Your Four Payments Are a Loan Now

A Product Designed Not to Be a Loan

The standard buy now, pay later product splits a purchase into four payments across six weeks and charges the consumer nothing. It was built that way deliberately. Four installments over less than a year, with no finance charge, falls outside the Truth in Lending Act's definition of a closed-end credit transaction requiring disclosure. That structure is not incidental. It is the reason the product exists in its current form.

One consequence of sitting outside that framework was that the obligation did not appear anywhere a lender would look. A consumer could hold six concurrent BNPL plans against six separate purchases and apply for a car loan the same afternoon with a credit report showing none of it. The underwriter saw the applicant's cards and installment loans. The underwriter did not see the six weeks of scheduled payments already committed.

That gap is closing, and it is closing through two separate mechanisms that are worth keeping apart.

The Furnishing and the Scoring Are Different Events

The first mechanism is furnishing. A lender decides to report account data to a credit bureau. Affirm now furnishes to Experian and TransUnion. Klarna furnishes to TransUnion. This is a voluntary act by each provider, made separately for each bureau, and the coverage is accordingly patchy. A consumer's Experian file and their Equifax file may describe two different borrowers.

The second mechanism is scoring. In fall 2025 FICO released two models, FICO Score 10 BNPL and FICO Score 10 T BNPL, the first scoring models built to ingest BNPL tradelines rather than ignore or mishandle them. The distinction matters because furnished data that no model knows how to read is not neutral. It can be read badly.

Both events have to occur for the data to reach a lending decision. The provider has to report to a bureau, and the lender pulling that bureau's file has to be running a model that understands what it is looking at.

Why a Six-Week Loan Confuses a Scoring Model

Conventional installment tradelines run for years. A scoring model reads them through account age, payment history, and the ratio of remaining balance to original amount. A BNPL plan opens, is repaid, and closes inside six weeks, and a consumer may open five in a month.

Fed through a legacy model, that pattern reads as a borrower opening and closing an unusual number of accounts in rapid succession, which is behavior legacy models are built to treat as distress. Average age of accounts falls. New account inquiries accumulate. The signal a legacy model extracts from responsible BNPL use can therefore be the opposite of what occurred.

The BNPL-aware models handle these tradelines as a distinct category rather than forcing them into the installment bucket. FICO's own figure is that for 85 percent of consumers, the presence of a BNPL account moves the score within a band of plus or minus 10 points. That is a narrow band, and it should be read as what it is: a statement about the median case, not a guarantee about any individual file. The 15 percent outside the band are disproportionately consumers with thin files, where a small number of tradelines carries disproportionate weight.

The Thin-File Case Cuts Both Ways

For a consumer with little or no conventional credit history, BNPL reporting is the first opportunity in the product's history to generate a positive payment record without a credit card.

The same mechanism operates in reverse. A missed payment on a $60 purchase, once furnished, is a delinquency on a credit file, and on a thin file it carries the weight of a scarce data point. The provider's late fee, where one exists, is the smaller cost. The reported delinquency is the durable one, and it persists on the file for seven years under the Fair Credit Reporting Act regardless of how quickly the underlying $60 was paid.

This is a real change in the risk profile of the product, and it is not the change the marketing describes. The pitch has always emphasized the absence of interest. The absence of interest was never the exposure. The exposure was the absence of a record, and that has now been removed on the downside while remaining incomplete on the upside, because not every provider furnishes and not every furnisher reports to every bureau.

Adoption Runs on the Lender's Timeline

A new FICO model does not deploy on release. It deploys when the institutions pulling scores decide to migrate, and mortgage lending in particular runs on scoring models several generations behind current release because of the requirements attached to loans sold to Fannie Mae and Freddie Mac.

The practical effect is a staggered transition of indefinite length. During it, the same consumer's BNPL history may be fully visible to a fintech personal lender running the current model, partially visible to a card issuer, and entirely invisible to a mortgage underwriter. None of those institutions is wrong. They are reading different files through different models.

A consumer cannot determine from the outside which is which. What a consumer can determine is what is in the file, and the mechanism for that is unchanged: the free weekly reports at annualcreditreport.com, pulled from all three bureaus rather than one, because the whole point of voluntary furnishing is that the three will not agree. The related question of what the scoring file is actually for is covered in how the score itself is sold.

Utilization, and the Thing That Does Not Show Up

BNPL balances do not enter revolving utilization, because they are not revolving accounts. That is favorable to the consumer and it is worth stating plainly, since the assumption runs the other way.

What does show up is the debt-to-income calculation, and that runs on a different track entirely. Manual underwriting for a mortgage or an auto loan involves a human reading a bank statement. Four recurring debits to Affirm appear on a bank statement whether or not any bureau was told about them. Underwriters have been counting them for some time. Furnishing changes what the automated pull returns. It does not change what a person reading two months of transactions can see, and for the loans where the decision matters most, a person is still reading.

What the Agreement Specifies

Each BNPL provider's terms contain a section on credit reporting, and it states which bureaus receive data and under what conditions. The conditions are the operative part. Several providers furnish only delinquent accounts, or furnish only their longer-term interest-bearing products while leaving the four-payment plans unreported. Under those terms the account can damage a file without any corresponding capacity to build one.

That asymmetry is disclosed. It appears in the terms accepted at checkout, in a section headed credit reporting or information sharing, and it is short. The document says what the account can do to the file in either direction, and for a product that reports delinquencies but not on-time payments, the document is describing an instrument with downside and no upside.

That is the version of the product a consumer should assume they hold until the terms say otherwise.