🧠 The Stablecoin Business Case for Banks.

There’s been a subtle vibe shift in the banks over the past three months. From loudly proclaiming that tokenized deposits are better than stablecoins, to increasingly benefiting from stablecoins' coexistence with tokenized deposits.

This was clear when I shared the stage with Sabih Behzad from Deutsche Bank on stage at Stablecon EMEA. Clients are asking banks for support to off-ramp stablecoins, access exotic currency pairs, and begin interoperating with them. Moreover, banks like Coastal Community Bank are now using stablecoins instead of correspondent banking for some flows. (Let that thought marinade if you’re a community or regional banker reading this.)

The key insight that I’ve been banging the drum about on Tokenized for the past month is simply this:

Tokenized deposits are money that rests. Stablecoins are money that moves.

Tokenized deposits cannot exist outside the bank, but can attract yield (easier), have no off-ramping issues, and handle unlimited volume. Typically, large corporations also receive preferential lending rates for parking large amounts of deposits with a bank. Stablecoins can exist on any platform with a compatible wallet, in any country, instantly. They’re more open-loop and not stuck inside the walls of one organization.

Some banks have spotted the opportunity on different sides of this table

JP Morgan Kinexys, Citi Token Services, and equivalents from GSIBs are incredible products; Kinexys alone has processed over $3 trillion cumulatively since 2020, dwarfing the estimated $390bn in annualized stablecoin payments. They offer instant, 24/7 transfers to 60+ countries, and for most large companies, major currency pairs (like EUR/USD) are the best and most price-effective choice.

But GSIBs are also increasingly on the stablecoin side of the table. Standard Chartered provides custody, off-ramping, and stablecoin card infrastructure. SocGen launched MiCA-compliant EUR and USD stablecoins, and Deutsche Bank clients are asking for stablecoin interoperability. The most innovative though are the smaller banks like Lead, Coastal, Cross River, Itau, Banking Circle, and LHV. They have chosen to support stablecoins as off-ramps and local payments partners. For both sizes of banks, stablecoins are another revenue-generating payment opportunity.

These are just some of the early movers. In the next two years, I predict stablecoins will become a major revenue driver for banks.

In the US, the GENIUS Act was passed, and regulatory rulemaking is expected in July. In the interim, the debate about “can stablecoins pay yield” has shifted to the CLARITY Act designed to provide the market structure for all tokenized assets (and define what the SEC vs CFTC does, for example). 

Without replaying the entire, exhausting, and intellectually dishonest debate, the CLARITY Act cleared the Senate Banking Committee 15-9 last week and is heading to the floor

But before the act is passed, the active lobbying by both banks and the crypto industry is masking a deeper shift in public sentiment. Tokenization and digital assets are now here, and while the regulation will take some time to settle, most large banks are simultaneously figuring out what role they need to play.

My fear here is that the regional and smaller banks see the lobbying activity and genuinely buy into the deposit flight risk. 

To which I ask: If Coinbase has been offering yield on stablecoins for the past three years, where was all the deposit flight then? 

We need to reframe this debate from one of risk to one of opportunity. 

The smartest and most obvious opportunity is the one many banks have chosen, which is to serve the existing demand of the stablecoin industry across several use cases.

This provides banks with immediate, accretive revenue by servicing the existing demand for fiat on/off-ramping and card support. Expect this category to expand increasingly into the regionals and GSIBs, many of which already count some of the largest crypto companies as clients.

If you’re a bank outside the top 20, chances are you dislike correspondent banking as much as most consumers and SMBs do. Non-GSIBs lack direct access to foreign central bank clearing systems (like Fedwire for USD or Target2 for EUR). To offer international services, they must maintain Nostro accounts (foreign currency accounts) with GSIBs.

Now, yes, they can markup those fees to their clients, and add FX spread, but that makes them less competitive, their costs are high, and their margins are thin.

But what if they used stablecoins instead? 

It would look something like this:

Far from being some future state, Coastal Community Bank is going live with this product.

The insight here is that most banks are sending payment messages today for cross-border payments. These payment messages help banks on both sides run all of their pre-payment compliance checks. With stablecoin wallets, the banks would still send a payment message, but instead of routing the settlement via correspondent banking Nostro accounts, they’d instead settle with a stablecoin wallet transfer.

