At Credit Week last month, the FCA’s Alison Walters described the UK consumer credit market as a caterpillar awaiting metamorphosis — fragmented, brittle, ready to be collectively reshaped into something more cohesive and trusted. It’s an optimistic frame. It’s also a diplomatic one, because it glides past the debate the industry actually wants to have.
I know, because I had that debate a few weeks ago — with the Chief Risk and Compliance Officer of a leading bank. Their view, put plainly: there is too much consumer protection in the UK, and it’s facilitating bad actors.
It’s a fair point. And there’s a distinction buried inside it that the whole market needs to get right, and mostly doesn’t:
Consumer protection exists to shield vulnerability. Too often it shields behaviour instead. And the two are not the same.
Vulnerability is circumstance: the borrower hit by illness, job loss, a relationship breakdown — someone whose situation, not their choices, put them in difficulty. Protecting them is the point of the framework, and rightly so.
Behaviour is choice: the strategic non-payer who knows enforcement is slow and reputationally costly for the lender; the speculative claim filed at industrial scale because complaining is free and someone else pays. When protection can’t distinguish between the two, it shields both — and the costs of the second group are priced onto everyone else.
The proof: what happened when complaining stopped being free
If you want evidence that the CRCO is describing something real, look at what just happened at the Financial Ombudsman.
Between April and December 2024, professional representatives were behind almost half — 47% — of the cases sent to the FOS. Yet only around a quarter of those cases succeeded, compared to over a third of complaints brought directly by consumers. Claims firms were flooding the system with poorly evidenced complaints because there was little commercial incentive to ensure the complaints they brought were well-founded — firms paid the case fee win or lose.
Then, from April 2025, the FOS started charging professional representatives £250 per case beyond ten free referrals a year. The result was immediate: complaints about perceived unaffordable and irresponsible lending fell by around 50% in a single quarter, and irresponsible-lending complaint volumes ultimately dropped by around 80% — driven almost entirely by the charging structure.
Read that again. The moment a modest cost was attached, the majority of that complaint volume evaporated. Genuine vulnerability doesn’t disappear when a claims firm has to pay £250. Behaviour does.
But notice what the fix wasn’t
Here’s the part the “too much protection” case misses. The FOS didn’t reduce consumer protection by a single degree. A consumer with a genuine complaint can still bring it themselves, for free, and keep every penny of redress. The vulnerable lost nothing. What changed was the incentive architecture — the system finally priced behaviour without touching vulnerability.
That’s the template. The answer to a framework that can’t tell behaviour from vulnerability isn’t less protection — it’s building the mechanisms that make the distinction. And nowhere does the market fail that test more expensively than at the point of decline.
The FCA’s answer sounds great. It’s been tried.
Walters’ speech contained what looks like the solution: declined mainstream customers should be referred to community lenders, where many would have a higher probability of approval. The FCA “sees no reason why this can’t be done”.
Policymakers love this story. I love this story. But those of us who work with credit unions know it’s been told before. Referral arrangements between major banks — Lloyds and Halifax among them — and credit unions have existed for years. The volumes were low. The funded outcomes were poor to non-existent. Great narrative, failed execution — because the technology, people and processes on the receiving end weren’t up to it.
And the harder truths sit inside the community lending sector itself. Most community lenders are geared to serving existing members, not acquiring new ones. Data skills and decisioning capability are thin across the sector, and processes remain grounded in manual underwriting — so the path of least resistance is to do nothing. And assuming credit unions are simply willing to absorb drastically higher risk is naive. A mainstream decline is a decline for a reason; the opportunity is in identifying the subset whose circumstances — not behaviour — put them there. That takes decisioning capability the sector largely hasn’t yet adopted — these services are readily available; the take-up isn’t there.
The FCA sees no reason it can’t be done. Nor do I — the technology to build the receiving end now exists, and the opportunity is there for anyone willing to take it up. Put a leading near-prime lender like Oakbrook or 118 118 Money on the sending side of working referral rails, and the difference in impact would be massive.
But leads alone are not enough. Two structural forces have kept this market where it is: incumbent players who have held it back, and operating models that simply cannot afford the data science talent needed to drive the change. A referral is only worth making if the receiving end is motivated — and equipped — to do something serious with it.
That is part of what I am trying to change as Vice Chair of the Credit Union Foundation: making education and upskilling in data science, analytics and conversion rate optimisation — the craft of making digital products genuinely work — far more readily available to credit union professionals. The sector does not need another pilot scheme that quietly fizzles out. It needs structural change, where community lenders are duly motivated, and duly capable, of making a big shift.
Protection isn’t the enemy. Indiscriminate protection is.
The CRCO is right: a framework that can’t distinguish behaviour from vulnerability facilitates bad actors and taxes everyone else. But the FOS episode shows the fix — price the behaviour, protect the vulnerability, and build the infrastructure that tells them apart.
At the point of decline, that infrastructure means affordability data that travels with the applicant, decisioning that community lenders can actually run, and referral rails that turn a mainstream “no” into an informed “yes” for the right borrowers — at a risk level community lenders can genuinely hold.
The regulator has said it sees no reason this can’t be done. It’s right. But the metamorphosis won’t come from another consultation. It’ll come when someone builds the plumbing that lets the market finally tell the difference.
← Back to News & Opinion