By John Downie, SteadyPay
Payments apps, wallets, personal finance tools, vertical fintechs – they’ve all built deep relationships with their users. They see income, spending patterns, cashflow gaps in real time. But when those same users need to borrow, be it for an emergency, a big purchase, to bridge the gap between paydays, they’re sent elsewhere. Back to a traditional bank. To a high-cost lender. To a BNPL provider who doesn’t know them at all.
Over half of our users tell us they’re borrowing to cover an unexpected expense. The Building Societies Association recently found that one in five UK adults couldn’t cover a £300 emergency, rising to more than a third of under-24s. There is clear demand for lending services. And there is the supply – it’s just that if the supply isn’t coming from you then you are going to lose business. And it’s someone else who then monetises the borrowing need – often badly, and often at a price the customer shouldn’t have to pay.
I’m not going to claim that people are blind to the opportunity. It’s just that the barriers are very real. Consumer credit authorisation, FCA reporting, affordability assessments – it’s a different regulatory world from payments or money management. Then there’s the problem that lending ties up the balance sheet at a time when most fintechs are burning cash on growth, rather than provisioning for loan books. Infrastructure can be an impossible task too. Building underwriting models from scratch is a multi-year, specialist endeavour. Get it wrong and you haemorrhage money. Get it publicly wrong and you damage a brand built on trust.
Around 20 million people in the UK fall outside the traditional credit system. Say there’s roughly 53 million adults in the UK and that’s 38% of that population. Are they high risk? Some might be. But, for the vast majority, they are simply unlucky enough that the infrastructure wasn’t built for how they live and earn. It’s the gig workers, the hourly earners, the immigrants, or anyone whose financial life doesn’t fit neatly onto a credit file. Many of them are already using fintechs for payments and money management. When they can’t access credit through that same app, they default to high-cost alternatives – or go without entirely.
What’s crazy is that the data needed to underwrite these borrowers fairly – like the real income, real spending, real cashflow – is already sitting inside the fintechs they use every day. Open Banking makes it accessible. At SteadyPay, we’ve originated over 600,000 loans to exactly this population, maintaining default rates below 5 per cent – less than half the industry average. These are creditworthy people. The system just couldn’t see them.
We saw the great unbundling in payments a few years ago, and this is now reaching credit. Stripe abstracted payment processing so that every internet business didn’t need a merchant acquiring licence, and plug-in credit rails now let fintechs offer lending without building the regulatory, capital, and underwriting infrastructure from scratch.
The split is clean: the infrastructure partner owns underwriting, capital, compliance, collections, and conduct obligations. The fintech partner owns the customer – the UX, the brand, the distribution, the relationship. Credit becomes a feature of your product rather than a separate business you have to run.
There’s a subtler version of this for banks and fintechs that already lend. Any institution with a banking licence operates under a lending policy reviewed and approved by the regulator (as it should be). But that necessarily means a significant segment of customers fall outside their criteria. Those customers don’t disappear. They go somewhere else, usually somewhere worse. Parallel decisioning lets an infrastructure partner catch those declines, serve the customer under the same brand, and graduate them back to the bank’s own book when their profile strengthens. The customer stays in the ecosystem. The bank expands its reach without compromising its lending policy.
None of this means building your own lending stack is always wrong. But before you commit, ask five honest questions:
- Do you have the balance sheet capacity, or will you need a funding partner anyway?
- Do you have credit risk specialists who’ve built and stress-tested models through a full cycle – not just data scientists?
- Is your regulatory appetite high enough for what is a 12-24 month authorisation process and a permanent compliance overhead?
- Can your leadership absorb the distraction when your core product still needs focus and scale?
- And will your investors – who backed an equity focused growth story – accept balance sheet exposure in the portfolio?
If you answer no to two or more, building your own stack will likely cost you more than it gains. It’s the same logic that made you use Stripe instead of building a payment gateway.
You can’t build everything in-house. But you can assemble the right infrastructure, and use it to serve the millions of people that traditional finance still pretends don’t exist. The infrastructure is ready. The industry just needs to learn to use it.
