Stablecoins in Business Payments and Treasury

Lesson 1 explained what a stablecoin is and why it holds a dollar. The obvious follow-up: so what? A dollar that moves on a blockchain is only interesting if moving dollars is currently a problem.

Domestically, it mostly is not. Paying a supplier in your own country is cheap and fast enough that no business is desperate for an alternative.

Cross-border is a different story. Sending money between countries is slow (often days), expensive (fees plus an FX spread), opaque (you frequently cannot see where the payment is), and constrained by banking hours and cut-off times. For a business paying overseas suppliers, running international payroll, or managing cash across subsidiaries, this is a real and permanent cost.

This lesson covers:

• Why cross-border payments are structurally slow, which is the necessary background.
• What stablecoins actually change, and the one property that matters most.
• Treasury effects, including the capital that today sits trapped.
• Tokenized treasuries and deposits, the adjacent developments.
• How to read the volume numbers, which are quoted constantly and mostly misunderstood.

The framing to hold: the interesting claim is not that stablecoins are digital. Bank money has been digital for decades. The claim is about what settlement means, and that requires understanding what happens today.

The slowness is not incompetence. It is structural, and the structure has a name: correspondent banking.

Start with the fact that surprises people: there is no global payment network. Your bank in one country has no direct relationship with a supplier's bank in another. So payments travel through a chain of intermediaries, each of which has an account with the next.

Those are nostro and vostro accounts, the same account seen from two sides: my account at your bank, your account at mine. A payment does not really "travel" at all. Instead, each bank in the chain debits one account and credits another, and the money is passed along by a sequence of bookkeeping entries between institutions that trust each other.

Two consequences follow, and they explain the cost:

• Every hop adds time, fee, and opacity. Each intermediary takes a cut and a delay, and each is a place your payment can stall for compliance review. A three-hop payment has three chances to sit for a day.
• The chain requires pre-funded money everywhere. For a bank to pay out on your behalf in another country, it must already hold cash there, sitting idle in a nostro account, waiting.

And the piece people most often get wrong: SWIFT is messaging, not settlement. SWIFT carries the instruction between banks, telling them what to do. It does not move the money. The money moves through those account entries. Understanding this is essential, because it tells you exactly what a new rail must replace, and what it must not bother replacing.

Now the claim can be stated precisely. With a stablecoin, the transfer is the settlement.

In correspondent banking, a payment is an instruction that a chain of banks acts on, with final settlement arriving later through their mutual accounts. With a stablecoin, moving the token is the movement of value. There is no chain of intermediaries passing entries along, no message that must be reconciled with a later settlement. The asset itself changes hands.

From that single property, the practical benefits follow mechanically:

• Speed. Settlement in minutes, because there is no chain to traverse.
• Always on. Blockchains do not observe banking hours, weekends, or cut-off times. A payment at 2am Sunday settles at 2am Sunday.
• No pre-funding in every corridor. You do not need idle cash parked in a foreign nostro account to pay someone there, which frees working capital.
• Fewer intermediaries, fewer fees. Each removed hop removes a cut and a delay.
• Programmable. The payment is software, so it can carry conditions and be automated directly.
• Transparent. You can see the transfer, rather than wondering which correspondent is holding it.

But be precise about the boundary. Stablecoins do not replace SWIFT's job, they replace the settlement layer underneath it, and remove the need for the messaging in the first place by making instruction and settlement the same act.

And the honest caveat, which Lesson 3 develops: the on-chain leg is fast, but getting in and out of the banking system at each end is not. The fast middle is bracketed by two slow ends.

Make it concrete. A company in the US owes a supplier in Southeast Asia 250,000 dollars.

The correspondent route:

Thu 16:10  Payment initiated, just after the cut-off
Fri 09:00  US bank debits account, sends instruction
Fri --     Hops to a correspondent, then another
Mon --     Compliance review at an intermediary
Tue --     Beneficiary bank credits supplier
Cost:      wire fees at multiple hops + FX spread
Visible?   No. "It has left us."
Pre-funded Yes. Someone holds idle cash in that corridor.

The stablecoin route:

Thu 16:10  Treasury converts USD to stablecoin (on-ramp)
Thu 16:15  Transfer settles on-chain
Thu 16:20  Supplier holds the tokens; may hold or off-ramp
Cost:      network fee + on/off-ramp spreads
Visible?   Yes, confirmed and auditable
Pre-funded No corridor float required

Three things deserve attention. First, the cut-off did not matter. In the bank route, arriving at 16:10 on a Thursday cost days; on-chain the clock is irrelevant. Second, the cost moved rather than vanished. You stop paying wire fees and correspondent cuts, and you start paying network fees and on/off-ramp spreads. Whether that is cheaper depends on the corridor, the size, and the provider. Third, the supplier now holds a claim on an issuer (Lesson 1), not a bank deposit. Whether they are content to is a real question, not a technicality.

