How BNPL regulation is forcing UK credit reference agencies to rethink data reporting

Different CRA coverage is nothing new. What’s changed is how closely the FCA is now looking at it. BNPL lenders are operating under CONC 5.2A, while the FCA’s CP26/7 consultation proposes requiring firms that report consumer credit information to one designated CRA to report it to all of them.

For lenders, that makes the current variation in CRA coverage harder to overlook. They’re expected to evidence robust affordability decisions even where some of a customer’s BNPL exposure is outside the bureau file being used.

CP26/7 could change that. Until then, lenders are still making decisions with the coverage available today.

Key takeaways CRA coverage has always varied. With BNPL now regulated and CP26/7 moving forward, that variation has greater implications for affordability decisions. CP26/7 could bring more consistency to bureau files. In the meantime, lenders are still working with the coverage they have today. Missing BNPL exposure can affect what firms see beyond the initial credit decision, particularly where customers are showing signs of financial difficulty or a change in circumstances. Greater use of AI for creditworthiness assessments puts more attention on the data feeding those models and whether it reflects the customer’s actual borrowing. CP26/7 also talks explicitly about competition and potential new data sources, giving alternative data providers a place in the conversation alongside the established CRAs.

Two regulatory tracks, one shared CRA data weakness

BNPL was already a sizeable credit market before the regulation. By 2024, lending had reached roughly £13 billion, with around one in five UK adults using it.

DPC lenders now have more to account for across the data they collect, use and report:

• Creditworthiness: CONC 5.2A applies, including affordability requirements.
• FCA reporting: firms writing more than £2 million in agreements a year are required to submit product sales data.
• CRA reporting: CP26/7 proposes that firms sharing consumer credit information with one designated CRA should share it with all of them.

 

For platforms built around fast, low-friction decisions, the reporting requirement is particularly relevant. Data captured during the customer journey now has uses beyond the lending decision, including FCA reporting and ongoing supervision.

CRA coverage is moving on a different timetable. CP26/7 is still going through the regulatory process, which means lenders are already operating under the new DPC requirements while differences between bureau files remain.

What the US learned from BNPL credit reporting 

The UK isn’t the first market to deal with inconsistent BNPL reporting. The US started bringing BNPL into credit reporting several years ago, but the approach varied between bureaus and providers.

• Experian created a separate file for BNPL accounts rather than immediately incorporating them into its main credit reports.
• TransUnion introduced BNPL tradelines separately from the data feeding standard scoring models.
• The CFPB was raising concerns about inconsistent BNPL reporting in 2022.
• Reporting among BNPL providers remained mixed, with some major lenders choosing not to furnish loan data to the bureaus.

 

Several years on, the US experience gives the FCA’s approach some useful context. CP26/7 proposes a mandatory route, which says firms reporting consumer credit information to one designated CRA would have to report it to all of them.

There’s another source of information available alongside the credit file. Data On Demand’s KYVC vulnerability data can surface more recent signs of financial pressure, including high-cost credit applications, borrowing for essential spending and changes in income or employment.

For BNPL lenders, those signals can add useful context to affordability and vulnerability assessments where bureau data doesn’t yet reflect what’s happening with the customer.

Further reading: Credit Bureau Data vs Alternative Data: What’s the Difference?

The vulnerability blind spot in fragmented CRA data 

Several concurrent BNPL agreements across different providers can be useful context, particularly alongside other borrowing or signs of financial pressure. If those providers report to different CRAs, a single bureau file may only show part of that exposure.

And there may be changes happening around the customer that won’t appear on the credit file straight away. A drop in income, employment change or high-cost credit applications can alter their position well before the bureau data reflects it.

Alternative data can surface some of those changes sooner. Data On Demand’s Know Your Vulnerable Customer (KYVC) provides more recent vulnerability and change-of-circumstance signals that firms can use alongside existing CRA and internal data.

For vulnerability teams, that can mean:

• Seeing changing circumstances earlier, with frequently refreshed data that can pick up changes between bureau updates.
• Identifying financial pressure beyond reported credit exposure, including high-cost credit applications and changes in income or employment.
• Prioritising customers for support sooner, using those signals to inform existing vulnerability processes.

