Data Both Helps to Underwrite Applicants Who Lack Traditional Credit History and Improves Risk Sorting Among Borrowers Who Are Ranked Similarly by Traditional Scoring Systems
More than 96 percent of American households have bank or prepaid accounts, and account records are increasingly easy to access electronically
Today’s credit markets depend heavily on information provided by credit bureaus to underwrite consumers and small businesses. The lack of easy access to reliable underwriting information also makes it hard for business start-ups to obtain loans.
Cash-flow information-such as data from consumers’ deposit and card accounts or from small business accounting software-is one of the most promising options for improving automated underwriting because it provides a more detailed and timely www.signaturetitleloans.com/payday-loans-sc picture of how applicants manage their finances than traditional credit reports.
FinRegLab has conducted one of the first ever independent studies of cash-flow variables and credit scores using data from six non-bank financial services providers – Accion, Brigit, Kabbage, LendUp, Oportun, and Petal – that have begun using the information in an effort to provide unsecured credit to consumers and small businesses who may have difficulty obtaining loans from traditional sources. The report, The Use of Cash-Flow Data in Underwriting Credit: Empirical Research Findings, can be found here. FinRegLab retained Charles River Associates to help design and conduct an independent analysis of the data with regard to its general predictiveness, its ability to increase access to credit, and its potential fair lending effects.
The study finds compelling evidence that the cash-flow variables and scores tested were predictive of credit risk across the diverse set of companies, populations, and products for which loan-level performance data was available. More specifically, the study has four main findings:
- General predictiveness: The predictiveness of the cash-flow scores and attributes was generally at least as strong as the traditional credit scores and credit bureau attributes studied. The results suggest that cash-flow variables and scores can provide meaningful predictive power among populations and products similar to those studied where traditional credit history is not available or reliable.
- Combined models: The cash-flow scores and attributes appeared to separate risk in somewhat different ways than traditional scores and attributes, such that the cash-flow data frequently improved the ability to predict credit risk among borrowers that are scored by traditional systems as presenting similar risk of default. These results occurred across traditional credit score bands.
- Inclusiveness: The participants appear to be serving substantial numbers of borrowers who may have historically faced constraints on their ability to access credit, although data limitations did not permit a consistent quantitative analysis to be applied across all companies. Among participants where such data was available, for example, the percentage of borrowers with traditional credit scores below about 650 was approximately 45 to 50 percent.
- Fair lending effects: When divided into subgroups based on likely race, ethnicity, and gender, the degree to which the cash-flow data predicted credit risk appeared to be relatively consistent across subpopulations. Moreover, when compared to traditional credit scores, the cash-flow based metrics appeared to predict creditworthiness within the subpopulations at least as well as the traditional scores, and better in selected cases. Overall, the cash-flow data appeared to provide independent predictive value across all groups rather than acting as proxies for demographic group.
In the small business credit market, use of cash-flow data in underwriting appears to be spreading relatively rapidly. As detailed in a second report by FinRegLab to be released in , use of electronic cash-flow data was originally pioneered by fintechs that used electronic bank account records, accounting software feeds, and payment records to facilitate faster underwriting of smaller loans to small business owners. Over the last decade, a growing range of market actors has also adopted automated underwriting systems that rely in whole or part on cash-flow data, including e-commerce platforms, payment processors, accounting software providers, community development organizations, and banks.
But an estimated 45 to 60 million consumers lack sufficient credit history to generate reliable credit scores, and millions more struggle to access to affordable credit because their scores are low
FinRegLab expects to release a third report in September that will provide a more detailed snapshot of the use of cash-flow data in consumer credit markets, where adoption has been slower, as well as an analysis of policy issues raised in both consumer and small business markets. In particular, the report will provide a detailed description of market, legal, and policy issues raised with regard to both the use of cash-flow data in underwriting and the underlying transfers of data between companies for credit and other purposes. The report builds on three working groups that FinRegLab convened in late 2018 to solicit insight and opinion from more than 80 representatives of fintech companies, banks, data aggregators, advocacy organizations, and research institutions.