Why
The correction is worth recording before the content, because the shape of the error is one this catalogue already has a rule for. Agorá started in Korea is a claim with no source, and it appears to be a confusion with Korea's separate domestic CBDC and tokenised-deposit work, which is real and substantial. fork-date-provenance sorts claims into confirmed, deferred, and circulating without a source. This one was in the third bucket and was being repeated as if it were in the first — and the tell was available without any expertise: a project with seven central banks in it has an owner, and the owner is named in the announcement.
What is actually true is less dramatic and more useful. The unified-ledger concept came from the BIS, the BIS Innovation Hub launched Agorá as an international project in April 2024, and seven central banks were in it from the start. Korea was not joining an American project and America was not joining a Korean one. Both were founding participants, which is a different and better fact than either misreading.
And inside that, one detail is load-bearing: Korea is one of only two non-reserve-currency jurisdictions in the group, alongside Mexico. Of the seven, the dollar, euro, yen, sterling and Swiss franc are all named reserve currencies in the IMF's own classification. The won and the peso are not. That makes Korea's seat structurally different from five of the other six — and the Bank of Korea's governor said as much from the outset, describing participation as a chance for a non-reserve-currency country to help set standards for cross-border payment and settlement.
Which is the part worth taking seriously, because standard-setting is exactly where a non-reserve currency's leverage lives. A unified ledger has to decide how a payment leg denominated in a currency nobody holds as reserves actually settles. There are only two shapes: the currency is a first-class leg on the ledger, or it is routed through a major currency and the non-reserve leg is an edge conversion. The first makes the won a settlement asset; the second makes it a local last mile. Nothing about the technology decides that — the rules do, and rules get written by whoever is in the room while they are being written. cross-border-rail-interop reached the same conclusion from a different direction: governance, not throughput, decides which rail wins.
And the project deserves to be read as the institutional answer to the same problem stablecoins are the outside answer to. Both target the friction in cross-border payments; they differ in who is allowed to issue and who bears the settlement risk. what-needs-a-stablecoin asks why this currency, in this transaction and finds that corporate cross-border has the most friction and the weakest answer for a won-denominated token. Agorá is where that same question gets answered by central banks instead of by product design — and if it answers it the first way, part of the case for a won stablecoin in cross-border settlement gets absorbed rather than defeated.
The honest caution is the base rate. A seven-central-bank project with dozens of private participants produces reports on a horizon measured in years, not a rail. build-rent-or-own-the-rail records what consortium governance does over time, and nothing here is exempt. The useful posture is to read the outputs for the one design decision above and ignore the rest, which is a much smaller commitment than following the project.
How it works
The claim, sorted
| Claim | Status | What settles it |
|---|---|---|
| Agorá started in Korea | No source — likely confused with Korea's separate domestic CBDC and tokenised-deposit work | The announcement names its owner |
| Korea joined an American project | False | Both were founding participants |
| The BIS proposed the unified ledger and the BIS Innovation Hub launched Agorá, April 2024 | Confirmed | BIS publications |
| Seven central banks from the start | Confirmed | The participant list |
| Korea is one of two non-reserve-currency jurisdictions in it | Confirmed, and checkable | The IMF's own reserve-currency classification |
The seven, and why the split matters
| Participant | Currency | Named reserve currency? |
|---|---|---|
| Federal Reserve Bank of New York | USD | Yes |
| Banque de France (Eurosystem) | EUR | Yes |
| Bank of Japan | JPY | Yes |
| Bank of England | GBP | Yes |
| Swiss National Bank | CHF | Yes |
| Bank of Korea | KRW | No |
| Banco de México | MXN | No |
Five of seven are writing rules for currencies that already settle everywhere. Two are writing rules for currencies that do not. That is not a complaint about the composition — it is the reason the two seats are worth more than the headcount suggests, and the reason the Bank of Korea framed participation as standard-setting rather than as membership.
The one design decision to read the outputs for
| If a unified ledger settles a non-reserve leg... | Then the won is | And the consequence |
|---|---|---|
| Directly, as a first-class leg | A settlement asset | Korean institutions hold and move won on the ledger; the FX conversion is a choice, not a requirement |
| By routing through a major currency | A local last mile | Every cross-border won payment carries a dollar leg and its spread, exactly as today with extra plumbing |
Nothing technical decides this. It is a rules question, which is why being in the room while the rules are written is the whole of the opportunity — and why the useful thing to extract from any published output is which of those two rows the worked examples assume.
Two answers to one problem
| Agorá / unified ledger | Stablecoins | |
|---|---|---|
| Who issues | Central banks and regulated commercial banks | Issuers under whatever regime applies |
| What is tokenised | Central bank money and commercial bank deposits, on one programmable platform | A claim on a reserve pool |
| Settlement risk sits with | The existing banking hierarchy | The issuer, and its redemption desk |
| Speed of arrival | Years — it is a standards process | Already here, unevenly |
| Where it is decided | In the room | In the market |
Read alongside what-needs-a-stablecoin, which asks why a won-denominated token in a given transaction and finds corporate cross-border to be the weakest case, and stablecoin-redemption-desk, which is the same settlement-risk question asked of the private answer. If Agorá lands on a first-class non-reserve leg, part of the won-stablecoin case for cross-border settlement is absorbed rather than beaten.
What this card does not claim
The project's current phase, findings and any published output after early 2026 are outside what this card can assert, and should be confirmed against BIS material directly. The participant list and the April 2024 launch are the durable facts; everything about progress is not.