This would work 24/7, and be especially beneficial to sponsor banks, who have large PSPs, fintech or crypto clients who want to offer that capability to their clients. These smaller banks would be price-competitive and offer arguably a more compelling product.

Of course, if the small banks do this, the big banks could equally find routes and corridors where their volume and pricing power are advantageous. I suspect, however, that the institutions will start with another use case. Off-hours cash.

Fiat doesn't move at weekends, but prices do.

When the Iran conflict escalated in March, oil spiked 30% over a Saturday. Traditional exchanges were closed. CLS was closed, FedWire was closed, and the exchanges' traders couldn’t settle trades. Traders piled into trade.xyz on Hyperliquid, a DeFi network offering 24/7 oil futures, which hit $1 billion in weekend volume. JPMorgan's own analysts flagged non-crypto traders moving to the platform.

That should terrify every GSIB. It should also excite them.

Kinexys and Citi Token Services can't solve this — they're closed-loop. Central bank money with cross-network tokenized deposit interoperability is plausibly ten to fifteen years out. But stablecoins are open-loop now.

The same bilateral settlement method that works for smaller banks in the correspondent banking use case works equally well for GSIBs, but as a way to open up larger corridors over weekends. Think of it like a semi-open-loop Kinexys. A bank-issued stablecoin "sleeve" scoped to institutional clients and weekend hours, where the deposits backing it are wholesale balances that are already the most expensive on the book. The fees banks could charge for off-hours settlement (weekend FX typically commands 5–20bps) would more than cover the balance sheet cost.

(Send that to your ALCO bankers)

If tokenized deposits are stuck inside the bank, and stablecoins can go anywhere outside but without the same protections and 1:1 redemption. Surely there’s an opportunity for a best of both worlds scenario.

SoFi just made this real inside a single bank. 

SoFi Bank is nationally chartered, OCC-regulated,  and launched SoFiUSD to its nearly 15 million members. The important part about the stablecoin is that reserves are held 1:1 at the Fed, and they’re available to all SoFi members alongside tokenized deposits. (More on this in 👀 Things to Know.)

When you have tokenized deposits and stablecoins, you can swap between them 24/7/365, redeem 1:1 for USD. This solves the traditional tradeoff; you could have fiat, but it didn’t work on weekends or bank holidays, especially internationally. You could use stablecoins, but they never traded at par and weren’t as widely accepted as deposits.

The obvious answer is to combine the two. 

I expect every bank will copy this within the coming years, because the demand for 24*7 is now irresistible.

I started writing this piece because I realized most of the conversation about stablecoins is a threat, and nobody was loudly making the stablecoin business case for banks. Some banks, some of the time, are doing interesting things. Local payments access, correspondent banking replacement, and weekend settlement are all revenue-positive. None of them requires offering stablecoins to consumers. None of them triggers the deposit flight that the yield debate keeps warning about.

Coinbase has been offering 4%+ on USDC for three years. Robinhood pays a yield on idle cash. Where's the deposit flight? It didn't happen. The customers who wanted yield have already left. The customers who stayed? Stayed for the relationship. The opportunity to build new revenue on the infrastructure your clients are already asking for.

The obvious gap I haven't addressed in this piece is compliance. Banks adding stablecoin capabilities will need wallet screening, chain analytics, and travel rule tooling alongside their existing AML stack. The tools exist, and the operational lift is manageable, but it's real, and any honest business case includes it (what can I say, I ran out of time).

But my question to you, as a bank or a bank partner, is:

The business cases are here. They're accretive. They work.

Which one will you pick?

ST.

PS. I've launched an advisory function at Tempo. We help banks (and non-banks) put together the full business case, operating model, and GTM plan. Drop a line if we can help.

1. Trade Prophet - Catalyst hedging for institutions

Trade Prophet lets institutions buy and sell exposure to specific binary outcomes like FDA approval decisions, tariff rulings, and M&A close/no-close. The product sits between hedge funds, pharma companies, and importers who want to hedge a specific event, and counterparties willing to take the other side. Think of it as bespoke event derivatives without going through a Goldman desk or prediction market.

