The CFPB recently issued a request for information (“RFI”) seeking information about the use of alternative data and modeling techniques in the credit process. According to the CFPB, alternative data and modeling techniques are changing the way some financial services providers conduct business, but come with potential benefits and risks. The CFPB issued the RFI to assist it in better understanding these benefits and risks.
In the RFI, the CFPB contrasts traditional modeling techniques that rely upon traditional data elements as inputs – such as income, assets, and other traditional data supplied by the nationwide consumer reporting agencies – with a broad range of new alternative data and modeling techniques which, according to the CFPB, firms are either using or contemplating using. The RFI provides examples of alternative data, including: (1) data showing trends or patterns in traditional loan repayment data; (2) payment data relating to non-loan products requiring regular (typically monthly) payments, such as telecommunications, rent, insurance, or utilities; (3) checking account transaction and cashflow data and information about a consumer’s assets, which could include the regularity of a consumer’s cash inflows and outflows, or information about prior income or expense shocks; (4) data that some consider to be related to a consumer’s stability, which might include information about the frequency of changes in residences, employment, phone numbers, or e-mail addresses; (5) data about a consumer’s educational or occupational attainment, including information about schools attended, degrees obtained, and job positions held; (6) behavioral data about consumers, such as how consumers interact with a web interface or answer specific questions, or data about how they shop, browse, use devices, or move about their daily lives; and (7) data about consumers’ friends and associates, including data about connections on social media. The RFI also provides examples of alternative modeling techniques, including: (1) decision trees (or sets of decision trees, such as “random forests”); (2) artificial neural networks; (3) genetic programming; (4) “boosting” algorithms; and (5) K-nearest neighbors.
The RFI indicates that potential benefits of using alternative data and modeling techniques may include: (1) greater credit access; (2) enhanced creditworthiness predictions; (3) more timely information; (4) lower costs; and (5) better service and convenience. Potential consumer risks identified in the RFI include: (1) privacy concerns; (2) data quality issues; (3) lost transparency, control, and ability to correct; (4) increased difficulty in changing credit standing through behavior; (5) increased difficulty educating consumers about the factors that led to a particular credit decision; (6) unintended or undesirable side effects (such as, for example, a false impression of instability generated by members of the military, who may move frequently); (7) discrimination; and (8) other violations of law, including ECOA, FCRA and Reg. V, as well as UDAAP violations.
A copy of the RFI is available here: http://files.consumerfinance.gov/f/documents/20170214_cfpb_Alt-Data-RFI.pdf.