Foundations: what this technology actually is

The ledger

Every payment system you have ever worked with is, underneath everything, a ledger — a record of who holds what. A bank account is a row in a ledger. A payment is two ledger updates: one balance goes down, another goes up.

The reason cross-border payments are slow and expensive is not that the messages are slow. A SWIFT message arrives in seconds. The problem is that there is no single ledger both banks share. A bank in Warsaw and the beneficiary’s bank in Manila keep separate ledgers, and the only way to connect them is through chains of intermediaries holding accounts with each other — the nostro accounts you already know, prefunded with idle capital, reconciled in batches, closed on weekends.

Here’s the reframe that makes everything downstream click: correspondent banking is already a distributed trust arrangement. No single institution operates it. It’s a network of bilateral balances. It works. It’s just slow, capital-hungry, and expensive — because “distributed” here means fragmented: thousands of separate ledgers that must be laboriously synchronized.

Idea one: a blockchain is a shared settlement ledger with no central operator

Strip away the vocabulary and this is what a blockchain is to a payments person: one settlement ledger that every participant can read and write directly, operated by no single institution.

Instead of your bank updating its ledger, sending a message, and waiting for three other institutions to update theirs, everyone writes to the same ledger. When the ledger updates, that’s not a promise to settle — it is the settlement. There is no separate reconciliation step, because there is nothing to reconcile against.

Who maintains a ledger that nobody owns? A network of independent operators who validate updates and are paid fees to do so honestly, with economic penalties if they cheat. How exactly that works — the consensus machinery — we will cover in detail in the next episodes.

What matters here is the property it produces: a ledger update, once confirmed, is final, and the ledger never closes. No cut-off times, no weekends, no batch windows.

One vocabulary item before we move on. An entry on this shared ledger — a unit of value that the ledger tracks — is called a token. That’s all a token is: a ledger balance, transferable directly by whoever holds it.

Idea two: a stablecoin is e-money in a new wrapper

A raw blockchain ledger tracks its own native asset, whose price floats. Floating-price assets are useless for payments — you know this instinctively, and you are right.

A stablecoin fixes that with a structure you will recognize immediately, because it’s the structure of e-money:

• A private issuer takes in dollars (or euros).
• It holds those dollars in reserve — cash and short-dated government securities.
• It issues tokens on the shared ledger, one token per dollar received.
• It promises to redeem tokens back into dollars, one-for-one, on demand.

If you have ever assessed an e-money institution, this is the same instrument: customer funds in, safeguarded reserves held, electronic claims out, redemption at par. The novelty is not the economics — it’s where the claim lives.

Instead of sitting in the issuer’s own database, the claim sits on the shared ledger, which makes it a bearer instrument: whoever holds the token holds the claim, and can transfer it to anyone else on the ledger, directly, without touching the issuer’s systems.

This isn’t just an analogy we are drawing for pedagogical comfort. It’s how European regulation formally classifies the instrument: under MiCA, a stablecoin pegged to a single fiat currency is an e-money token (EMT), and its issuers face requirements recognizably descended from the e-money directive.

For concreteness: McKinsey’s analysis of the largest regulated issuers describes Circle’s USDC reserves as roughly 85–90% short-term US Treasuries and repurchase agreements, with the remainder in cash for immediate redemptions (as of mid-2025). The reserve composition is the whole ballgame for stability — which is why it’s the first thing regulation everywhere reaches for.

One caveat that belongs in an honest introduction, because it’s the kind of thing vendor decks omit: the legal strength of the redemption claim is not uniform. Depending on jurisdiction and issuer structure, a token holder may not have a direct legal claim on the reserve assets — in an issuer insolvency, holders could rank as unsecured creditors. Newer regulatory regimes are closing this gap; it is one of the main things they exist to close. We’ll treat it properly when we get to diligence criteria.

Idea three: mint and burn — or, where is the money?

