment performance. These allowed the client to identify a population of accounts to bring in house / utilize for further segmentation, as well as use when purchasing data from the credit bureaus in the future. As the variables are further incorporated in their strategy, this foundational work will enable them to increase liquidation rates and ultimately improve profitability for the firm.