Stablecoins settle fast. They also require 100% prefunding. That tradeoff sits at the center of a new debate after 17 of the largest U.S. banks said they would use The Clearing House to clear and settle tokenized deposits on-chain rather than rely on stablecoins.
In Simon Taylor’s telling, the real story is not blockchain by itself. It is netting. Banks have spent decades building systems that let them offset payment obligations against one another, sharply reducing the amount of cash they have to park in advance. The question raised in the piece is whether stablecoins can eventually borrow that same playbook.
Fast payments are not always the whole answer
From a merchant’s perspective, instant settlement is hard to argue against. Goods are sold, inventory is gone, and the cash may not arrive for days. In B2B settings, the wait can stretch into weeks or months. That leaves a financing gap that has to be borrowed against or absorbed as operating risk. Consumers also tend to respond well when payment is immediate.
Corporate treasurers and banks do not always see the equation the same way. Slower payment can be cheaper on the balance sheet because major clearing systems use netting and liquidity-saving mechanisms to minimize the amount of cash that must actually move. Taylor places TCH, Faster Payments, and CLS in that category.
This is why the distinction matters now. Stripe, Visa, and Mastercard are leaning into stablecoins as a path to faster settlement. Banks are backing tokenized deposits, aiming for many of the same around-the-clock benefits while keeping the liquidity advantages of established clearing design.
How netting turned payment plumbing into a capital-efficiency machine
Payment once meant moving physical cash or paper records from one place to another. A debt was settled only when the money or the instrument arrived. Banks developed netting to avoid shifting every gross payment in full. If one bank owed $10 million to the rest of the system but was also owed $9 million, only the $1 million difference needed to be moved through a central operator.
Over time, that accounting shortcut became software. Banks built liquidity-saving mechanisms, or LSMs, that queue payments, search for offsetting flows, and settle only the residual amount that cannot be canceled out.
Taylor points to CHIPS, operated by The Clearing House, as the clearest example. Large banks use CHIPS for high-value dollar payments. Each transaction enters a central queue, where a proprietary algorithm continuously looks for offsetting positions. A $50 million outgoing payment from one bank matched against a $48 million incoming payment to another means only the difference touches actual funds. Settlement still happens within seconds, but liquidity barely moves.
The numbers cited in the article are stark:
- CHIPS settles about $2 trillion in average daily payment value.
- It does so with roughly $96 billion of prefunded liquidity.
- TCH estimates that running the same flows on a transaction-by-transaction gross basis would require $442 billion.
- That works out to 26:1 liquidity efficiency.
- Other major high-value payment systems average 6.6:1.
- TCH estimates CHIPS participants save $5.5 billion a year.
CLS is presented as the foreign-exchange version of the same idea. It settles more than $8 trillion a day across 18 currencies on average, and multilateral netting cuts the funding members must provide by more than 96%, according to the article. On a record day, CLS used $72 billion of actual funding to settle $19.1 trillion in payment instructions. That is 0.38%, or $380,000 in funding for every $100 million settled.
RTP shows the opposite tradeoff. The Clearing House also runs RTP, a U.S. instant-payments rail that settles each payment individually by debiting a prefunded account at the New York Fed. To get immediacy, RTP gives up netting. Taylor says stablecoins have made the same choice.
Where stablecoins gained ground
Netting works when flows can offset one another. In large parts of the world, and in large parts of the day, the pipes required for that simply were never built.
The article lists several areas where stablecoins have filled gaps left by the banking system:
- Global South to Global South corridors. CLS covers 18 currencies. If a currency sits outside that set, foreign exchange settles with more risk. Taylor cites Bank for International Settlements research saying about one-third of global FX turnover, more than $2 trillion a day, settles without payment-versus-payment protection.
- Weekends. Fedwire is closed on Saturday. CLS and Europe’s T2 shut for the full weekend. The article says trade.xyz settled $1.5 billion of institutional notional trades on USDC margin over an April weekend while bank rails were inactive. Fedwire’s six-day expansion is not expected before 2028, and Saturday is still excluded.
- Merchant settlement. Sell on Saturday morning, get paid on Saturday morning, restock on Saturday afternoon.
- Global dollars. If a business earns revenue in a currency it does not trust, a dollar instrument that can be held and transferred without a U.S. bank account becomes economically meaningful.
