skip navigation

Sunday, October 4, 2015

[ – ] Text Size [ + ]  |  Print Page

Update Newsletter: Summer 2008

Alternative Data and Its Use in Credit Scoring Thin- and No-File Consumers

Recent economic events remind us that accurately predicting a particular consumer's credit risk can be a difficult task, especially among consumers with little or no credit history. At the same time, consumer lenders know that within the pool of the 35 to 70 million estimated U.S. consumers with "thin" or nonexistent credit files, there are many who represent potentially bankable customers. As a result, a growing interest has developed in determining whether there are other payment behaviors not captured by traditional credit reporting agencies that might lead to profitable underwriting approaches. To gain further insight into this area and better understand the challenges surrounding the use of alternative payments data in credit scoring, the center hosted a workshop with Arjan Schütte, associate director of the Center for Financial Services Innovation (CFSI). Arjan and CFSI have been researching alternative scoring models as part of their broader mission to improve opportunities for access to financial services for low- and moderate-income consumers.1 Industry Specialist Julia Cheney summarized the findings presented at the workshop in the discussion paper titled "Alternative Data Alternative Data and Its Use in Credit Scoring Thin- and No-File Consumers and Its Use in Credit Scoring Thin- and No-File Consumers."2 PDF

During the workshop, Schütte noted that one of the key challenges in developing the market for alternative data is determining which data are the best predictors of risk. Research to date suggests that insights can be gained by determining if the underlying transaction is "cash-like" or "credit-like." The more credit-like a transaction is, the more helpful it should be in determining the likelihood of whether a thin- or no-file consumer will make future payments on traditional credit products. The extent to which transaction types are used is also relevant. The more widely used, the more efficiently data analysis standards can be applied across a larger population. Conversely, if coverage is limited, the incremental benefit derived from the data may be less than needed to justify the costs of gathering the data. On the supply side, the structure of the particular data-furnishing industry is also an important determinant of the feasibility of using alternative data. If data furnishers are highly concentrated, scale efficiencies are gained, making it more likely that those furnishers' efforts to report this information will be successful. Based on these criteria, Schütte noted that utility and telecom payments represent good examples of credit-like alternative payment transactions that are broadly used by consumers and where the data-furnishing industries are relatively concentrated.

Data furnishers, including utility and telecom companies, are one of three types of organizations on the supply side of the alternative data market. Furnishers supply payment data to repositories that manage the databases storing the data that are ultimately used by the third party in the supply chain, data scoring firms that apply analytics to generate a credit risk score. Schütte noted that the supply side of the market for alternative data is rapidly evolving with many new and established companies contributing data and providing risk analysis that incorporate at least some elements of alternative data in order to improve underwriting decisions. However, he noted that further growth of the market depends largely upon there being a ready and regular supply of data. As the benefits to sharing information with data repositories is determined to outweigh the costs of reporting, the supply of alternative data is expected to expand.

On the demand side, lenders' interest is contingent on a number of factors. First, data sources must be broad and deep, offering redundancy and high hit rates. Second, data delivery systems need to be improved so that they integrate both alternative and traditional data and allow lenders to use existing channels and sources. Third, risk managers must build trust with new systems by realizing the incremental benefits gained by incorporating alternative data into credit underwriting and other business decision models. In sum, Schütte concluded that lender demand will grow as it becomes evident that incorporating alternative data into credit scoring models will allow the profitable expansion of portfolios.

Incorporating alternative payments data in current credit scoring practices presents additional challenges, including the costs of modifying legacy systems, the costs and complexities of changing IT infrastructures, legal and regulatory hurdles, and the broad economic impact of extending the market for consumer credit. Ultimately, though, the continued evolution of supply and demand for alternative data in the credit information markets will center on the strength of the business case to motivate furnishers of alternative data to voluntarily share payment information with data repositories and whether, then, the data can be used effectively to improve underwriting decisions.