Something fundamental changed in finance, and almost nobody noticed.
For 250 years, there was one rule: to access clearing networks, you needed to hold reserves at a central bank. This wasn't a technical limitation - it was the structure of money itself.
Banks could hold reserves. Banks could clear with each other. Everyone else was a guest in the system.
Then stablecoins arrived, and this iron law shifted.
What Actually Changed
Self-custody in crypto usually means individuals holding their own keys. But the real disruption in fintech is organizational self-custody. For the first time in modern finance, a payment company can hold its own reserves without a bank.
Think about what this means. A PSP in the Philippines can hold USDC. Not in a bank account. Not through a correspondent. Directly, on-chain, in their own wallet that they control completely.
This seems small. It's not.
Why Network Topology Matters More Than Rails
One of the most critically neglected dimensions in the stablecoin landscape is topology. While discourse obsesses over payment rails - how fast, how cheap, how many currencies - far less attention is paid to the architecture that underlies it all: the shape of the network, the way value flows, and the web of obligations that drive it. These are not secondary effects of the system; they are the system. The shape of the network determines everything else from the core.
Traditional finance has hub-and-spoke topology:
Hub-and-Spoke Clearing Model (Banks centralize clearing; PSPs must route through their bank)
Each level can only interact with the level above it. PSPs can't hold central bank reserves, so they can't access interbank clearing. They're permanently peripheral as a non-participating partner.
But with stablecoins, the topology flattens:
Flat structure allowing PSPs to work together directly
Every PSP is now a first-class node. They hold reserves (stablecoins) directly. They can clear with anyone else who holds reserves. No hierarchical permission needed.
Recent industry report (Artemis) shows over 30 major PSPs now processing stablecoin payments globally, with the largest corridors connecting major financial hubs in Asia and the Americas. Each maintains separate bilateral relationships - creating exactly the n² complexity problem that clearing networks were invented to solve.
The Clearing Privilege
Here's what most people miss: clearing isn't just about operational efficiency. It's about the fundamental pricing power.
When banks transact with each other, they don't settle every payment gross. They net obligations down to tiny fractions. A bank moving $1 billion might only need $34.5 million in actual liquidity. That's 29:1 leverage through netting. For example, CHIPS (which processes 95% of USD large-value payments) operates at 29:1 efficiency - for every $1 in capital, banks can settle $29 in volume.
This efficiency translates directly to pricing. If Bank A needs 29x less capital than PSP B to move the same volume, guess who wins on price?
The infrastructure that enables this has always been bank-exclusive. It's not that PSPs couldn't theoretically net with each other - it's that the infrastructure was never built for them. They weren't allowed in the club.
From Second-Class to Self-Sovereign
Traditional PSP economics were brutal:
You need a bank partner to hold funds
Banks charge for access (account fees, transaction fees, FX spreads)
You prefund everything with each counterparty
Your capital efficiency is 1:1 at best
Banks achieve 29:1 through clearing infrastructure
You literally cannot compete on price
With organizational self-custody through stablecoins:
You hold your own reserves
No bank intermediation required
But you still prefund everything with each counterparty
Your efficiency remains 1:1
Banks still have 29:1 through their clearing networks
The pricing disadvantage remains
The right to self-custody without the infrastructure to clear is like freedom without power.
The Infrastructure Gap
Take a real PSP processing $100k daily. Typical floats at $100-200k - essentially 1-2x their daily volume locked up just to operate. Scale that to yearly lockups, opportunity cost, and then across every single PSP, and it becomes clear that this is the most fundamental problem in payments infrastructure.
The numbers are staggering. PSPs are now processing over $3 billion monthly in B2B payments alone, with average transaction sizes exceeding $200,000. What's invisible in these headline numbers is the massive capital infrastructure required - the prefunding requirements that scale linearly with volume while banks enjoy logarithmic efficiency through clearing.
Even sophisticated PSPs doing internal netting still face this reality: when settling with external parties, it's always individual gross settlements.
Wise can net internally across their own corridors, achieving some efficiency. But when they need to settle with a bank or liquidity provider? Full amount, every time. No netting. No capital efficiency. Just 1:1 gross settlement.
Here's the brutal truth: the cost of payments isn't in the movement - it's in the capital locked up for settlements, the operational overhead and associated risks.
Blockchains solved transaction speed, but not capital efficiency. A USDC transfer settles in seconds, but you still need:
Prefunded accounts with every counterparty
Manual treasury operations to manage positions
Settlement risk buffers
Capital locked up proportionally to volume
Moving money fast doesn't make it cheap if you need 100% collateral for every transaction while your bank competitors need 4%.
The Uncomfortable Truth About Where Stablecoins Win
Here's what nobody wants to admit: stablecoins only succeed where banks fail at clearing.
Look at domestic payments. US ACH costs pennies. European SEPA is often free and instant under €100k. Why? Because banks have mature clearing networks achieving high capital efficiency. They can simply price PSPs out of existence.