Taylor draws on comments from Guillermo Goncalvez, CEO of El Dorado, on the Tokenized podcast. Goncalvez described customers in Bolivia, Paraguay, and Ecuador, economies that depend heavily on imports and treat China as a major trade partner. Multinational firms can route payments through U.S. subsidiaries, he said, but small importers of electric vehicles, medical equipment, and basic materials do not have that option.
According to the article, those smaller businesses now pay Chinese suppliers with stablecoins. Goncalvez said Toyota Bolivia accepts USDT directly, five or six of Bolivia’s largest banks offer USDT wallets during weekday business hours, and El Dorado has opened a physical office in Bolivia because small importers want a place to ask questions and complete customs paperwork.
He said: "A few months ago I stopped going to crypto conferences. Now I go to coffee union meetings, car show meetings in Paraguay and Bolivia, agriculture meetings in Argentina... These customers do not necessarily care whether we use USDC or USDT. They just want to complete payments to China at the lowest possible cost."
Taylor reduces the distinction to a simple framework. Stablecoins are moving money. If you have a compatible wallet, they can travel anywhere at any time. Deposits are stationary money. They live inside banks and need a trusted clearing intermediary before another bank will accept them.
The hidden cost of instant settlement
Stablecoins settle in seconds, but only if the sender already holds the full amount. Taylor argues that this prefunding requirement is not unique to crypto. It applies to almost every form of instant payment.
He also notes that under the GENIUS Act, stablecoin issuers are barred from paying yield on float, citing Section 4(a)(11). In that setup, the fiat sitting underneath a rapidly moving stablecoin usually remains locked in place, often through weekends, without generating the kind of liquidity efficiency seen in systems like CLS.
The article gives a stylized example. A bank moving $1 billion a day through a corridor with a two-day settlement cycle would have about $2 billion parked under gross settlement. At a 4% opportunity cost, that represents $80 million a year in friction. If netting compresses gross liquidity by 90%, the bank only needs to fund $200 million over the same cycle. Same payments. One-tenth of the capital.
Taylor frames the difference this way: netting shrinks how much you owe, while speed shrinks how long you owe it. Stablecoins, as used today, settle the gross number instantly and ask you to prefund it.
He adds that industry estimates put global correspondent account balances at $15 trillion or more, with some figures as high as $27 trillion depending on methodology. Even after netting, liquidity-saving, and compression, that is still a vast amount of capital whose full-time job is to wait for rails to open.
Can stablecoins be netted?
Taylor says both sides of the argument may be underestimating the next phase. Many bank insiders say stablecoins cannot net and will lose on efficiency. Many crypto insiders treat netting as a workaround for slow rails and see no reason to care. He expects both camps to be surprised.
His view is that stablecoins and tokenized deposits are both payment promises issued by someone. They are negotiable instruments. Once enough obligations are gathered into one system, those obligations can be netted, regardless of whether the record sits on paper, in a database, on clay tablets, or in tokens.
He writes: "Stablecoins and tokenized deposits have in common that they are just payment promises from an issuer. They are negotiable instruments. Whenever you put a bunch of IOUs in a jar, you can net those obligations. You can demonstrate this at any party: have 10 people write each other a pile of IOUs and throw them in the middle. Then calculate that the net payments are trivial. That is how CHIPS and every other clearing system works. It does not matter whether the IOUs are made of paper, database entries, clay tablets, or tokens."
In his argument, prefunding is not a law of the universe. It is what happens when there is no central intermediary such as a clearinghouse. Nothing about tokens prevents queueing, matching engines, or net settlement cycles. A blockchain is a shared ledger with programmable rules, which makes it a strong candidate for netting infrastructure if someone writes the right smart contracts.
The harder problem is money equivalence. One dollar has to equal another dollar. Taylor says that is not true across stablecoins today. USDG, USDC, and newer coins such as KlarnaUSD and MoneyGram’s MGUSD do not trade into one another frictionlessly at par. He notes that USDC itself fluctuates between $0.9995 and $1.00013. Nor can all of these instruments be redeemed back into bank dollars without friction. And for now, there is no mature netting system sitting above them.
He says a clearinghouse effectively does three things:
- makes different forms of money interchangeable,
- makes them redeemable,
- nets fund flows among members.
The token-era clearing stack is beginning to emerge
The article points to several companies trying to rebuild those functions for tokenized money.
