Buy Now, Pay Later, the "pay in four" option at checkout that splits a purchase into four interest-free installments, gets described two completely opposite ways, and the reason is not spin. It is that the honest answer depends on which of three numbers you look at, and those three numbers were built to disagree.

Line them up, because almost no coverage does:

All three are real. The first two measure loans that were never repaid at all, the rare, serious failures. The 47% measures how many people missed a payment, even by a few days. That is why the industry can honestly call BNPL low-risk while nearly half of users have personally paid late. And the gap is not static: self-reported late payment has climbed 13 points in two years, which is the trend line that actually matters.

The saving grace inside that scary number: about 72% of late payers were roughly a week late at most. So the typical "late payment" is a minor slip, not a default. That is the whole BNPL story in miniature. Widespread small stumbles, rare catastrophic falls.

The trap isn't the loan. It's the stack.

Here is the mechanism that turns a harmless four-payment plan into genuine trouble, and it is the thing single-loan statistics are structurally blind to.

One pay-in-four plan is easy to track. The problem is that BNPL lenders generally do not know whether you have loans from other BNPL lenders, and until very recently reported almost nothing to the credit bureaus. That creates "loan stacking": a shopper can run four or five simultaneous BNPL plans across Affirm, Klarna, Afterpay, and PayPal, each individually small and manageable, that together add up to a payment load nobody, not even the lenders, can see in full.

This is why BNPL feels frictionless in a way that is precisely the danger. A credit card shows you one climbing balance. BNPL fragments the same spending into a dozen separate little schedules with different due dates, which makes the total remarkably easy to lose track of. The convenience and the risk are the same feature. It doesn't feel like debt, which is exactly what makes it easy to accumulate more of than you meant to.

Who actually gets hurt

The data draws a fairly clear line between the users for whom BNPL is fine and the ones for whom it is a warning sign.

The Richmond Fed's read is reassuring at the macro level: BNPL reached about $70 billion in 2025 but is still only 1.1% of credit card spending, and at that scale its impact on financial stability appears limited, with no clear evidence it causes higher unsecured debt. So this is not a systemic crisis, and for a financially stable person splitting a planned purchase, it is a genuinely useful, interest-free tool.

But the CFPB's profile of who uses it is where the caution lives. BNPL borrowers are more likely than non-users to show signs of financial distress, more likely to carry revolving credit card balances, to be highly indebted, and to also use payday loans, pawn loans, and overdrafts. Eighteen percent of BNPL borrowers had a delinquency on another account, versus 7% of non-borrowers.

The crucial honesty here, which the CFPB states plainly and most hot takes ignore: the data cannot tell you which way the causation runs. It is unclear whether BNPL use leads to distress, or whether people already in distress reach for BNPL to stretch thin budgets. Probably both, in a feedback loop. Either way, the practical signal is the same: if you are turning to BNPL because you cannot afford the thing outright, that is the situation where it stops being convenience and starts being a symptom.

What changes in 2026: your BNPL habits start counting

A significant shift is underway, and it cuts both ways.

For years, BNPL was a credit blind spot: lenders mostly didn't report it, so it neither helped nor hurt your credit score. In 2026, that is changing. BNPL repayment behavior is beginning to appear on credit reports and in FICO scoring.

For responsible users, this is an opportunity. Many BNPL users are younger with thin credit files, and on-time BNPL repayment can now help them build credit faster. For everyone else, it is a new exposure: the late payments that used to vanish quietly can now ding your score, and the loan stacking that was invisible may start to show. The same behavior that was consequence-free is about to have consequences in both directions. Note the wrinkle, though, that lenders have been ambivalent about handing pay-in-four data to the bureaus, so the rollout is uneven and worth watching rather than assuming.

So: convenient tool or debt trap?

It is genuinely both, and which one you get is close to fully determined by how you use it. That is not a cop-out; the data supports a specific dividing line.

BNPL is a convenient tool when you are splitting a purchase you could already afford, into a short, interest-free schedule you will not lose track of, ideally one plan at a time. Used that way it is often better than a credit card, because there is no interest and, done right, it can now build credit.

BNPL becomes a debt trap when it is the reason you can afford the purchase, when you are running several plans at once, or when you are using it alongside other high-cost credit to keep your budget afloat. In that mode the frictionlessness works against you, the stacked payments are hard to see, and a "harmless" week-late slip is now the leading edge of a score-damaging pattern.

The clean test is a question the checkout button is designed to stop you from asking: if you had to pay the full price today, in one payment, would you still buy this? If yes, BNPL is a fine way to spread it out. If the four payments are the only reason it feels affordable, the tool is not solving your problem.

It is postponing it, three payments at a time.

Further reading