Student loan office casts net for ‘creative’ credit analysis
The Department of Education’s Office of Federal Student Aid is looking outside the box for ways to determine the credit worthiness of those applying for student loans.
In a recent request for information, the office said it was developing new statistical models to gauge portfolio risk, predict individual borrower outcomes and test hypotheses about borrower behavior.
The FSA said it wanted to use the models to support its decision-making about borrowers as well as to augment its existing research with additional credit, debt, income and other risk related information.
That included finding out about new sources of borrower credit information, including data on credit scores and what “other national databases could be leveraged and integrated into these analyses.”
For example, “any diverse, creative or unconventional use” of credit score data used by the retail banking sector to better identify customer segments, inform decision-making or manage credit risk may be helpful, the FSA said.
The FSA said it had broad access to demographic characteristics of borrowers from the Free Application for Federal Student Aid and other sources, but wished to know whether that information could be even further leveraged to improve the analysis of an applicant’s credit risk.
Additionally, the office said it also wanted to know whether “continuous or semi-continuous data feeds” were available to help analyze the data.
The RFI may well yield a set of technologies enabling the federal student loan office to pursue “creative and unconventional” credit testing methods.
Louis Beryl, the chief executive of Ernest, a “data-driven” bank based in San Francisco, is using such methods and analytics to weigh credit worthiness of its loan customers.
“We are building the consumer bank of the future,” Beryl told the New York Times.
In assessing loan applications, the bank uses software that scans thousands of bits of information related to customers’ behavior to help the company make a credit decision. The information might be related to applicants’ purchasing patterns, bill-paying timeliness or even where they attended college.
While the FSA is unlikely to adopt these methods soon, the RFI signals its clear interest in pursuing more sophisticated analytics to manage the loan program.
Connect with the GCN staff on Twitter @GCNtech.