The FCA’s latest mortgage consultation (CP26/18, out in June) proposes letting brokers run the “tailored interactive dialogue” that tests whether an interest-only borrower has a credible way of repaying the capital. Writing in FT Adviser, Chloe Cheung captured the response from the intermediary market in her piece “FCA mortgage proposals could put more liability on advisers” — and it was almost unanimously wary. It should be the lender that polices this. Where does the liability sit? Our PI premiums will go up. Networks will cap how much interest-only we can write.
Every one of those concerns is reasonable. The rules are vague, the liability really is unclear, and no one should sign up to carry a risk they can’t define. But step back from this one consultation and there’s a bigger prize in plain sight: the instinct, every time, is to push work — and with it, information — back towards the lender. Resist that instinct and the same front-end work becomes the most valuable asset a broker could own.
The inefficiency nobody’s pricing
Look at how a case actually moves today.
A broker wins a lead. They assemble a relatively thin file and fire it at a panel of lenders. Each lender then independently pulls its own bureau data, runs its own affordability and scorecard checks, and comes back with a decision. The broker takes the cheapest yes to the client and gets paid a procuration fee. Job done.
Now count the waste. The same borrower is credit-searched and underwritten five, six, seven times over — once per lender — on the strength of a file the broker never really owned. Every lender prices in the cost and uncertainty of underwriting a case they can’t fully see. And the broker, who is closest to the customer and did the actual work of finding and understanding them, captures none of that duplicated cost as value. They’re a lead router paid a flat fee, not a data owner with leverage.
What “belts and braces” actually buys you
Picture the same case run the other way. The broker collects the full picture up front — verified income, expenditure, the borrower’s real credit file, the repayment-strategy dialogue the FCA is now formalising — and presents lenders with a clean, pre-qualified, data-rich case. Instead of a panel of lenders each doing the same expensive work in parallel, you have lenders bidding on a file that’s already been done properly, once.
Three things follow from that, and none of them are theoretical.
The cost of underwriting collapses, and a broker with a clean file has every right to demand the saving be passed through as sharper pricing for the client. The customer experience improves — one thorough conversation, not seven duplicated searches and a fortnight of back-and-forth — and the broker’s standing rises with it. And the commercial one that brokers keep missing: if you’ve stripped real cost out of a lender’s process, you can negotiate a share of it. Higher proc fees stop being a favour and start being a trade — you saved them money, you get paid for it.
That’s the reframe. The “extra responsibility” the FCA is nudging towards isn’t a liability to be batted back to the lender. Handled well, it’s the foundation of a better-paid, higher-status role. Quality over quantity: fewer, better-qualified cases that each earn more.
The catch, and it’s a real one: the data
Here’s the honest problem, and it’s the reason the instinct to push work back exists at all. Brokers can’t run a proper belts-and-braces process today because they can’t see the data that matters. They’re largely confined to public information — the electoral roll, CCJs, insolvencies — which seldom lines up with the private scorecards lenders actually decide on. You can do all the front-end work you like; if you can’t see what the lender sees, you’re still guessing.
That’s not a broker failing. It’s a structural one, and it sits with the bureaux and the principle of reciprocity — the long-standing rule that you can only see the credit data you contribute. Brokers don’t lend, so they don’t contribute repayment data, so they don’t get the full file. Full stop.
Which is exactly why the FCA’s Credit Information Market Study matters more to brokers than they realise. The final report (MS19/1.3, December 2023) put reform of the reciprocity regime and a new governance body — the Credit Reporting Governance Body, replacing the old Steering Committee on Reciprocity — squarely on the table. The follow-up consultation, CP26/7 (February 2026), goes further: mandatory reporting across all the designated credit reference agencies, and an explicit reopening of the Principles of Reciprocity and which use cases are permitted.
That is the door. If reciprocity is being redrawn anyway, the intermediary market should be arguing — loudly, now — for a defined route to full-file access and a seat at the table where those principles are set. Not as a favour, but because a broker holding a customer’s verified, permissioned data is doing real work to improve the quality and completeness of what flows through the system.
“But won’t we just sell less?”
The devil’s advocate case writes itself: more scrutiny means more declines, fewer completions, less volume. I don’t buy it, for two reasons.
First, search costs go up, not down — for everyone. Brokers running proper searches, and lenders still running theirs, means the data itself becomes something worth paying for rather than a sunk cost buried in each lender’s underwriting.
Second, and more interesting, it opens a product that doesn’t exist yet. Right now a credit search is one-to-one: one requester, one pull, one fee. But a broker sitting on a clean, permissioned file could serve it one-to-many — a single search surfaced to every lender that wants to see it, with each paying to view it. Tokenise the search and you’ve got a neat mechanism: one broker pull, charged out to the panel that requests sight of it, with a clear audit trail of who accessed what and when. The buy-to-let space, where cases are often more bespoke and the panels more specialist, is the obvious place for a data product like this to take root.
This is already working next door
If that sounds far-fetched, it’s already running in the room next door. In collections, IE Hub has spent years doing exactly this: a customer builds their income-and-expenditure picture once, in one place, and shares it — on their permission — with every creditor they owe, rather than repeating a distressing 45-minute affordability interview a dozen times over. One dataset, many recipients, the customer in control. Creditors get accurate, consistent affordability and vulnerability data; the customer does the work once.
That’s the one-to-one-to-one-to-many model, live, in a regulated corner of consumer finance. There is no reason the same architecture can’t sit at the front of the lending journey instead of the back of the collections one.
The opportunity brokers are running from
So yes — the FCA is asking brokers to shoulder more of the front end, and yes, they should demand clarity on liability before they take it on. But the deeper instinct to hand the work, and the data, back to the lender is the wrong one. The information brokers are being asked to gather is the one asset in this market they could actually own — and owning it is how you get to cheaper pricing for clients, a higher-status role, and better-paid deals.
The consultations are open. This is the moment to ask for the data, not to give back the work.
← Back to News & Opinion