March 7, 2013
Big Data ushers in new ways to determine creditworthiness
In line with Predictive Personalization, one of our 10 Trends for 2013, a number of startups, along with some traditional banks, are taking a Big Data approach to consumer lending. As Slate recently reported, several companies are analyzing unstructured sources of data (such as that generated by social media) in combination with conventional information to offer more precise assessments of creditworthiness, often targeting consumers with no credit history or poor credit.
U.S.-based LendUp, which positions itself as “a better alternative to payday loans,” offers short-term, high-interest loans based on data it verifies through sources such as Facebook and Twitter to improve an applicant’s credit profile. ZestFinance says it “takes an entirely different approach to underwriting using machine learning and large-scale big data analysis,” focusing on people with bad credit. Lenddo, which emphasizes “financial empowerment,” extends credit in the Philippines and Colombia—where traditional credit bureaus don’t exist—by using applicants’ social media contacts as references. And Kreditech, a German startup, touts a technology that analyzes more than 8,000 data points derived from e-commerce behavior, location, mobile device data and social media info, among other things.
Meanwhile, traditional banks are pursuing their own big data strategies to further refine their lending practices. In the U.S., JP Morgan Chase, Bank of America, Citigroup and Wells Fargo are among those moving beyond traditional gauges of creditworthiness, considering everything from how often consumers eat out to whether they shop mostly at upscale department stores or discount emporiums. Data sources include credit and debit card receipts, as well as social media profiles. Regulators are questioning the practice of using personal data without an individual’s knowledge to make life-affecting financial decisions. But the biggest threat for lending may be the “creep factor”—that is, when loan applicants become so creeped out by how much data lenders can gather that they decide not to apply in the first place.