Why Making Banking Data Portable Isn’t Always Good for Borrowers
In theory, rules requiring banks to share consumer data with third parties increase competition and help consumers. In practice, it’s not so simple, according to a new study co-authored by Yale SOM’s Jidong Zhou.
After decades of dominating the global credit industry, big banks are finding themselves challenged by the fintech lending firms entering the market. In theory, increasing competition by requiring banks to share their customer data with fintech challengers should be good for consumers. However, new research from Yale SOM paints a more complicated picture.
Big banks have long enjoyed one major advantage when making lending decisions: data, and lots of it. For example, if a customer who already has a checking account with Chase applies for a loan, Chase doesn’t have to rely only on their credit report or FICO score. The bank can examine years of transaction history to determine if the customer is creditworthy.
Traditional banks’ hold on customer data has made it hard for upstart fintech lenders to compete. While these challengers may use more sophisticated algorithms to make lending decisions, an algorithm is only as good as the data that goes into it. Without access to the information held behind banks’ locked gates, fintech disruptors will always be one step behind.
In part to level the playing field, regulators in the European Union and many other countries have passed laws requiring banks to share data with third parties such as fintech lending firms when consumers request it. The United States takes a more laissez-faire approach. While banks sometimes share customer data with third parties, they aren’t required to do so. Your bank will probably allow you to share your checking account information with a budgeting service such as Mint, but it will probably balk at giving the same information to a competing lender.
On paper, the European approach looks better for consumers. By making it easier for fintech firms to compete with banks, the thinking goes, open banking rules give borrowers more and better options. But in their new paper, Yale SOM’s Jidong Zhou, along with Zhiguo He of University of Chicago and Jing Huang of Texas A&M University, outline some unexpected downsides to open banking.
“The initial motivation of open banking is to increase competition, and usually we believe that more competition is better for consumers,” explains Zhou. “Here, we tried to point out that it’s not as simple, and in some circumstances [open banking] may backfire.”
The researchers developed several mathematical models that helped them investigate the effects of open banking rules. First, they modeled a setting without open banking. Under this scenario, big banks have more data and therefore more ability to screen borrowers, separating those likely to pay back their loans from those likely to default. Fintech firms, despite their sophisticated algorithms, have a much tougher time figuring out who is creditworthy and therefore struggle to compete. As a result, the model shows, “banks will earn higher profit than the fintech lenders, and the banks will also charge a higher interest rate,” Zhou explains—not great for consumers.
Now imagine the (admittedly unlikely) hypothetical scenario on which banks had to share their data with fintech firms whether consumers wanted them to or not. “Now the situation can be reversed,” says Zhou. “with the same or even better data—the fintech lenders often also have some independent data sources such as digital footprints—and more advanced algorithms, the fintech lenders can become much more powerful than the traditional banks.” Once again, a situation that is not great for consumers.
While this is an extreme hypothetical, it illustrates a key point. The crux of the issue is what the researchers call the “screening ability gap” between banks and fintech firms. When sharing banks’ data shrinks the gap, competition will indeed intensify and consumers benefit. However, if it gives too much power to the fintechs, the banks will struggle to compete and consumers pay the price.
Finally, the researchers modeled the scenario that exists in Europe and many other countries today, where consumers can opt to share their data with fintech firms or not. This scenario affects different borrowers in different ways, they found.
Borrowers who opt to share their information will likely be perceived by lenders as highly creditworthy—because why would they share information unless it made them look good? Borrowers who opt not to, whether it’s because they genuinely aren’t creditworthy or simply because they are privacy conscious, risk appearing to lenders like a bad bet.
For borrowers who opt to share, the situation starts to resemble the researchers’ extreme hypothetical: here, the fintech lender has lots of data and better tools to analyze it, resulting in a possibly increased screening ability gap and reduced competition. Borrowers who opt not to share will be perceived as less creditworthy, and therefore have fewer loan options and higher interest rates.
“Both types of consumers are suffering from open banking,” says Zhou, “but they are suffering via different channels.”
Does this mean we should abandon open banking altogether? Not at all, Zhou says. Open banking backfires only when it makes the fintech lenders too powerful. Plus, some of the consumer downsides the research identifies are likely short-run issues: over time, open banking could spur traditional banks to improve their analytical tools, once again closing the screening ability gap between banks and fintech firms. Rather, the point of the paper is that open banking does have short-run consequences worth considering.
Ultimately, it’s probably too soon to say whether open banking is mostly good or mostly bad for consumers and what regulatory approaches work best. Perhaps countries like the U.S. will develop novel systems for sharing consumer data—for instance, ones in which banks sell data to third parties for a fee. Perhaps regulators will find ways to protect privacy-conscious consumers from negative inferences about their creditworthiness.
So, for now, the jury is out. “Different countries are experimenting with different approaches,” says Zhou. “In the future, perhaps we will be more able to tell which ones work best.”