Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Ripple Is Quietly Winning Bank Settlement Payments
Transcript
- Lucas: So here's a number that stopped me this week: banks still hold about one point six trillion dollars in pre-funded nostro accounts just to grease cross-border payments. That's roughly the GDP of Australia, just sitting in accounts waiting to settle. Luna: One point six trillion? That's the correspondent banking float — the money that's basically parked in case someone wants to wire yen to Mexico. Lucas: Exactly. And it's been a known inefficiency for decades. The question is: can fintech actually dent it? Most of our episodes have been about consumer-facing fintech — Stripe, Chime, Revolut. Today I want to look at the plumbing underneath. Specifically, Ripple. Luna: Ripple the token, or Ripple the company? Lucas: The company. And specifically their product called On-Demand Liquidity, or ODL. It uses the XRP token as a bridge currency between two fiat currencies. So instead of a bank pre-funding a dollar account in Mexico, they send dollars to Ripple's network, which converts them to XRP, moves that in three to five seconds, and converts back to pesos on the other side. The bank doesn't need to hold peso balances upfront. Luna: Right. So the XRP token is the settlement layer — it replaces the pre-funded nostro. The bank only holds XRP for the few seconds the transaction takes. Lucas: That's the thesis. And it's been controversial because, you know, 2017 was the peak of crypto hype and XRP had that massive run. But since then, Ripple has been signing actual bank contracts. Quietly. Bank by bank, regulator by regulator. Luna: Give me the concrete example. Who's actually using it in production? Lucas: Santander is probably the best-known. They launched One Pay FX in 2018 using Ripple's technology, and it's been live across multiple corridors — Spain to Mexico, Spain to Chile, UK to Poland. Standard Chartered's SC Ventures did a pilot with Ripple across their ASEAN corridors. And SBI Holdings in Japan — they're not just a client, they're a strategic partner. They run SBI Ripple Asia and have been pushing xrp based settlement inside Japanese banking. Luna: So these are real, live payment rails, not just proof of concepts that never leave a sandbox. Lucas: Correct. And the interesting shift is geographic. The US has been a regulatory mess for Ripple — the SEC lawsuit started in 2020 and dragged on for years. So Ripple leaned into Asia and the Middle East. The Monetary Authority of Singapore gave them a Major Payments Institution license in 2023. The UAE's financial center, ADGM, did the same. Ripple now has licenses in Ireland, Singapore, Dubai, and New York — but the New York one is the limited BitLicense, not a full bank charter. Luna: So they're going where the regulation is clear. That's smart. If today was actually useful to you, the way these stay ad-free is listener support — buy me a coffee dot com slash fexingo. Keeps us independent and digging into these specific use cases. Lucas: Yeah, and we really mean that. Every coffee funds a deeper look. Anyway, back to the geography — the US overhang means Ripple's biggest wins are outside the dollar system. And that's actually where the inefficiency is worst. Emerging market corridors have the highest correspondent banking costs. A remittance from the UAE to the Philippines can cost 7 percent. Ripple claims ODL can bring that under 1 percent. Luna: But there's a tension here. The XRP token is volatile. If the settlement takes three seconds, volatility isn't a huge risk. But if the bank has to hold XRP for any time — say, because settlement fails or there's a queue — they're exposed. Lucas: That's the big design challenge. Ripple's answer is that ODL uses a market maker algorithm that locks in the exchange rate at the start of the transaction, and the actual XRP holding window is measured in seconds. In practice, they say the volatility risk during settlement is about one basis point, which is less than the currency fluctuation risk in the traditional model. Luna: One basis point is basically noise. But if Ripple wants to scale beyond niche corridors, they need banks to trust that the token will be liquid enough. Lucas: And that's the chicken and egg. XRP's daily trading volume has fallen from the 2021 peaks. Ripple itself is the largest holder — they hold about 45 billion XRP in escrow, releasing one billion per month. That's a supply overhang that makes institutional investors nervous. You're not just using a settlement token; you're betting that the issuer won't flood the market. Luna: So the same centralization concern that critics have always raised. Ripple controls the supply. They say it's transparent and regulated, but it's still a single company's token. Lucas: Right. And that's why the comparison with SWIFT's gpi is useful. SWIFT gpi — global payments innovation — came out in 2017 and now handles over 80 percent of SWIFT payments. It cut settlement time from days to minutes by standardizing the tracking and fee disclosure. No new token, no new ledger. Just better process. Luna: But SWIFT gpi still uses the correspondent banking model. It's faster and more transparent, but it doesn't eliminate the float. Banks still need those nostro accounts. Ripple's advantage is that it collapses the float to zero — or at least to a three-second holding of a volatile asset. Lucas: That's the trade-off. Speed and capital efficiency versus trust in a new settlement asset. And the data so far is suggestive but not definitive. Ripple says ODL saved partners over forty million dollars in fees in 2025. But that's across maybe a dozen active corridors. The total addressable market is one point six trillion dollars of float. Forty million is a rounding error. Luna: What would make it tip? What's the trigger for mass adoption? Lucas: Two things. First, a clear regulatory framework in the US. If the SEC provides a no-action letter or if legislation like the Stablecoin Trust Act passes, that removes the legal cloud. Second, a major central bank issuing a digital currency — a CBDC — that interoperates with Ripple's ledger. Ripple already has pilot projects