Europe's money runs on someone else's rails.
Olbra Chain is a layer-1 for regulated European money. Gas is paid in the money itself. Validators are named European institutions under contract. Every transfer is private to the public and open to a supervisor.
Nine cells in ten. The share of every block this chain reserves for payments, at the protocol, where nobody can bid it away.
- Chain
- Olbra Chain · EVM
- Consensus
- BFT · sub-second finality
- Gas
- PLNY · EURY · USDY
- Validators
- Named EU institutions
- Privacy
- Confidential · auditor key
- Ethereum
- Issuer mint and burn
- Status
- In design · devnet targeted Q4 2026
Olbra ApS is an e-money institution. Finanstilsynet, the Danish financial regulator, authorised us on 11 February 2025, reference 17547. That authorisation is passported across the European Economic Area.
Seventeen licences. Almost no money.
Europe spent two years building the licence regime and then watched the money go somewhere else.
ESMA's interim register of 24 April 2026 lists 17 authorised e-money token issuers across ten member states. Olbra ApS is one of them. Between them they have issued about €680 million of euro tokens — a quarter of one percent of a $311 billion stablecoin market.
The demand figures are worse than the supply figures. EURC does real volume on 89,714 monthly active addresses, against USDC's 15.66 million. The whole euro complex is backed by roughly $18 million of on-chain liquidity, and the deepest euro pool in decentralised finance is EURC against EURCV — euro tokens trading against each other. Coinbase has delisted the EURC pairs.
Deutsche Bank has a BaFin licence, Galaxy and Flow Traders as partners, ten chains and a Deutsche Börse listing. EURAU stands at €380,000, fourteen months after launch. Qivalis has 37 banks in 15 countries behind it and has issued nothing at all. The seventeen G-SIBs who organised on 1 September 2026 are launching in dollars.
Every one of those issuers put its token on a rail it does not own. That is the part nobody has changed.
- 17
- Authorised EMT issuers in the EU. We are one of them
- 0.25%
- Of the stablecoin market denominated in euro
- €380k
- Deutsche Bank's euro token, fourteen months in
Sources: ESMA interim EMT register, 24 April 2026. DefiLlama stablecoin supply, 17 September 2026, corrected for the EUR/USD conversion the endpoint applies. Artemis monthly active addresses. AllUnity's own supply page.
We rent four rails.
Four owners, four fee markets, four roadmaps. Our money is a guest on all of them.
- Ethereum mainnet
- PLNY, EURY, USDY — deployed 23 Jul 2026, CREATE3 v2.0.0, verified. EURY 5,000 in issue; PLNY and USDY supply 0, issuance commencing.
- Base 8453
- Addresses reserved via CREATE3, verified unoccupied. Nothing deployed. An integration pointed at chain 8453 finds no contracts.
- Canton
- Daml packages written and tested. Not deployed, not on TestNet.
- Robinhood Chain 4663
- Tokenised equities inside the app.
Ethereum prices our transfers in ether. Base is Coinbase's and shares a fee market with everything else on it. Canton has no EVM on MainNet and a developer ecosystem of a hundred repositories. Robinhood Chain belongs to a broker. Every one of them can change its fees, its upgrade path or its business without asking, and Circle proved in September 2026 that an issuer can delete a chain's reason to exist with a press release — USDC and CCTP are being withdrawn from Noble entirely.
Olbra Chain.
A sovereign layer-1, EVM throughout, built for money that has an issuer, a licence and a regulator.
Everything a developer knows still works. Solidity, Foundry, Hardhat, viem, the same addresses, the same signatures, the same tools. What changes is underneath: fees are quoted and settled in PLNY, EURY and USDY, a payment lane is reserved at the protocol so transfers never queue behind speculation, validators are named European institutions under contract, and a confidential transfer carries an auditor key that a supervisor holds and a sender cannot route around.
And because we issue the money, there is no bridge. We mint on Ethereum and we mint here, against one supply ledger and one redemption obligation.
- <1s
- Deterministic finality. No reorganisations, no confirmation counting
- 90%
- Of blockspace reserved for payments at the protocol
- 0
- Gas tokens a user has to hold, own or understand
Targets, not measurements. Nothing on this page is built yet. See the roadmap for what would be true when.
Five layers and one supply ledger.
The diagram scrolls sideways on a narrow screen.
A payment never queues behind a memecoin.
Blockspace is the scarce good. A payment chain that auctions it to the highest bidder is a payment chain until the first mania.
