Six major BNPL lenders originated 335.8 million loans worth USD 45.2 billion, at an average of USD 135 each, and between 34% and 41% of users report having made at least one late payment. That last pair of numbers is the tension in this market: formal delinquency looks benign, with 4.1% of loans drawing a late fee and 1.83% charged off, while self-reported payment difficulty runs an order of magnitude higher. BNPL remains roughly 1% of credit card spending volume, so the systemic question is not size but whether the hard data is measuring the right thing. The figures below come from CFPB market reporting and survey work.
TL;DR
- Six major BNPL firms originated 335.8 million loans (CFPB)
- Those loans totalled USD 45.2 billion (CFPB)
- The average loan size was USD 135 (CFPB)
- 53.6 million consumers took out at least one BNPL loan (CFPB)
- That was a 12% increase on the prior year (CFPB)
- 4.1% of loans were assessed a late fee (CFPB)
- That is down from 5.2% the prior year (CFPB)
- Charge-offs fell to 1.83% from 2.63% (CFPB)
- Between 34% and 41% of users report at least one late payment (CFPB survey)
- BNPL is roughly 1% of total credit card spending volume (CFPB)
- The CFPB notes indicative evidence of worsening delinquency since (CFPB)
- The regulatory treatment of BNPL has shifted and remains unsettled
- Detailed origination data lags the current market by more than a year
1. The Market in Numbers
The origination figures describe a product built on volume rather than value. Six major BNPL companies originated 335.8 million loans totalling USD 45.2 billion, at an average of USD 135 per loan. Hundreds of millions of transactions producing under fifty billion dollars is a distribution that no conventional credit product matches.
That structure has consequences for how the product should be assessed. Underwriting cost per loan has to be near zero for the economics to work, which is why BNPL approvals are near-instant and shallow. It also means a lender’s exposure to any individual borrower is small while its exposure to a general deterioration in consumer conditions is spread across an enormous number of accounts. The risk profile is closer to a payments business with credit attached than to a lending business, which is part of why the regulatory classification has proved contentious.
| Metric | Value | Source |
|---|---|---|
| Loans originated by six major firms | 335.8 million | CFPB |
| Total origination value | USD 45.2 billion | CFPB |
| Average loan size | USD 135 | CFPB |
| Consumers with at least one loan | 53.6 million | CFPB |
| Year-over-year change in consumers | up 12% | CFPB |
| Implied loans per consumer | approx. 6.3 | Derived from CFPB figures |
| BNPL share of credit card spending volume | roughly 1% | CFPB |
| Firms covered by the detailed data | 6 | CFPB |
Payment-method context sits in our digital wallet statistics. Source: CFPB, the Buy Now Pay Later market.
2. Late Payments Versus Defaults
This is where the dataset gets genuinely interesting, because two credible measures disagree by roughly a factor of nine. 4.1% of loans were assessed a late fee and 1.83% were charged off, while CFPB survey work found 34% to 41% of users reporting at least one late payment.
Both can be true simultaneously. The loan-level figures count individual loans, most of which are repaid on time; the survey figures count people, and a user with a dozen BNPL loans in a year has many opportunities to be late on one. A 4.1% per-loan late rate applied across six loans per user produces exactly the kind of user-level incidence the survey found. Understood that way, the numbers are not in conflict, and the survey figure is the better guide to how common the experience of falling behind actually is.
| Metric | Value | Source |
|---|---|---|
| Loans assessed a late fee | 4.1% | CFPB |
| Same figure the prior year | 5.2% | CFPB |
| Loans charged off | 1.83% | CFPB |
| Same figure the prior year | 2.63% | CFPB |
| Users reporting at least one late payment | 34% to 41% | CFPB survey |
| Implied loans per user | approx. 6.3 | Derived from CFPB figures |
| Ratio between user-level and loan-level incidence | roughly 9 to 1 | Derived |
| Whether the two measures conflict | no, different denominators | Derived |
Source: CFPB, consumer use of Buy Now Pay Later.
3. The Hard Data Is Improving, the Soft Data Is Not
Both formal risk measures moved the right way. Late fee incidence fell from 5.2% to 4.1% and charge-offs from 2.63% to 1.83%, which on its own would describe a maturing market with improving underwriting.
The CFPB attaches a caveat that undercuts the reassurance: there is indicative evidence from consumer surveys that late payment and delinquency behaviour has increased since the period covered by the detailed data. That leaves the sector in an awkward evidentiary position where the rigorous numbers are old and improving while the timelier numbers are softer and worsening. Anyone citing the 1.83% charge-off rate as evidence that BNPL is low-risk should note that it describes a period now well behind us.
| Metric | Value | Source |
|---|---|---|
| Late fee incidence, latest detailed year | 4.1% | CFPB |
| Prior year | 5.2% | CFPB |
| Change | down 1.1 points | Derived from CFPB figures |
| Charge-off rate, latest detailed year | 1.83% | CFPB |
| Prior year | 2.63% | CFPB |
| Change | down 0.8 points | Derived from CFPB figures |
| Direction of survey-based evidence since | worsening | CFPB |
| Tension between the two | hard data older, soft data timelier | Derived |
Household credit context sits in our student loan debt statistics. Source: CFPB BNPL market report.
