Credit files omit most BNPL loans held by the 16% of consumers using them

The Federal Reserve's survey also describes who holds the loans the bureaus miss. Users were more likely to have lower incomes, lower credit scores, trouble repaying debt or revolving credit card balances ^[7]. Most of that already shows up in an application, because scores and revolving balances sit on the bureau file and get priced. The applicant whose decision can change is the one whose file looks clean because it is short. Young adults were more likely to use BNPL ^[8]. Anna Kooi, a partner at Wipfli, said that until the loans are reported, "banks should assume debt-to-income is understated for a meaningful share of applicants, particularly younger and thin-file borrowers" ^[11].

How big the gap is depends on which bank is looking. "Banks already have the data to see BNPL behavior in their customers' accounts," Kooi said ^[12]. The bank holding the checking account sees the recurring payments to BNPL providers ^[12]. From that ledger it can check whether a customer runs several plans at once, whether the payments set off overdrafts or nonsufficient-funds charges, and whether the spending has moved into essentials ^[13]. A card issuer lending to someone who banks elsewhere has only the bureau file, and most of these loans are not in it ^[2]. The bureau route costs that issuer nothing extra. The deposit route costs analyst hours, and only the lender with the primary checking relationship can take it.

The gap could close from the bureau side. Kooi framed her advice as lasting until the loans are reported ^[11]. If providers start furnishing them, every lender sees the same file and the deposit holder's edge goes away. The balances may also be too small to matter. "Most pay-in-four loans don't appear on a credit report," Kooi said ^[4], and if the typical plan is a few installments on a modest purchase, a missing one may barely move a debt-to-income ratio. The article does not report outstanding BNPL balances, so how far debt-to-income is understated is an assumption, not a measurement.

I think the risk is real and sits mostly in thin files, where the record is short enough that one missing obligation changes the picture. Loan performance would prove this wrong. If thin-file BNPL users default on bank cards at the same rate as non-users with the same scores, the missing loans were not changing outcomes. "Every credit cycle has a blind spot, and BNPL is this one's," Kooi said ^[14].

Distribution is still widening. Tanuj Parikh, head of revenue for Cash App and Afterpay, said in a statement that being picked as Fanatics' preferred BNPL partner "is a meaningful milestone for Afterpay as we continue expanding into high-intent shopping categories where flexibility matters" ^[17]. Daniela Hawkins, a partner at Capco, told American Banker: "The whole economic system is driving people to use these products, which is ultimately creating more risk in the market" ^[16].