 

For BNPL lenders, those signals can be particularly useful where the bureau file only shows part of a customer’s reported credit activity. They provide more current evidence of how that customer’s circumstances are changing alongside it.

Further reading: What FCA-aligned vulnerability identification looks like in operational teams

What CP26/7 signals for the data market

The FCA’s CP26/7 proposals go further than addressing inconsistent CRA coverage.

Two remedies stand out:

• Remedy 2A: wider data sharing. Once a firm chooses to report consumer credit information to one designated CRA, it would be required to share that information with all designated CRAs.
• Remedy 2D: data accuracy. Firms would face stronger requirements around the accuracy of credit information and the correction of disputed data.

 

The FCA is also working on a new Credit Information Governance Body. As Alison Walters, Director of Financial Services at the FCA, has said, the proposals are intended to give lenders better information when making lending decisions and make it easier for consumers to correct errors on their credit files.

New sources of credit information

CP26/7 also talks about increasing competition and supporting new sources of credit information. For lenders already using alternative consumer data, that part of the consultation is worth noting. Data On Demand’s Know Your Vulnerable Customer (KYVC) data can supplement CRA and internal information with signals around:

• changes in income and employment
• financial pressure
• vulnerability and changing customer circumstances

 

Firms can use those signals within existing credit, affordability and vulnerability processes, including where customer circumstances change between bureau updates.

Even with more consistent CRA reporting, the bureau file won’t contain every change in a customer’s circumstances. Alternative data gives lenders access to some of that information as it changes.

Further reading: How to combine bureau and alternative data without rebuilding your whole strategy

What fragmented CRA data means for AI decisioning 

CRA coverage also has implications for automated affordability decisions and model validation.

The FCA expects DPC lenders to make greater use of AI and open banking data as BNPL volumes grow. Firms using these technologies will need to test and track whether their models support responsible and sustainable lending outcomes.

A model can perform well across a portfolio even when individual customer records don’t contain all of their reported borrowing. If BNPL activity sits with a different bureau and isn’t present on the CRA file used to train, validate or run the model, an existing commitment that could affect affordability may be absent from the decision.

For model validation teams, there are three areas worth looking at:

1. Coverage: whether the data captures the customer’s existing credit commitments across the sources being used.
2. Currency: what may have changed since that information was last updated.
3. Customer circumstances: whether there are signs of financial pressure outside reported credit activity.

 

The last point is already being addressed through a broader mix of data. Open banking can provide current transaction information, while alternative consumer data can pick up changes in income, employment and financial pressure that haven’t appeared in bureau data. Data On Demand’s Know Your Vulnerable Customer (KYVC) provides these types of signals, so they can be considered alongside CRA, internal and open banking data when assessing affordability.

CP26/7 should improve consistency in CRA coverage if the proposals are implemented. The FCA has proposed a 12-month implementation period following publication of its Policy Statement. Until then, DPC lenders will continue making affordability decisions using the CRA coverage available today.

Don’t wait for CRA reporting to catch up

CP26/7 could make BNPL reporting more consistent. With a proposed 12-month implementation period following publication of the FCA’s Policy Statement, lenders will still be making affordability decisions and validating models against today’s CRA coverage for some time.

Changes in income and employment, high-cost credit applications and signs of financial pressure are already available outside traditional bureau data. For lenders reviewing the strength of their affordability data, those signals are worth testing against the customer records they already hold.

A Data On Demand Proof of Value does exactly that. Using a sample of your own customer records, alternative consumer data is assessed alongside the CRA and internal data already used by your teams.

You’ll see:

• which additional customers and signals are identified when alternative data is included
• how early changes in customer circumstances can be picked up
• where those signals affect affordability and vulnerability assessments
• what the results look like against the measures your teams already use

 

You can assess the impact using your own portfolio before making any changes to existing data sources, systems or processes.

Book a Data On Demand Proof of Value and see what the results look like on your portfolio.