The Financial Conduct Authority says the share of UK credit card accounts in persistent debt fell from 7% in 2018 to 3% by early 2023, after rules that made lenders contact customers who pay more in interest and charges than they repay. The regulator does not credit the rules alone, and its own early look at newer data shows the share rising again.
What the FCA tested
In November the FCA published Evaluation Paper 26/1, a look back at the persistent debt remedies it set in policy statement PS18/4. A customer is in persistent debt when the balance is above £200 and, over an 18-month period, the customer has paid more in interest, fees and charges than toward the principal.
The rules work in steps. At 18 months, a lender contacts the customer and prompts a change in repayment behaviour if the customer can afford one. At 27 months, a lender sends a reminder if it looks likely the customer will still be in debt at 36 months. At 36 months, the lender proposes ways to repay over what the FCA calls a reasonable period, usually three to four years, and shows forbearance where the customer cannot pay faster. Customers who have been shown forbearance, or who do not respond, have their cards suspended.
The evaluation covers January 2017 to August 2023 and draws on application and statement data from ten large card lenders. It sits in a large market: the paper cites UK Finance figures of 38.5 million active credit card accounts and more than £72 billion in outstanding balances as of February 2026.
What changed after the letters went out
The headline result is the drop from 7% to 3% of accounts, which the paper puts at a fall of around 57%. The FCA’s blog post of November 18 adds the customer-level detail. In the six months after receiving a first persistent debt letter, over half of customers raised their average repayments, around two-thirds reduced their average debt, and a majority cut the share of their available credit they were using.
Exits followed. Of the customers who received an initial letter, more than half had left persistent debt within 18 months, and almost two-thirds within 27 months. Monthly interest and fees for customers in persistent debt fell over the period, and the share of consumers paying fees decreased.
Reach was partial. Where lenders could monitor it, 50% to 60% of customers opened electronic communications, so between 40% and 50% of those letters went unopened.
Why the FCA will not claim the whole drop
The evaluation period includes the Covid-19 pandemic, the cost of living crisis and a run of rising interest rates. The paper says these “undoubtedly affected” consumers’ use of credit, so the persistent debt rules cannot be credited with the entire fall. The decline also took longer than the two years its original cost-benefit analysis assumed.
The FCA calls its findings “indicative rather than precise”. Data quality problems required analytical adjustments and assumptions, and some inconsistencies in lenders’ submissions remain unresolved; the regulator judged that further work to resolve them would be disproportionate. The paper says it could not assess whether lenders changed who they lend to, how much, or at what price in response to the rules, so any unintended effect on credit supply is unmeasured.
What compliance cost
Lenders’ post-implementation compliance costs came to an estimated £281.6 million, slightly below the final pre-intervention range of £294.6 million to £334.1 million. That estimate rests on self-reported data from seven firms representing around 43% of the market, scaled up to the full market. The FCA reads the result as costs within the range available before implementation.
Where letters stop working
The paper finds the remedies less effective for customers who cannot afford to repay faster, and a significant minority remained in persistent debt 27 months after the first communication. Alison Walters, Director of consumer finance at the FCA, wrote in the blog post that “lenders need to do more than send communications”, pointing to the earlier intervention rules and the Consumer Duty, which require firms to support customers who show signs of financial difficulty and to act to prevent foreseeable harm.
Newer data points the other way. Walters wrote that credit card use and defaults have increased since the period the evaluation covers, and that initial analysis of new product sales data suggests rates of persistent debt have risen, although they remain below the level at which the rules were introduced. The FCA says that product sales data will let it monitor persistent debt levels going forward.
What it means for the finance leader
The paper’s lessons for lenders are practical. In rare cases lenders applied the persistent debt definition inconsistently, including how refunds count as negative transactions when calculating repayments and how missing statement data is handled. The FCA suggests that talking such cases through with supervisors, and sharing practice through trade associations, could make outcomes more consistent, since lenders’ varied approaches to supporting customers led to different results.
In our reading, three jobs follow for a card issuer’s finance and risk teams. First, the 18-month test is a calculation over statement data, so refunds, missing statements and balance thresholds need one written rule that data engineering and compliance both sign off. Second, the 50% to 60% open rate makes channel data a control: a lender that tracks who opened which letter can see who has not engaged before the 27-month step arrives. Third, the customers still in persistent debt after 27 months are the ones the Consumer Duty reaches, which puts hardship routing and forbearance tooling in the same budget conversation as letter templates.
Two earlier pieces cover neighbouring ground. In early November, FCA Supports Car Finance Settlement Offers; Barclays Partner Finance Will Make Some covered what the regulator expects from lenders offering redress. Banks Adding Earned Wage Access Should Test It Against the CFPB’s Four Criteria looked at a related consumer-credit product, earned wage access.
For a head of finance at a UK card lender, the practical step this quarter is to pull the persistent debt population from the latest statement data, compare it with the 3% early 2023 level the FCA reports, and take the result to the board alongside the lender’s own hardship numbers.
Source: FCA Evaluation Paper 26/1
Source: FCA


