The secret to increasing lending, expanding credit access and fighting bias is plentiful and readily available — if only financial institutions can use it.
Those are all potential benefits of analyzing and modeling cash flow data, which is the information generated by customers when they use and receive funds from their deposit accounts. The technological capabilities, including machine learning models, now allow financial institutions to use customer transaction data to paint a broader, better picture of that customer’s borrowing capability and ability to repay, potentially uncovering additional borrowers. Not every consumer or small business has a long and robust credit history, consistent payroll income or unbroken employment — attributes that factor into traditional underwriting. That doesn’t necessarily mean those customers would be a poor credit risk.
“It’s amazing the data that a bank or credit union has within their walls, that they’re often not using to make these types of decisions,” says Laura Kornhauser, CEO and cofounder of the fintech Stratyfy, on this episode of Reinventing Banking. “Oftentimes, [it’s] not because they don’t want to use it, or they don’t get that [it’s] valuable. They just don’t have the right tools and technologies to use it. You’re not seeing those banks and credit unions leverage that information.”
Stratyfy helps financial institutions with risk decisions tied to credit underwriting, credit access, and fraud using cash flow data. It counts Oakland, California-based Beneficial State Bank, Pittsfield, Massachusetts-based Berkshire Bank, Boston-based Eastern Bank and Olympia, Washington-based WSECU, the credit union founded for the state’s public employees, as customers. Kornhauser and Stratyfy COO Deniz Johnson explain in this episode how their firm uses machine learning to interpret data, how they think about customer data privacy and the importance of explainability in their results. They also share the impact they believe cash flow data can have on the financial services space, based on results they’ve already experienced.
Regulators see great promise in cash flow underwriting’s ability to expand credit access and inclusion into the bank space, particularly for individuals with low scores in the traditional credit scoring system or who don’t have enough data to even register for a score. In a recent speech, Federal Reserve Vice Chair for Supervision Michael Barr highlighted the Fed’s 2019 guidance about the benefits and risks of leveraging alternative data in credit underwriting, particularly for borrowers that have been locked out of conventional borrowing.
“With the share of households with bank accounts much larger than the share with credit scores and especially with prime or near-prime credit scores, this information has the potential to allow underwriting of a much larger pool of potential reliable borrowers,” he said.
Previous