with the central banks of Bhutan, Palau, and Montenegro. If one of the G20 central banks joins, that changes the signal. Luna: But aren't central banks more likely to build their own infrastructure? The Fed has FedNow, the ECB has TARGET Instant Payment Settlement. Why would they use a private token? Lucas: Because interoperability is hard. Every central bank builds its own system, and then you have bilateral links between each pair. That's a quadratic problem. Ripple's network is a hub and spoke: connect once to the RippleNet and you reach every other member. For smaller central banks, that's cheaper and faster than building bilateral connections. Luna: So Ripple's real pitch is: don't build your own bridge, just join the network. And they're betting that the token is a necessary part of that — not just a technology choice, but an economic incentive layer. Lucas: Exactly. And that's where the conversation gets really interesting. Because if you accept the token, you're accepting a new monetary settlement asset alongside central bank money. That's a huge leap for most finance ministers. Luna: Let's bring it back to the numbers. What's Ripple's revenue model? How do they make money? Lucas: They have three streams. One: selling the software license for RippleNet — that's a recurring fee to banks. Two: transaction fees on ODL — a tiny per-transaction cut. Three: they hold a lot of XRP, and as the token appreciates, their balance sheet grows. In 2025, Ripple reported about four hundred million dollars in revenue, up from two hundred million in 2023. But the bulk of that is still software licensing, not transaction fees. Luna: So the transaction volume on ODL isn't big enough yet to drive meaningful fee income. They're still in the land and expand phase. Lucas: Correct. And that's actually healthy — it means they're not dependent on speculation. If the token price crashes, they still have recurring revenue from banks. But it also means the network effects haven't really kicked in. You need more banks and more volume for the network to be self-sustaining. Luna: Let's contrast Ripple with another fintech we've covered: Adyen. Adyen built a single platform for merchants to accept payments globally. Ripple is building a single platform for banks to settle payments globally. Both are trying to eliminate intermediaries. But Adyen succeeded because they focused on the merchant experience — checkout, fraud, reporting. Ripple is focused on the back end — settlement, liquidity, compliance. Lucas: That's a good comparison. And Adyen's revenue is in the billions now. Ripple is at four hundred million. But the total addressable market for settlement is much larger than merchant acquiring. The question is whether Ripple can execute a similar flywheel: more banks join, more liquidity, lower costs, which attracts more banks. Luna: What's the biggest risk you see? Not the SEC — we've talked about that. Something structural. Lucas: The biggest risk is that banks decide they don't need a separate settlement layer. Look at what JPMorgan did with JPM Coin — it's a deposit token on a private ledger, used for intraday settlement within their own network. No public token, no volatility, fully regulated. If every big bank builds its own private settlement coin, Ripple becomes a solution looking for a problem. Luna: But JPM Coin only works for JPMorgan clients. It doesn't connect different banks seamlessly. Ripple's value is interoperability across banks that don't trust each other. Lucas: Right. But that interoperability relies on the XRP token being trusted by all parties. And trust in a private company's token is harder to build than trust in a consortium or a central bank. That's why I think the most likely outcome is that Ripple becomes a technology provider for central bank digital currencies — not the settlement token itself, but the software layer that connects different CBDCs. Luna: So Ripple as the plumbing for central bank digital currencies, not the settlement asset. Lucas: Exactly. And they've already repositioned that way. Their website talks about 'real-time settlement for the internet of value' — not 'XRP is the future of money'. The token is still there, but it's not the headline. The headline is speed, transparency, and lower costs. Luna: A pragmatic pivot. So where does this leave us for the next few years? Lucas: I think 2026 and 2027 are make or break for ODL's volume. If they can get to ten billion dollars in daily settlement volume — that's about one percent of the correspondent banking market — then the network effects become visible. If they stay at one to two billion, they're a niche product for a few corridors. Luna: That's a clear metric to watch: daily ODL volume. And the second metric? Lucas: Number of live corridors. Right now they have about seventy. If that doubles to one hundred forty in two years, and the volume per corridor is growing, that's a strong signal. If it plateaus, banks are treating it as an experiment. Luna: And what about the token price? Does it matter for the business case? Lucas: Only indirectly. If XRP is too volatile, banks hedge it and that adds cost. If it's too illiquid, ODL doesn't work for large payments. But the business model doesn't depend on price appreciation. That was the old narrative. The new narrative is: we sell software, we take a tiny fee, we make money whether XRP is at fifty cents or five dollars. Luna: That's a much healthier business. And it's also more boring — which is fine. Lucas: Boring is good in settlement. You don't want your cross-border payment to be exciting. You want it to arrive in seconds, cost a few cents, and not require a spreadsheet to figure out the fees. Ripple is closer to that than they were in 2017, but there's still a long way to go. Luna: And that's the real story: a fintech that survived the hype cycle, absorbed a multi-year SEC lawsuit, and is now quietly signing bank deals in Asia and the Middle East. It's not the revolution we were promised. But it might be something more durable. Lucas: A boring, functional settlement layer. Which, if it works, could eventually reshape that one point six trillion dollar float. And that's worth watching.