Circle launched Arc on 16 September 2026 with BlackRock, DTCC, Visa, Mastercard, ICE and Standard Chartered as founding validators. It is the most institutional validator set ever assembled. Day one was 7.83 million transactions, of which about 624,000 were USDC transfers. Traders turned it into a memecoin casino inside twenty-four hours, and every one of those blocks was blockspace a payment did not get.
Tempo answered it in the protocol. Its block gas limit is 500 million, of which general computation is capped at 30 million — roughly 94% of every block is reserved for stablecoin transfers and cannot be bid away. We measured it live on 18 September 2026: chain 4217, 0.56 second blocks, the 500 million limit exactly as documented.
Olbra Chain reserves the lane. A transfer of regulated money settles in its own space at a price that does not move when something unrelated becomes popular.
What the lane guarantees
A fixed share of every block belongs to EMT transfers. Congestion elsewhere on the chain cannot raise the price of sending money or delay it past the next block.
Why it matters more than cheapness
An instant euro credit transfer has ten seconds in law. A fee that is usually a tenth of a cent and occasionally two euro is unusable for a payments business. Predictability is the product.
Pay the fee in the money you are sending.
A user who holds euro should be able to send euro. Every chain in production makes them buy a second asset first.
Across 53 non-custodial neobanks we surveyed, nobody makes a user hold a gas token, and the operators who took the problem most seriously moved it into the chain rather than papering over it in the app.
The mechanism with production evidence is Celo's. One extra transaction field names the currency; a governance-controlled directory holds each token with its oracle and its intrinsic gas; an adapter handles six-decimal tokens; the client reads the rate once before the block and fails closed if the oracle is unavailable; debit and credit hooks reserve and refund around execution. Close to half of all Celo gas is now paid in stablecoins, across 21 registered fee currencies.
Its decisive property is what it does not require. It works with a plain externally-owned account — no smart-contract wallet, no bundler, no relayer, no account-abstraction stack. The first transaction a new user ever sends costs them nothing to prepare.
Quoted in the currency
A fee is priced in PLNY, EURY or USDY at a rate fixed for the block. What the wallet shows is what the user pays.
Sponsored when it should be
An account with a zero balance is the real gap. Onboarding, claim links and first receipts are sponsored at the protocol, so a funded signup works on arrival.
Honest about the cost
Paying in a stablecoin costs more gas than paying in a native token — roughly nine times on Celo after its 2026 recalibration. Sub-cent either way, and budgeted rather than discovered.
Fee abstraction of this kind lives inside the execution client, so we own a client fork for as long as the chain runs. Celo's migration from op-geth to op-reth took several quarters and a full reimplementation in Rust. That is a standing cost, and it is the reason nobody has built this for the euro.
Private to the public. Open to a supervisor.
Visa's own read: many banks treat the lack of privacy as a dealbreaker for moving anything meaningful on-chain.
No treasurer will publish payroll, supplier terms and counterparty balances to a block explorer, and no bank will let a client do it. Canton was built for that objection and the institutions went there: Broadridge reported approximately $7.5 trillion in monthly tokenised transactions in its FY2026 annual report, and DTCC ran its first live production transactions on 15 July 2026 with JP Morgan, Société Générale, Citadel Securities and BNP Paribas.
Canton buys that privacy with a closed world. There is no EVM on its MainNet, the developer ecosystem is about a hundred repositories of which nearly all belong to one vendor, and by its own documentation settlement across synchronizers is not atomic — a contract is unassigned on one domain, unusable in between, and assigned on the other. The sequencer still sees who is transacting with whom.
Olbra Chain takes the privacy property onto an EVM chain, with the auditor built into the proof rather than bolted beside it.
The auditor key, enforced
Avalanche's Encrypted ERC has been on mainnet since March 2025 with two completed audits. Balances are twisted ElGamal ciphertexts updated homomorphically; every private transfer is a Groth16 proof constrained in the circuit to carry an encryption of the exact amount to the auditor's public key. With no auditor set, a private transfer reverts. There is no version of the transaction that omits the supervisor.
Revocable, scoped disclosure
Zama's protocol has been on Ethereum mainnet since 30 December 2025 with $72 million of shielded value and sixteen confidential Morpho vaults curated by Steakhouse, Wintermute and Flowdesk. Its delegation is a triple of delegator, delegate and contract, with an expiry and a revoke — which is what an audit engagement actually looks like.
No design on the market gives per-account, revocable, retroactive, threshold-governed auditor access that also hides the counterparty graph. Every option trades at least two of those away. Amounts are the part that is solved; the graph is the part that is open. Circle has conceded the direction commercially — confidential USDC is shipping on Canton, Aleo and Miden — and Arc lists auditor viewing keys as a design and as a notable omission on the same launch page. Arriving there first is a real position. Claiming it works today is not one.