Now the question the whole episode has been circling. Follow one dollar through its full life cycle:

Mint. A company owes $1,000,000 to a supplier on another continent — an ordinary invoice that would normally travel the correspondent chain. Instead, the company wires the $1,000,000 to a stablecoin issuer. The dollars land in the issuer’s reserve account — a real account at a real bank (or a Treasury portfolio at a custodian). In exchange, the issuer creates — mints — 1,000,000 new tokens on the shared ledger and delivers them to the company’s ledger address. The money is in the issuer’s reserves. The token is a claim on it.

Transfer. The company sends those tokens to the supplier. This is a single update to the shared ledger. No correspondent chain, no nostro debits, no messages between institutions. The money hasn’t moved at all — it’s still sitting in the same reserve account. Only the claim changed hands. If that sounds strange, notice it’s exactly how commercial bank money works: when you pay someone at your own bank, the bank’s assets don’t move either; the liability just gets reassigned. A stablecoin transfer is that same trick, except the “bank” is a shared ledger spanning every participant on it.

Burn. The supplier redeems the tokens with the issuer. The issuer destroys — burns — them, removing them from the ledger, and wires $1,000,000 from reserves to the supplier’s bank account. The money exits the reserve and re-enters the ordinary banking system.

The mint and the burn are the on-ramp and off-ramp — the doors between the ordinary banking world and the shared ledger. Everything between the doors is on-chain (recorded on the shared ledger, settling in seconds, running around the clock). Everything outside them is off-chain (the familiar world of bank wires, business hours, and local payout rails).

Hold onto that boundary. It’s where this technology’s costs, risks, and business models all live, and it’s the single most important structural fact in this series.

The honest one-paragraph case for why you should care

Here it is, without adjectives: a stablecoin transfer settles with finality in seconds, at any hour of any day, between any two parties on the same ledger, without either side prefunding a nostro position — because the “nostro” is the issuer’s reserve, funded once at mint rather than maintained pairwise across every corridor. For treasury settlement across time zones, for weekend and out-of-hours movement, and for corridors where correspondent chains are long and liquidity is thin, that combination is genuinely new. Whether it’s cheaper, all-in, once you price both ramps and the FX inside them — that’s an empirical question, corridor by corridor, and it’s exactly the question the rest of this series is equipped to answer.

What this looks like in the real world

If the mint → transfer → burn loop still feels abstract, here is the same machinery wearing a familiar brand. MoneyGram — a name you can find in any remittance pricing comparison — operates cash ramps for USDC: a person can walk into a participating MoneyGram location, hand over physical cash, and receive stablecoins in a digital wallet; at the other end of the corridor, the recipient can walk into another location and convert tokens back into local cash, with the service reaching over 170 countries.

Map that onto the vocabulary you just learned and nothing new is happening: the cash desk on the sending side is the on-ramp, the hop between wallets is the on-chain transfer, and the cash desk on the receiving side is the off-ramp. The remarkable part is what sits in the middle. In MoneyGram’s classic model, the middle is prefunded agent liquidity and settlement across correspondent accounts. Here, the middle is a single shared-ledger update — while the edges of the business, the physical locations where people actually touch money, stay exactly where they were. Incumbents’ agent networks turn out to be one of the scarcest assets in this new architecture: everyone can access the shared ledger, but very few can turn a token into cash on a street corner in 170 countries.

That last observation — what happens when an incumbent decides its network is the ramp, and goes as far as issuing its own token — is a story with real strategic stakes, and it gets a full treatment in Episode 3.

What we didn’t tell you, on purpose

You now know what a blockchain is (a shared settlement ledger with no central operator), what a stablecoin is (e-money whose claim lives on that ledger as a bearer instrument), and where the money is at every step (in the issuer’s reserves, until someone burns their way out).

You do not yet know how big any of this actually is — and if you’ve seen headline numbers, be suspicious: the gap between raw on-chain volume and genuine payment volume is enormous, and quoting the former as the latter is the single most common analytical error in this space. Measuring it honestly is Episode 2.

Terms to internalize

Every term below was introduced in this episode. If each one-liner reads as obvious, you’ve absorbed everything Episode 2 assumes.