🧠 Hedge funds and pharma desks already price these probabilities; what they lack is a clean way to transfer the risk. But the question I have is, who’s the counterparty? Who actually wants to be the seller on "FDA rejects this drug"? Kalshi and Polymarket have volume-building order books. Trade Prophet is standardizing the wrapper for much more niche events. But can they get sellers and buyers?

2. Rex - AI agents for finance ops, starting with AR

Rex puts AI agents inside enterprise finance teams, beginning with accounts receivable. The agents chase invoices, run dunning, reconcile payments, and follow up with customers; they log into existing ERPs the way a human collections analyst would. The plan is to expand across the rest of the finance function over time.

🧠 AR is the smartest wedge in finance ops because the ROI is cash, not productivity. But the question I have is, is this a company or the next feature inside Ramp? The field is already loud: Tesorio, HighRadius, Tabs, plus whatever Brex and Stripe Billing ship next. So what does Rex have that the spend management platforms don't?

3. ZeroDrift - The Compliance Firewall “for the AI era.”

ZeroDrift sits between an employee's outbound messages across email, browser, CRM, social, AI tools, and the outside world, validating each one against SEC/FINRA rules and firm-specific policies before it gets sent. Most compliance tooling reviews messages after the fact and triggers alerts; ZeroDrift's bet is to block non-compliant content pre-send. They're starting in financial services with RIAs, asset managers, broker-dealers and wealth platforms, then expanding into insurance, healthcare, ESG and AI governance.

🧠 The "AI era" framing is doing real work here. Generative AI has turned every junior associate into a content producer at scale, and the existing compliance stack (Smarsh, Global Relay, Theta Lake) is built to surveil after the fact, not block in flight. Pre-send blocking is a different bet, and a riskier one, because every false positive is a sales rep telling their CCO this tool is a productivity tax. What's the false positive rate at production volume, and how much friction does this actually add to client outreach?

4. Bilrost - AI for commercial real estate loan processing

Bilrost automates the document-heavy parts of commercial real estate lending, like classifying and extracting fields from rent rolls, T-12s, OMs, and leases. It then pulls those outputs into the bank's loan origination system. They claim to compress a 90-hour underwriting workflow into under 3 minutes, with field-level lineage and immutable logs for audit. Five weeks after launch, they say they're already in pilot with 5 of the top 10 US banks.

🧠 CRE document extraction has been a graveyard for startups. Blooma, Cherre, Ocrolus, and a long list of others have all swung at this. The pitch is usually the same: OCR / machine learning will finally crack the unstructured chaos of a CRE deal package. The genuinely new bet is that LLM-grade reasoning over noisy docs is materially better than 2018-era OCR plus machine learning. Five top-10 banks in five weeks is either a real signal or death by pilot. But banks are adopting AI at a real clip.

1. SoFi launches SoFi USD stablecoin to 15m members.

SoFi just became the first national US bank to launch a stablecoin to its entire user base that is 1:1 redeemable for USD. The stablecoin SoFiUSD is issued by SoFi Bank, a nationally chartered, OCC-regulated US bank. With reserves are held 1:1 in cash at the Federal Reserve. The stablecoin is now live on Ethereum and Solana to its nearly 15 million members 

🧠The stablecoin is available via the regular mobile app (i.e., not some white-label partnership with Coinbase). SoFi is taking a bigger role in custody, management, and fully integrating with their core. 

🧠 Pay attention to the fact that SoFi offers a stablecoin and a tokenized deposit, side by side. Why? Because stablecoins move 24*7, globally to any compatible wallet, but without interest or FDIC insurance. Tokenized deposits can only stay inside SoFi walls, and earn interest. 

🧠 When you swap tokenized deposits for stablecoins, 24*7*365, and redeem 1:1 for USD. This does two things.

🧠 My question: How wil SoFi drive adoption beyond its membership? If a non SoFi customer received SoFi USD, they're likely to want to swap that pretty quickly. You could imagine other banks wanting access to the stablecoin to defend deposits, and offer a new correspondent banking rail. 

🧠 Also what are the edge cases here? The OCC rulemaking for stablecoins for banks is still not final. Things could yet go wrong, and, there's always a risk that the sins of today will be punished by a future administration. 