But cross-border? Banks are stuck with correspondent relationships, bilateral prefunding, and all the same inefficiencies PSPs face today. That's why a stablecoin payment from US to Philippines can beat a SWIFT wire. Not because blockchain is magic, but because banks don't have efficient cross-border clearing.
The evidence is clear. BIS 1265 research shows stablecoin payment flows have +25% elasticity to traditional remittance costs - they only grow where the alternative is expensive enough. In efficient corridors, stablecoins lose. In expensive corridors, they win. It's pure arbitrage on inefficiency.
This isn't a technical accident. It's network topology determining market structure. Where banks can clear efficiently (domestic), PSPs can't compete. Where banks can't clear efficiently (cross-border), PSPs have an opening.
How long will this last?
Banks Are Building the Bridge
Figure 5: Connecting digital islands
Project Nexus and SWIFT + CLS clearing experiments are already exploring tokenized clearing. Central banks are designing wholesale CBDCs with - surprise - permissioned access where only banks can participate directly. Fintechs build in the open and iterate fast; banks build pilots and launch once. If you look at the progress closely, they are not behind, potentially ahead, and the final straw in regulations will open the flood gates that will accelerate this development.
The SWIFT+CLS report explicitly states the game plan:
"Netting capabilities embedded into the CLS processing engine significantly reduce the amount of liquidity needed to settle FX transactions, especially when fragmentation of liquidity due to pre-funding of accounts is required."
Read that again. Banks know exactly what problem they're solving - the same capital inefficiency PSPs face today. They're not building this for fun. They're building it because whoever controls clearing controls pricing power in payments. Operating any other way takes us back to 1775.
The report goes further, describing how they'll achieve "material liquidity optimisation" through "standard matching and netting" while exploring "molecular settlement" - a balance between instant risk-free settlement and capital efficiency.
This isn't theoretical. It's operational. And it's being built for "central bank money" - meaning PSPs need not apply.
What happens when banks achieve their typical 29:1 capital efficiency in cross-border tokenized payments with CLS providing risk-free payment-vs-payment mechanisms while PSPs still operate at 1:1 fire fighting risk openings at every corner?
The math is extinction-level. Any corridor where tokenized bank clearing launches becomes instantly unviable for PSPs operating without clearing infrastructure. It's not competition - it's a pricing power gap too wide to bridge.
We're watching this play out again:
PSPs innovate in neglected markets (cross-border)
PSPs validate demand and build solutions
Banks notice the validated market
Banks build the same technology stack for that market
PSPs get priced out by superior capital efficiency
PSPs must become banks to survive
This is why Adyen has better margins than Stripe. This is why every successful PSP eventually gets a banking license. When you can't compete on price due to structural disadvantages, you have two choices: join the structure club or standby.
The old fintech adage proves itself again: fintechs want to become banks, and banks want to become fintechs. But banks have had a structural advantage, until now.
The Value-Added Services Trap
"We'll compete on user experience"
"Our API is better"
"Faster onboarding"
"Real-time dashboards"
"Compliance"
When you can't compete on price due to structural capital inefficiency, "value-added services" becomes the only play. Better UX, slicker interfaces, prettier dashboards.
We think we're innovating, but we're really just decorating our structural prison cell.
No amount of UI polish overcomes a 29x capital efficiency disadvantage. No API elegance matters when your competitor needs 96% less working capital. We're fighting physics with marketing.
These aren't useless features - they're necessary for adoption. But competitive advantages built on UX have expiration dates. Structural advantages don't.
Why This Moment Won't Last
Organizational self-custody through stablecoins is a historical accident. It broke 250 years of bank exclusivity, but the window is closing fast.
The urgency is real: stablecoin payment volumes are exploding - B2B alone grew from under $100M monthly to over $3B in just two years. Once banks see these validated flows and launch tokenized clearing, they'll recapture this market overnight with their 29:1 capital efficiency advantage.
Banks aren't sleeping. They're building the infrastructure to recapture this moment. When tokenized clearing networks launch with the same permissioned access as traditional systems, PSPs will be right back where they started - guests in someone else's system.
The Infrastructure Imperative
For 250 years, the shape of financial networks determined who captured value. Hub-and-spoke meant banks won. Self-custody temporarily broke this model, giving PSPs the same foundational right banks always had.
But rights without infrastructure are temporary privileges, not permanent advantages. We have the keys to the kingdom but no roads between castles.
Every day PSPs operate at 1:1 capital efficiency while banks leverage 29:1 through clearing networks is another day closer to the window closing. When tokenized bank clearing launches - and it will - the pricing power gap will make today's inefficiencies look quaint.
PSPs aren't just payment companies anymore. They're participants in a historic disruption. For the first time in modern finance, we have the opportunity to build our own clearing infrastructure.
The question isn't whether we'll build our own PSP-driven clearing infrastructure. It's whether we'll build it before banks make it irrelevant.
Next: How we are unknowingly recreating 18th-century banking inefficiencies - and what the path forward looks like
If you're a PSP dealing with float and capital efficiency pain or building stablecoin infrastructure, reach out: Twitter | Telegram
Latest updates
