The consumer lending industry is abuzz about the Federal Reserve Bank of Cleveland’s recent report on debt consolidation and online lending. This excellent piece of research concludes that, on average, online installment loan borrowers fall into more debt after taking out a loan, experience hits to their credit score and history as a result, and take out online loans despite having access to traditional banking and credit channels.

The first two conclusions are damning, especially as these loans are often marketed as a way to help consumers consolidate credit card debt and improve their finances. At the end of the day, a lender’s duty is not merely to avoid losses. Any loan must be suitable for the customer — which means it should be made only if the lender believes it is improving the customer’s financial health. A lender not guided by that principle should be prepared for severe criticism as well as elevated losses down the road. What is particularly concerning is the evidence that the credit characteristics of online installment borrowers at the time of repayment are consistently worse than they were at the time of borrowing. That should be a sobering thought for online lenders whose credit models have not been fully tested in a protracted credit downturn.

The study examined three claims sometimes made by online lenders and their advocates: that P2P loans allow borrowers to refinance or consolidate credit card debt, help consumers improve credit scores, and widen access to credit for the underbanked. The authors found the opposite to be true in each case.

Fintech lending advocates — led by the Marketplace Lending Association — counter that the loans examined in the study included relatively few originated by fintech “marketplace” lenders, despite the study’s use of the term “P2P” to describe the loans included in the study, and that the conclusions of the study thus don’t apply to fintech loans. This may be true. The Cleveland Fed has since acknowledged that its use of that term might not accurately reflect the types of loans and lenders examined in the study. The Cleveland Fed research, the MLA contends, was based largely on data about installment loans made by lenders other than fintechs, such as more traditional finance companies. TransUnion has also raised questions about how the Cleveland Fed characterized the loan data provided by the credit bureau and used in the report.

But it would be nonsensical to discredit or ignore the study because of these concerns. The study has been carefully designed and executed, using “control” consumers as well as online borrowers as the statistical basis for its analysis and conclusions, which appear well-supported by the data presented. And consumer finance lenders, whether they are called P2P lenders, online lenders, marketplace lenders or something else, are all in the same business and often use the same marketing and delivery channels. All sorts of lenders originate installment loans through direct mail using algorithmic scoring and fulfill their loans online.

Instead, we should ask every online lender —whether a fintech-type marketplace lender or not — included in the study to show the public, capital providers and regulators why the study’s conclusions don’t apply to its loans and borrowers. Lenders should use their own data to try to rebut the study’s two conclusions that: (1) that borrowers were, on average, increasing other leverage (particularly cards) after getting an online installment loan instead of de-levering, and (2) that borrowers' FICO scores and other credit characteristics were, on average, deteriorating after taking out a loan. If a lender can't rebut these assertions, then we can reasonably assume that the study’s conclusions apply to its borrower base and that it is, on average, generating riskier loans than advertised while damaging the customers it purports to be helping. While we would normally expect lenders that charge higher rates to borrowers with lower credit scores to perform relatively poorly, and lenders that charge lower rates to borrowers with high credit scores to do better when compared to the study sample, there could be plenty of surprises. If, for example, the decision to take out an online installment loan is itself a sign of financial stress, the problem could be widespread among lenders.

Independent of the study itself, it should not come as a surprise that online “debt consolidation” loans fail to reduce a consumer’s debt. By design, they can be used for any purpose at all. Lenders make no effort to force the borrower to use loan proceeds to reduce other debt and, even if the borrower pays down maxed-out credit cards with loan proceeds, that only means that the borrower now has a large new “open to buy” position on his or her cards. Given this dynamic, it is also hardly surprising that many borrowers run into credit issues after re-levering with credit cards or other borrowing.

As a society, we really should know which online lenders are adding to consumer financial health and which ones are detracting from it. And why stop there? This is not an online-only issue. Financial health questions like this should be asked by regulators during any bank, credit union or finance company examination. Online lenders tout their products as an upgrade in credit options for a diverse borrower base. But let’s hope that the Cleveland Fed study forces them to think more about their customers’ overall financial health, like doctors who make this oath: “First, do no harm.”

Updated November 16, 2017 at 4:51PM: This op-ed has been updated to reflect uncertainty over how the Cleveland Fed used data provided by TransUnion in the study, as well as which specific types of institutions were included in the data.

Todd H. Baker

Todd H. Baker

Todd H. Baker is a senior fellow at the Mossavar-Rahmani Center for Business and Government at the Harvard Kennedy School, and managing principal of Broadmoor Consulting LLC.

BankThink submission guidelines

BankThink is American Banker's platform for informed opinion about the ideas, trends and events reshaping financial services. View our detailed submission criteria and instructions.