🧠 Still, kudos to SoFi here. I'm sure they've not taken these steps lightly, and this product is market-leading and much-needed.

🧠 Do not underestimate how much is happening in tokenization. SoFi is just the first, but many, many others will follow. Institutional-grade tokenization is coming, bringing the promise of true 24*7 transfers, trading, and settlement. That is monumental. As someone with 20+ years in payments, that is NOT hyperbole

2. Robinhood is letting its users' AI agents trade stocks.

Users create a separate brokerage account, load a balance into a dedicated wallet (virtual card), and connect their AI agent via MCP. The agent can analyze portfolios, assess concentration risk, read analyst notes, and execute stock trades. Gold Card holders can connect their AI agent to Robinhood's banking MCP server and let it make payments to buy stocks, which means you get monthly spend limits and optional approval gates.

🧠 Imagine being Fidelity or Vanguard watching this. There’s no way you could ship a competitive product in the next 6 months. The compliance surface area alone is enormous. Robinhood gets to move fast because they built for it (and frankly, have a wider risk appetite).

🧠 Sounds risky, but they’ve limited the surface area. Letting agents, who hallucinate, trade? Are you mad! Well, it’s only in beta, only for stocks, all managed on a virtual card, and Robinhood says they’ll manage disputes or fraud risks.

🧠 There’s a segment of tech-forward users who will love this. I have Claude code scrape together all kinds of admin for me. Always with a human QA, but the speed them can move, and the surfaces they can operate across (like email, documents) makes them super useful. Those beta users of today are helping build infrastructure that everyone will use in 5 years.

3. Cash App has gone live with USDC

Customers can now send, receive, or hold stablecoins on Ethereum, Solana, Arbitrum, and Polygon. When received, stablecoins will instantly be converted 1:1 to USD. Stablecoin transfers are free (for a limited time), but are not available to New York residents. Cash App also noted that outbound sends to incorrect wallet addresses will not be refunded, and funds may be lost.

🧠 Love that they’re converting 1:1 USD, that should be the user experience default and slowly is becoming that (SoFi and Ramp do the same). Odd that sends to unsupported networks are non-refundable, surely they could block that feature in the app? Seems like a bad UX? I’m surprised too, that they’re not sponsoring fees with account abstraction. The blog post says using stablecoins is “a step closer to bitcoin,” which seems like unnecessary cope at this point.

1. Bye the Index

Nasdaq’s flywheel used to be that IPOs drove its index business QQQ, launched on its sister venue the American Stock Exchange. “QQQ now sits on $456 billion of net assets, making it the fifth largest exchange-traded fund in the market.” It spent $749m on marketing over 5 years, with the fund earning $854m annually for Nasdaq. Now as listings flounder, Nasdaq is trying to reverse the flywheel. But now companies can join the index in a matter of weeks if they’re large enough. 

They’re also changing how the indexing dates work, so a company could list less shares, see more demand (from lower supply), and potentially then grow in value and weighting in the index. S&P is countering with similar moves, but has a human backstop that kept Microstrategy out of the index and delayed Tesla. 

🧠How the index works now determines when, and where you list. Given the rebalancing of the Nasdaq in June, guess when SpaceX is due to IPO? Yep. June.

🧠With SpaceX, Anthropic, and OpenAI about to list, the indexes are about to get tested. If Antropic hits $100bn this year and can keep anything like that growth next year, by mid-2027, they’d be among Apple and NVIDIA level revenues.

Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)

Want more? I also run the Tokenized podcast and newsletter.

(1) All content and views expressed here are the authors' personal opinions and do not reflect the views of any of their employers or employees.

(2) All companies or assets mentioned by the author in which the author has a personal and/or financial interest are denoted with a *. None of the above constitutes investment advice, and you should seek independent advice before making any investment decisions.

(3) Any companies mentioned are top of mind and used for illustrative purposes only.

(4) A team of researchers has not rigorously fact-checked this. Please don't take it as gospel—strong opinions weakly held

(5) Citations may be missing, and I’ve done my best to cite, but I will always aim to update and correct the live version where possible. If I cited you and got the referencing wrong, please reach out