Can credit unions use artificial intelligence, machine learning, or business analytics — whatever you want to call it — to create real-time debt alerts that inform a careful young spender pre- and post-purchase how said purchase will impact their budget?
No speaker addressed that exact idea, but in a session titled “How is AI Being Used Right Now and What is the Upside for Lenders” panelists did stress that vast stores of data combined with the willingness to consider nontraditional measures of risk present financial institutions the ability to better compete in a crowded, disrupted market.
That’s especially effective and growing in importance as young Americans and established career professionals from abroad present the problem of “thin files” when it comes time to rate their risk.
“Traditional thin files are being scored by prohibitive models,” said panelist Nick Rose, senior scientist of VantageScore. “We’re seeing growing use of machine learning to deconstruct and rebuild credit reports that can better tell us what we should be looking at.”
Lenders should be ready to explain that to examiners, too, as federal regulators are also working to understand emerging ways of rating risk and granting credit.
“There’s a willingness to see how this kind of math is used in underwriting,” said Kareem Saleh, executive vice president at ZestFinance, who says he was in Washington, DC, recently working with regulators on that issue.
“It’s ultimately about fairness,” he said. “It’s also the 800-pound gorilla in the room that we have to keep in mind as we learn about how different factors interact and relate and how we can act on them.”
Saleh mentioned the Office of the Comptroller of the Currency and the Federal Reserve as the agencies he visited. The NCUA, as a member of the FFIEC with those regulators, typically pays close attention to developments among its peers and often follows their lead.
Panel moderator Penny Crosman of American Banker said there are predictions that AI will displace 250,000 loan officers by 2030. Does that mean business intelligence specialists will data analyze themselves out of jobs? It doesn’t have to be.
Saleh, the ZestFinance executive, responded: “You can put the people you have to better use. You can focus on building better predictions rather than cleaning up data.”
Some bench strength could help. Credit unions are affected by the same economic realities as all financial institutions, and right now there’s a growing possibility of continuing squeeze on the profitability of credit cards, typically a financial cooperative’s most profitable product.
Rob Mau, a partner with Oliver Wyman, said in a morning session that his research firm has found interest margins are trending down while issuers are spending more on rewards. In fact, rewards costs are currently growing 14% a year.
His call to action: innovation framed by lessons from big tech. He illustrated his example by recounting how travel assistance has evolved from the basic map to the AAA TripTik to the Garmin to GPS on smart phones to live traffic updates and rerouting on those same phones.
Ironically, perhaps, two of the scheduled morning speakers were not able to attend because of flight issues. So, ultimately, it still comes down to people. You need people.
Want more credit union strategies? Sign up for the CreditUnions.com free newsletter.