Better Money Company
Taylor describes Better Money Company as a clearinghouse built for par exchange across compliant stablecoins. It was founded by former a16z investor Sam Broner and raised a $10 million seed round led by a16z crypto. Paxos, Bridge, MoonPay, and other partners are already live, with support for settlement across 15 assets.
Ubyx
Ubyx was founded by former Citi payments executive Tony McLaughlin. Taylor describes it as a clearing system where banks and fintech firms can redeem any supported stablecoin into ordinary accounts at par under a shared rulebook. Ripple, Paxos, and more than a dozen other issuers have signed on, according to the article. Barclays has invested, and BitGo serves as settlement agent.
Glacis Labs
Glacis Labs built ZeroDelta, a multichain clearing layer that matches and nets digital-asset transfers across more than 40 chains before settling only the net amount on-chain. Taylor says it has cleared more than $1 billion so far. Six days before he wrote the article, the company announced a $6.8 million seed round led by Lightspeed Faction, with participation from Franklin Templeton and Coinbase Ventures.
Cycles
Cycles takes a more aggressive approach. It maps who owes whom across the network and removes circular obligations. If A owes B, B owes C, and C owes A, that triangle can be cleared without additional liquidity. The company was founded by Cosmos co-founder Ethan Buchman, who calls clearing a "financial superpower." The article says it has just raised $6.4 million.
Circle Payments Network
Even Circle is moving in this direction, Taylor argues. Circle Payments Network, or CPN, is a closed permissioned network where member banks and fintech firms settle directly in USDC under Circle’s rules on a 24/7 basis. In his view, that already looks like a clearing club. The missing brick is a netting layer, and several players are now trying to build it.
Why the big U.S. banks chose tokenized deposits
Taylor brings the discussion back to the original question: why would The Clearing House not just use stablecoins?
On June 5, 17 of the largest U.S. banks, including JPMorgan, Citigroup, Bank of America, and Wells Fargo, said The Clearing House would build on-chain clearing and settlement for tokenized deposits, plus a connectivity layer tied to RTP and CHIPS, operating around the clock.
That gives banks 24/7 fund transfer capability and the 26:1 netting ratio the article attributes to CHIPS, while staying compatible with the systems they already run. In that sense, the move is logical. Banks do not need stablecoins to achieve this because TCH has acted as an interbank intermediary for clearing and accepting different banks’ deposits since 1853.
Taylor’s reading is blunt. Banks looked at stablecoins and saw gross, prefunded settlement. They already had something more efficient, and interoperability among themselves was solved long ago. For current interbank flows, especially large wholesale payments, tokenized deposits connected to existing clearing rails amount to a substantial upgrade.
That said, he does not treat tokenized deposits as a universal answer. TCH is large, but it is not the only ledger that matters. Stablecoins can move across clearing networks and payment rails to any wallet running compatible software. They do not require The Clearing House, the U.K.’s Faster Payments, India’s UPI, CLS, and many other institutions to strike bilateral integrations under a single stablecoin rulebook.
As he puts it, stablecoins work and they travel. Compliance is still unresolved in places, and final stablecoin rules under GENIUS are still pending. Even so, he expects both stablecoins and tokenized deposits to become part of a 24/7 money loop.
The race is toward instant and net settlement
Taylor closes by splitting payment design into four boxes:
- slow and gross, as in correspondent banking before CLS, where everything sits for days;
- instant and gross, as in today’s stablecoins and, in his framing, RTP, where all funding is parked upfront for a short period;
- slow and net, as in CHIPS and CLS, where little funding is parked but operating hours remain constrained;
- instant and net, the destination he sees as the end state, where little funding is parked and the system runs continuously.
He argues that nearly everyone is now moving toward the fourth square from different starting points. Large banks are extending their existing model into a 24/7 environment. Stablecoin startups are trying to add clearing and liquidity-saving layers to tokenized money. He says much of his own day job at Tempo is focused on how to build a settlement chain and smart contracts with embedded liquidity-saving mechanisms for deposits, central bank money, or stablecoins.
That is why the opening question, instant payment or two-day payment, turns out to be the wrong framing in his view. Treasurers were never against instant settlement. Merchants were never against efficiency. The winning rail, he argues, will be the one that settles small net amounts quickly, anywhere and at any time.
He also adds one final note on correspondent balances: estimates for nostro and vostro accounts range from $15 trillion to $27 trillion depending on who is counting. The figure is directional, not definitive.