The honest summary: the gain is largest where the old route is worst, exotic corridors, weekends, urgency. On a well-served corridor between two big banks, the advantage narrows.

For a large business the more interesting effect is not a single payment. It is treasury, the management of cash across the whole company.

Corporate treasurers spend their lives on a small set of problems, and stablecoins touch several directly:

• Trapped liquidity. To pay quickly in a currency or country, you must pre-fund there. Multiply that across dozens of corridors and a significant sum sits idle, purely as positioning. Settlement that does not require corridor pre-funding releases that capital, and it was earning nothing.
• Settlement timing. Money in flight for days is money you cannot use or account for. Shortening the float improves working capital and forecasting.
• Weekend and holiday gaps. Cash needs do not respect the banking calendar. A 24/7 rail removes a class of timing problems entirely.
• Cost per transaction. This is where scale bites: shaving settlement times, and even a few basis points off each transaction, is significant for a company moving billions annually. A basis point is trivial once and material a million times.

Notice the pattern. The treasury case is about capital efficiency, not novelty. Nothing here requires believing anything about crypto. It is the same argument a treasurer would make about any faster settlement rail, and it is why the enterprise interest is coming from finance departments rather than technologists.

Which also predicts where adoption lands first: businesses with many corridors, thin margins, and urgent settlement. Cross-border marketplaces, remittance providers, commodity traders, global payroll. Where the old rails are worst, the case is strongest.

Two adjacent developments are often confused with stablecoins and are genuinely different.

Tokenization of real-world assets (RWA) is putting claims on conventional financial assets on-chain. The activity has concentrated in the most cash-like instruments: on-chain representations of cash, treasuries, and money market instruments crossed 36 billion dollars in 2025.

That concentration is informative. Tokenization has been experimented with for over a decade across every asset class, and what actually gained traction is the boring end: government paper and money funds. The reason connects to Lesson 1, these are the assets where an on-chain claim is nearly as good as the asset, because the asset is simple, liquid, and its value is not in dispute. A tokenized bond is useful. A tokenized painting still has a painting problem.

It also creates an obvious complement: a treasurer can hold value on-chain in something yield-bearing rather than in a stablecoin, which, as Lesson 3 explains, is legally barred from paying them interest.

Tokenized deposits are the other one, and the distinction is sharp. A stablecoin is a claim on an issuer, backed by segregated reserves. A tokenized deposit is a claim on your bank, an ordinary bank deposit represented on-chain, still inside the banking system, still the bank's liability, still covered by the regime that covers deposits.

So the real question for a business is not "crypto or not." It is whose liability do you want to hold: a reserve-backed issuer, or your bank?

You will constantly meet a headline statistic, and it needs careful handling. Stablecoins are reported to have accounted for an estimated 46 trillion dollars in transaction volume in a recent year, described by a16z crypto as more than 20 times PayPal's volume and close to 3 times Visa's.

The number is real. The implied conclusion is not. On-chain transaction volume is not payment volume, and the gap is enormous:

• Trading and arbitrage dominate. Most stablecoin movement is people moving value between exchanges and trading strategies, not buying goods.
• Automated activity inflates counts. Bots, market makers, and liquidations transact constantly, at machine frequency.
• One economic payment can be many transfers. Moving through wallets, bridges, and contracts is counted at each step.
• A dollar is counted every time it moves. Card networks measure purchases; a chain measures transfers.

So comparing 46 trillion to Visa is comparing categories, not competitors. Visa's figure is people buying things. The stablecoin figure is everything, mostly financial plumbing.

This is not a reason to dismiss stablecoins. It is a reason to evaluate them on the mechanism rather than the scoreboard. The genuinely strong claim from this lesson is narrow and defensible: settlement that is the transfer itself, available 24/7, without corridor pre-funding, is a real improvement over correspondent banking for specific use cases.

That claim needs no trillion-dollar statistic, and it survives the scrutiny the statistic does not. When a number sounds too good, ask what is being counted, exactly the discipline the numbers in Lesson 3's regulatory picture do not require, because rules are checkable.

A cross-border bank payment traverses a chain of correspondents passing bookkeeping entries, each adding time, fee, and compliance stalls, and requires pre-funded nostro accounts; a stablecoin transfer settles directly, though it is bracketed by on-ramp and off-ramp into the banking system.