Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / Why Neobanks Are Buying Their Own Core Banking Software
Transcript
- Lucas: So Monzo spent something like fifty million dollars a couple of years ago to buy a core banking platform. Not to license it — to own it. Luna: And that raised a lot of eyebrows in the fintech world because the whole neobank model was built on not owning that stuff. You rent the plumbing. Lucas: Exactly. The original pitch for companies like Monzo, Chime, N26 was: we'll be the user-friendly front end, and we'll plug into someone else's banking infrastructure — usually a partner bank or a core processor like Jack Henry or Fiserv. Luna: Right. Don't build what you can buy. That was the mantra for years. Lucas: And it worked at small scale. But as these neobanks crossed a million, then five million, then ten million customers, the unit economics started to look different. The per-account cost of that third-party infrastructure stops scaling down the way you want. Luna: So the fifty million dollars — that's a bet that owning the stack saves money on a per-deposit basis once you're big enough. Lucas: Right. And Monzo has been pretty open that after the acquisition, their cost to service a deposit improved by about fifteen basis points. That's huge when your margin is measured in basis points. Luna: Fifteen basis points on fifty billion in deposits — that's real money. Lucas: Seventy-five million dollars a year in savings, roughly. So the acquisition pays for itself inside the first year if you hit those numbers. Luna: And Chime — they took a different route, right? They started building their own core in-house. Lucas: Yeah. Chime has been quietly developing a proprietary core banking system for the last three years. They haven't bought a company — they've hired a lot of engineers from the traditional core vendors and built it themselves. Luna: That feels slower. Why not just buy like Monzo did? Lucas: Because you have to integrate whatever you buy. Monzo bought a platform that was already running in production, but they still spent eighteen months migrating accounts. Chime's approach gives them more control over the architecture from day one, but the time to value is longer. Luna: And there's the risk that by the time you finish building, your needs have changed. Lucas: Exactly. That's the classic make versus buy trade-off in technology. And what's interesting is that both approaches are now being validated by the market. Monzo's deposit margin has improved noticeably since the migration. Chime hasn't disclosed their numbers yet, but they're clearly betting that owning the stack is the only path to sustainable profitability at scale. Luna: What about the smaller neobanks? The ones with a few hundred thousand customers. Does this logic apply to them? Lucas: Probably not. The economics don't work until you have a certain deposit base. If you've got two billion in deposits, saving fifteen basis points is only three million a year. That doesn't justify a fifty-million-dollar acquisition or a multi-year build project. Luna: So there's a threshold. Do we know where it is? Lucas: Roughly, you need at least ten billion in deposits before owning your core starts to make sense on a pure cost basis. Below that, the flexibility of renting infrastructure outweighs the margin benefit. Luna: But there are other reasons to own the core beyond cost. Lucas: Right. Speed of product development is a big one. When you're on a third-party core, every new feature — instant payments, lending products, cryptocurrency trading — requires the vendor to build it or at least certify it. That can take months. When you own the core, your engineers can push changes in days. Luna: Monzo has talked about that. They launched a lending product six months faster after the migration. Lucas: Exactly. And for a neobank trying to increase revenue per customer, time to market for new products is everything. You're not just saving money — you're unlocking revenue you couldn't get before. Luna: So the strategic calculus has shifted. When you're small, rent. When you're big enough, own. But what about the very largest players — the JPMorgans of the world? Lucas: They've always owned their core. JPMorgan spends something like fifteen billion a year on technology. They've been building their own systems for decades. The difference is that now, the neobanks are starting to act like incumbents in terms of infrastructure strategy. Luna: Which is ironic because the original pitch was that they were disrupting the incumbents by not building infrastructure. Lucas: It is. But that's the natural evolution of any disruptive model. You start by borrowing someone else's rails, you prove the demand, and then — if you're successful — you eventually have to build your own. The question is when to pull the trigger. Luna: And there's a risk of getting it wrong. If you buy or build too early, you waste capital. Too late, and your competitors who moved earlier have a cost advantage that's hard to close. Lucas: Yeah, it's a timing play. And it depends a lot on your specific product mix. A neobank that's mostly deposits and debit cards might need different infrastructure than one that wants to do lending, investments, and insurance. Luna: That's a good point. The more products you offer, the more you benefit from controlling the core. Lucas: Because each product has its own data requirements. If you own the core, you can build a unified customer profile that works across all of them. If you're on a third-party system, you might have to patch together data from multiple sources, which creates latency and risk. Luna: So what's the next frontier? Are we going to see more neobanks follow Monzo and Chime? Lucas: I think so. There are probably a dozen neobanks globally that are at or near that ten billion deposit threshold. I'd expect two or three of them to announce core migrations in the next twelve months. Luna: And the core vendors — the Fisers and Javelins of the world — how are they reacting? Lucas: They're watching their biggest customers consider leaving. So they're trying to innovate faster, offering more modular, api friendly versions of their platforms. But it might be too late. The neobanks have seen what owning the core does for their margins, and they're not going to unsee it. Luna: It's a fascinating shift. And it really underscores how the fintech landscape is maturing. The era of pure-play front-end neobanks might be ending. Lucas: Yeah, the winners are becoming full-stack financial institutions. They started as thin layers on top of old infrastructure, and now they're digging down into the stack. Luna: It's almost like the opposite of the unbundling thesis that drove fintech a decade ago. Lucas: Right. First you unbundle the bank, then you rebundle the infrastructure. The companies that survive will be the ones that own their own rails. Luna: That's a good place to leave it. And hey, speaking of independent control — this podcast is completely ad-free. We don't run any sponsorships or commercials, and we want to keep it that way. Lucas: Yeah, that's important to us. If you find these conversations useful for what you're building or running, the best way to support that choice is at buy me a coffee dot com slash fexingo. It's a simple way to keep the show independent. Luna: Even a small contribution helps. And it means we never have to tailor the content to an advertiser's agenda. Lucas: Exactly. So back to the core banking shift — it will be interesting to see if any of the larger neobanks decide to spin off their core platforms as a product for smaller banks. That would be a whole new business model. Luna: That would be a twist — the disruptors becoming the infrastructure providers for the incumbents. Lucas: It's happened before in other industries. Amazon built infrastructure for itself and then sold it as AWS. Could a neobank do the same with banking software? Luna: That's a great question for a future episode. For now, thanks for listening.