TransUnion Study Reaffirms Its Position on Using Non-Traditional Credit Data

According to a joint study released today by The Brookings Institution and the Political, Economic Research Council (PERC), including “non-traditional” credit data (utility and telecommunications payment information) in a consumer’s file significantly levels the credit risk playing field for “thin-file” consumers with those in the credit mainstream. TransUnion provided more than 8 million anonymous TransUnion credit files that had a strong focus on the underserved credit population as well as analytic support to the study entitled “Give Credit Where Credit Is Due.”

TransUnion has always stated that more information included in a consumer’s file will lead to better decisions about that consumer. For many years, TransUnion has accepted positive and negative payment histories from entities operating in the utility, telecom, and rental housing industries and has publicly noted its intention to expand its efforts in these markets.

“The study’s results continues to validate TransUnion’s position on the many benefits full file credit reporting provides,” said TransUnion’s Chet Wiermanski, vice president, Analytics and Decisioning. “Including this type of data could allow the credit industry to offer credit to a broader and more diverse set of consumers who traditionally may have been overlooked by the conventional risk assessment tools deployed today.”

Other key findings of the “Give Credit Where Credit Is Due” study include:

• The inclusion of nontraditional data potentially would give greater access to more affordable mainstream sources of credit, while decreasing the credit risk to lenders. In some instances, the number of minorities accepted is 22 percent greater when nontraditional data is included in credit files than it would be without the inclusion of nontraditional data.
• Based upon this research, access to utility and telecommunications data may possibly assist lenders with reducing bad loans by as much as 29 percent.

Along with improving individual consumer credit profiles and their credit scores, the study also concluded that using more comprehensive data can improve the performance of credit scoring models. The study analyzed several different credit scoring models, including the industry’s newest generic scoring model, VantageScore^SM. Across the board, a significant rise in model performance was evident. In fact, among thin-credit file consumers model performance improved up to 300 percent, which could possibly enable lenders to make larger loans – at lower interest rates to these consumers – because they can be more confident in the ability of credit risk models to predict future credit performance.

The complete study, along with an executive summary, can be found on PERC’s website at http://www.infopolicy.org/.