Director Kathy Kraninger lamented on the implications of using “alt data” not typically considered in traditional credit scoring, something fintechs have been doing with more frequency. The alternative data takes into account information not normally used in credit scoring.
“Some of the examples include rent and utility payments, employment history, and cash-flow information,” she said during the conference. “In recent years, an increasing number of financial institutions, especially fintechs, have been using alternative data to make credit and pricing decisions. Understanding how such information affects consumers in credit markets has important policy implications.”
Kraninger also spoke broadly on small business lending research and data as well as the overall accuracy of credit reports.
“Prevalent use of alt data holds the promise of potentially significant benefits for some consumers, like extending credit more quickly, but may present certain risks for others,” she added. “For example, if the data used in credit and pricing decisions become more complex, and they are fed into more complex models, it could become difficult for some consumers and even financial institutions to understand how credit and pricing decisions are made.”
"A comparison between 2018 and 2020 shows the percentage of financial firms not using alternative data at all has plummeted from 30% to 3%. Social media, Web scraping and credit card data are the most frequently used categories of alternative data. https://refini.tv/3nwZGbg"
According to Kraninger, CFPB research shows traditional scoring has left “critical gaps” in some consumers’ credit access. Citing data from 2010, 26 million people, which represented approximately 11% of the nation’s adult population, were classified as “credit invisible.” As such, there was no file for them logged with any major credit bureau. Further, another 8.3% were considered “unscorable” as their files were too stale or thin to create a reliable score.
Credit reporting agency Experian argues many lenders are ignoring a wealth of data by limiting themselves to traditional credit info alone. They say things like rental payments, consumer-permissioned data and “full-file public records” are extremely valuable.
“These insights help lenders to expand their credit universe, identify and prevent fraud, determine an applicant’s ability to pay, and verify an identity all while mitigating risk,” according to Experian. “Ultimately, this new data drives greater access to credit for consumers and profitable growth for lenders through more informed lending decisions.”
Kraninger says the CFPB is hoping to learn more about potential developments in this area, and understand how market players can mitigate consumer risk. She called for feedback on a number of questions. Namely:
- “What are the most important questions to answer when it comes to alternative data?”
- “What types of privacy concerns should we be considering in our research?”
- “Do you have any guidance on how to assess the use of alternative data on access to credit?”