The regulatory case nobody has made.
The received wisdom is that a bank cannot touch a public chain. We read the instruments. The received wisdom is looser than the text.
- Operational resilience is the argument, and almost nobody uses it. DORA has applied since 17 January 2025. Article 3(19) defines an ICT third-party service provider as an undertaking. Article 30 requires one written contract with named processing locations, termination rights and — at 30(3)(e)(i) — unrestricted rights of access, inspection and audit that no other agreement may limit. All nineteen designated critical providers are incorporated companies with signable contracts: AWS, Microsoft, Google, IBM, Oracle, Bloomberg, Equinix. A permissionless protocol satisfies none of it and there is nobody to breach it, so the dependency falls back onto the financial entity's own framework under Articles 5 to 16. A chain whose validators are named, contracted European entities turns an unsignable dependency into a compliant one.
- The capital treatment in Europe is not what people repeat. The EU applies CRR3 Article 501d, in force since 9 July 2024, and its transitional regime is textually technology-neutral — no permissionless test, no traceability test, no validator test, no infrastructure add-on. Basel's SCO60 contains no ban on permissionless chains either; that reading comes from a 2023 consultation note. SCO60.19 expressly allows validators to be subject to appropriate risk-management standards as an alternative to being regulated, and SCO60.53 sets the infrastructure add-on at zero. In March 2026 the Federal Reserve, OCC and FDIC said plainly that the capital rule draws no distinction between permissioned and permissionless blockchains.
- There is a live defect worth raising. Article 501d(2)(b) gives a 250% risk weight to asset-referenced tokens whose issuers comply with MiCA. E-money tokens are not named there at all, so on the plain text a fully compliant EMT falls to paragraph (c) at 1250%. The Commission's proposal under 501d(1) was due on 30 June 2025 and has not been made. That is an advocacy position for a licensed European issuer with a chain to point at.
- Finality is a designation, not a confirmation count. The Settlement Finality Directive defines a system as a formal arrangement, governed by a chosen member-state law, designated and notified to ESMA. Every DLT settlement system authorised under the Pilot Regime carries an exemption from the CSDR finality article. The Eurosystem's own answer is the Pontes design: DLT as the coordination layer, with finality achieved inside T2. Olbra Chain is designed to be designable — a closed, identified participant set, a chosen governing law, a published moment of irrevocability.
- The controls a supervisor asks for are in the protocol. Travel-rule data under Regulation 2023/1113 with no de minimis, and proof of control on a self-hosted address above one thousand euro. Sanctions freezing as an obligation of result under Article 2 of Regulation 269/2014, which is the reason every credible token carries a blacklist. Segregation on the ledger under MiCA Article 75(7). Named operators for every function, each contracted, each auditable.
Basel reopened the standard on 19 November 2025 and no revised text has appeared, so the floor moves. Singapore's consultation would reward validator dispersion rather than validator identity, which cuts against a permissioned set. And MiCA itself imposes no obligation to control the chain you issue on — Article 34(5)(f) reaches only a proprietary ledger. Owning the rail is a DORA and a commercial argument before it is a MiCA one.
Why a layer-1 and not a rollup.
The rollup route is cheaper to start and it fails on four specific things.
The gas token is constrained
OP Stack's custom gas token permits eighteen decimals only, which excludes our six-decimal EMTs outright. Arbitrum Orbit allows a custom fee token on AnyTrust data availability alone, set immutably at genesis. The one feature this chain exists for is the one a rollup makes hardest.
Blockspace belongs to someone else
A rollup shares a fee market and a sequencer policy it does not set. Reserving a payment lane is a protocol decision, and on a rollup it is somebody else's protocol.
Exit latency is measured in hours
Circle credits USDC after 65 Ethereum blocks on ZKsync Era and treats Linea as six to thirty-two hours. A rail whose worst case is worse than T+1 is not a payment rail.
There is nobody to contract with
A shared sequencer, a shared upgrade admin and a security council are exactly the arrangement DORA Article 30 cannot be satisfied against. Sovereignty is what makes every operator nameable.
We bootstrap our own security rather than inheriting Ethereum's, and we own an execution-client fork for the life of the chain. The precedent runs our way — Circle built Arc, Stripe and Paradigm built Tempo, Tether backs Plasma and Stable, and all four are sovereign — but precedent is not proof. This is the single decision on the page that is hard to reverse.
The issuer is the bridge.
Bridges were 1.4% of crypto losses in 2025. They are 40.6% of 2026 to date, across 42 incidents — more than the previous five years combined.