4. Scale Against the Rest of Consumer Credit
Context prevents both overstatement and complacency. BNPL represents roughly 1% of total credit card spending volume, which means that as a share of consumer credit it remains marginal even after years of rapid growth.
The systemic risk argument therefore does not rest on size. It rests on stacking and invisibility: BNPL loans have historically not been reported to credit bureaus consistently, so a borrower can hold several simultaneously without any single lender, or any conventional underwriter, seeing the full picture. A product that is 1% of card volume but invisible in credit files can still distort affordability assessments for the other 99%. That is the mechanism worth watching, and it is precisely the one the available data cannot quantify.
| Metric | Value | Source |
|---|---|---|
| BNPL share of credit card spending volume | roughly 1% | CFPB |
| Total BNPL origination value | USD 45.2 billion | CFPB |
| Consumers using the product | 53.6 million | CFPB |
| Average loan size | USD 135 | CFPB |
| Loans per consumer | approx. 6.3 | Derived |
| Primary systemic concern | stacking across lenders | CFPB |
| Historic credit bureau reporting | inconsistent | CFPB |
| Whether current data quantifies stacking | no | Derived |
Merchant-side context sits in our live shopping statistics. Source: Richmond Fed on recent BNPL developments.
5. Why This Data Is Thinner Than It Should Be
BNPL is the largest consumer credit category with the weakest public statistics, and the reason is regulatory rather than technical. The CFPB moved to treat BNPL under credit card rules and subsequently stepped back from enforcing that approach, leaving supervision unsettled and mandatory reporting undefined.
The practical result is that the best available figures come from a voluntary market-monitoring exercise covering six firms, published irregularly, describing a period that recedes further into the past with every quarter that passes without an update. For a product used by more than fifty million people, that is a thin evidentiary base, and it is why any BNPL statistic should be read with its vintage attached rather than as a description of the current market.
The gap also shapes what circulates. Where official data is sparse and dated, market research estimates and lender press releases fill the vacuum, and those figures are constructed to different standards with no obligation to disclose methodology. Much of the BNPL commentary in circulation traces back to sources of that kind rather than to the supervisory data used here. Preferring a slightly older number with a documented methodology over a fresher one without is the right trade in this category, which is not a conclusion that would hold in most other consumer credit segments.
| Metric | Value | Source |
|---|---|---|
| Firms covered by CFPB detailed data | 6 | CFPB |
| Nature of the data collection | market monitoring, not mandatory reporting | CFPB |
| Publication cadence | irregular | Derived |
| Regulatory status | unsettled | CFPB |
| Consumers affected | 53.6 million | CFPB |
| Loans originated | 335.8 million | CFPB |
| Lag between data period and now | more than a year | Derived |
| Consequence | statistics describe a past market | Derived |
Fraud-side context sits in our online scam statistics. Source: Analysis of the CFPB’s BNPL regulatory approach.
Summary: BNPL by the Numbers
| Metric | Value | Source |
|---|---|---|
| Loans originated by six major firms | 335.8 million | CFPB |
| Total origination value | USD 45.2 billion | CFPB |
| Average loan size | USD 135 | CFPB |
| Consumers with at least one loan | 53.6 million | CFPB |
| Year-over-year change in consumers | up 12% | CFPB |
| Implied loans per consumer | approx. 6.3 | Derived |
| Loans assessed a late fee | 4.1% | CFPB |
| Prior year late fee rate | 5.2% | CFPB |
| Charge-off rate | 1.83% | CFPB |
| Prior year charge-off rate | 2.63% | CFPB |
| Users reporting a late payment | 34% to 41% | CFPB survey |
| Ratio of user-level to loan-level incidence | roughly 9 to 1 | Derived |
| BNPL share of card spending volume | roughly 1% | CFPB |
| Direction of survey evidence since | worsening | CFPB |
| Firms in the detailed dataset | 6 | CFPB |
| Regulatory status | unsettled | CFPB |
Methodology and Sources
- Origination counts, loan values, average loan size, consumer counts, late fee incidence, and charge-off rates come from the Consumer Financial Protection Bureau’s BNPL market reporting (market overview, market report PDF, earlier consumer use report).
- User-level late payment incidence comes from the CFPB Making Ends Meet survey (CFPB).
- Market context and regulatory analysis come from the Richmond Fed and legal commentary (Richmond Fed economic brief, University of Baltimore Law Review, Consumer Financial Services Law Monitor).
- Data watch: the detailed origination and delinquency figures cover six major lenders and a reporting period now more than a year in the past, and the CFPB has itself flagged indicative survey evidence that delinquency has worsened since. Treat the improving hard numbers as historical rather than current. The six-firm scope excludes smaller providers, retailer-embedded programmes, and card-issuer instalment features, so market totals understate the category. Loan-level and user-level incidence use different denominators and the roughly ninefold gap between them is arithmetic, not contradiction. BNPL reporting to credit bureaus has been inconsistent, so no dataset reliably captures borrowers holding multiple loans across lenders, which is the principal consumer risk. Regulatory treatment has shifted and no mandatory reporting regime exists. Rows marked as derived are arithmetic on published figures.
- Last updated: August 2, 2026. We update this roundup quarterly as the CFPB and Federal Reserve banks publish new BNPL analysis.