If you're reading this and still haven't signed up, click the subscribe button below!

This week Simon Taylor & Cuy Sheffield are joined by:

🎙️ Listen to the latest episode of Tokenized here.

📷 Watch on YouTube.

This episode was sponsored by Modern Treasury!

Trusted by companies like Procore, Navan, and Morse, and backed by over $600 billion in payments, learn how to adapt to changing payment rails and scale with confidence at moderntreasury.com

Rain published a detailed breakdown of what happens after a customer pays with a stablecoin card. The post argued that settlement still runs on banking hours, forcing issuers to prefund several days of outflows around weekends and holidays. The funding problem is real, but it may not apply equally across the issuer market. Simon illustrated Rain’s claim with a hypothetical program settling $1 million a day. In his example, the program could need $4 million available to cover a long weekend.

Simon also summarised the pushback Rain’s post received. Critics argued that large issuers benefit from netting, earn interest on their balances and operate successfully in capped-interchange markets such as the EU. Cuy agreed with the narrower point - for an investment-grade US bank, settlement may not be the constraint Rain’s argument implies.

Cuy Sheffield split the market by size and credit quality. Smaller BIN sponsors, fintechs and stablecoin-native neobanks do not receive the same collateral and credit terms. More frequent settlement matters more to them because it reduces how much liquidity must sit idle before a card program can scale. Over time, Cuy also expects banks, particularly those outside the US, to use stablecoin settlement.

Simon separated the customers each side had in mind. The online debate concerned large banks with scale and working economics, while Rain focused on the customers that model leaves behind.

Augustus is trying to become one of the first new national banks built for both conventional dollar clearing and stablecoins. The OCC granted conditional approval in May 2026, followed by FDIC insurance approval on July 31.

Ferdinand Dabitz's case for Augustus starts with a gap in the fintech record. Every other layer of the bank stack got challenged; dollar clearing did not, because a charter and direct Fed access were far harder to obtain than a banking app.

His answer for why that is changing now is the de novo charter wave. A handful of new national banks are being approved with the one thing a fintech partnership never provided.

The charter provides access, while stablecoins target the balance-sheet cost that comes with correspondent banking. Banks pre-position cash across nostro and vostro accounts to keep payments moving between jurisdictions and time zones. Ferdinand sees stablecoin rebalancing as a way to release some of that trapped liquidity without removing the regulated bank from the transaction.

Anna framed the underlying cost plainly.

Augustus now sits further along in the OCC charter and Fed master-account race than it did in July, though it still has to capitalise by May 2027 and open by November 2027. The thesis still depends on final approval, but the mechanism is clear. Regulated access gets a challenger into dollar clearing, and stablecoins give it another way to manage liquidity once it arrives.

Robinhood CEO Vlad Tenev renewed his call for US regulators to create a path for tokenized stocks, pointing to the company’s rollout in Europe. The product is fully backed, but backing is not the same as ownership. Robinhood’s own key information document for European stock tokens describe them as economic exposure without shareholder voting rights, and Tenev restated this week that holders do not directly own the underlying shares.

The regulatory route Tenev wants is still unsettled. The SEC’s innovation exemption for tokenized securities has been delayed twice amid pushback from the White House and SIFMA, leaving Robinhood to press its case publicly rather than file into an established US framework. The Senate's procedural vote on the CLARITY Act on 15 September is the next thing that moves it. If the bill advances, the exemption becomes less necessary.

Anna Wroblewska used that news to draw a line between the products being grouped together as tokenized stocks. She inferred that Tenev was seeking a US path for synthetic tokens because Dinari already offers what the SEC refers to as a custodial model.

Tenev's own framing points the other way. He said he expects regulatory progress to support tokens that eventually carry the full rights of ordinary shares, which is closer to the custodial model than to a synthetic one. Anna's distinction holds regardless of which path Robinhood is asking for.

The wrapper matters most when the token moves beyond trading. Its legal structure determines ownership, corporate actions and insolvency claims. A lender also needs to know whether it is financing a security entitlement or a contract that tracks one. Brokers and asset managers cannot sell the collateral utility without explaining the instrument underneath.

The debate started with a familiar claim - blockchains are needed because bank payment rails do not operate around the clock. Ferdinand pushed back on the premise. Fedwire clears 22 hours a day, while 3pm cutoffs often come from the correspondent-banking stack wrapped around it. The rail-level gap is weekends, when Fedwire does not clear.

That distinction changes what a bank must fix. A blockchain can extend settlement availability, but it cannot remove the organisational constraints institutions have built around their existing rails. Simon carried that argument inside the bank. A continuous subledger does not produce a continuous institution unless compliance, risk and treasury operations follow it around the clock.

Stripe’s agreement to acquire OpenRouter shows where new demand for continuous money may come from. The New York Times put the price at about $7.5 billion, although Stripe has not disclosed terms. The acquisition matters here because OpenRouter has also been testing endpoints on the Machine Payments Protocol (MPP), the agent-payments standard from Stripe and Tempo. That lets agents buy inference on demand and pay as they consume it.

The economics explain why Stripe wants both layers. Stripe acquired Metronome last year, giving it a ~2.5% take rate on token billing. OpenRouter adds a 5% markup on inference. Where both apply to the same flow, Simon calculated that Stripe’s combined take could approach 7.5%. That was an illustration, but it shows how model routing and billing can compound within the same transaction.

The operational implication runs in both directions. Cuy argued that agents making financial decisions will require payment rails that stay open. Ferdinand’s response was that AI could also help banks automate parts of the control layer once the payment infrastructure runs continuously.

That brought the conversation back to how quickly banks add stablecoins to the correspondent stack. Ferdinand put a date on bank adoption.

Correspondent banks don’t need to choose between stablecoins and the existing banking system. If stablecoins reduce trapped liquidity while preserving the same compliance and risk controls, banks will add the rail.

Ferdinand’s five-to-ten-year timeline may be closer than it sounds.

Thanks so much for reading the Tokenized Newsletter!

Please share this edition or share it with your colleagues if you enjoyed it!

Disclaimers