The two large failures of 2026 involved no compromised key. Kelp lost $293 million on 18 April when compromised RPC nodes inside LayerZero's own infrastructure fed a forged packet to a single verifier. Liquid lost $320 million on 6 September to a range-proof cache-key collision, with federation keys untouched. The 2022 remedy — more signers, better custody — addresses neither.
Circle's answer is now 32.4% of all cross-chain volume, and it is not a bridge at all: CCTP burns USDC on one chain and mints it on another against Circle's own attestation. Institutions decided a known issuer's operational risk beats a bridge's cryptoeconomic risk. Tether reached the same place by a different route with USDT0.
We are the issuer. Our redemption obligation under MiCA Article 49(4) runs against us on every chain the token exists on. Moving PLNY between Ethereum and Olbra Chain is a burn and a mint on one supply ledger, with no third party holding collateral and no message to forge.
One supply, audited once
Total issuance is the sum across chains, reconciled against one reserve and one daily attestation. A partner integrating either chain integrates the same token.
No wrapped asset
There is no wrapped PLNY with a different risk profile, no lockbox to drain, and no depeg that originates in the bridge.
Ethereum stays canonical
Liquidity, listings and DeFi integrations stay where they are. Olbra Chain adds a settlement venue and takes nothing away from mainnet.
Everyone is building this. Nobody is building it in euro.
Measured on 18 September 2026, from chain registries, public RPC endpoints and DefiLlama.
| Olbra Chain | Arc | Tempo | Plasma | Celo | Canton | |
|---|---|---|---|---|---|---|
| Gas paid in | PLNY, EURY, USDY | USDC only | Any TIP-20 dollar token | XPL. Paymaster still in development | CELO plus 21 registered tokens | Canton Coin |
| Who holds the licence | Olbra ApS, Danish EMI, EEA passport | Circle, US and EU | None. Stripe is the sponsor | VASP acquired; MiCA CASP and EMI pursued, not granted | None. Mento is the issuer | None. Each participant holds its own |
| Confidential transfers | Designed in from layer 3 | Announced. Listed as a notable omission at launch | Zones give an operator full visibility | Module under active development | None | Sub-transaction privacy, in production |
| Supervisor key | Enforced in-circuit | Viewing keys, stated design | Operator visibility, not a key | None published | None | Participant-scoped, sequencer sees the graph |
| Validators | Named EU institutions, contracted, staking $OLBRA | 11 to 12 founding, proof of authority | Permissioned. Visa, Stripe, Zodia. No token | Proof of stake still in development | Centralised sequencer, stage 0 | ~11 on-chain operators behind 59 listed seats |
| Finality | Sub-second, deterministic | ~0.5s blocks, sub-second | 0.56s measured | 1.00s blocks, ~7 tx/s measured | 1.00s blocks, L2 proof latency on top | Deterministic within one synchronizer |
| Link to Ethereum | Issuer mint and burn. No bridge | CCTP | Bridged | Bridged | OP Stack optimium, EigenDA | No EVM on MainNet |
| Stablecoin float | EURY 5,000 in issue today | $639m | $167m | $1.24b, TVL down 91% from peak | $113m, down 46% in two years | Reported in trillions of monthly value, including private subnets |
| Denominated in euro | Yes, and in złoty | EURC listed among 32 tokens | Dollar-denominated by design | No | EURm exists, negligible | No euro EMT |
Arc launched 16 September 2026. Tempo figures measured at rpc.tempo.xyz and its own public benchmark. Plasma throughput measured at rpc.plasma.to. Celo measured at forno.celo.org. Canton volumes as reported by the Foundation, including private subnets; Global Synchronizer throughput in Q1 2026 averaged about 10 transactions per second.
Where does the first hundred million come from?
A chain with no money on it is a press release. Deutsche Bank proved that a licence, a balance sheet and ten chains produce €380,000.
Olbra Chain does not launch into an empty market and hope. It launches under a business that already has the flows, and every one of these is on the existing roadmap with an owner and a date.
Our own ladder
The savings shelf, collateralised lending on Morpho, the index and the card all settle in EMTs. They are the first tenants and they do not have to be recruited.
Qpexa
The rebranded Mobilum app: roughly 400,000 pre-verified Polish and central European users, rolled into a channel that already exists.
WhiteBIT
An agreed four-pillar launch — global listing, Polish marketing, merchant acceptance through WhiteBIT Pay, and the Nova card.
Kanga
Deposit custody under MiCA Article 60(4), with more than 800 physical points of presence for cash in and out.
Business accounts
Treasury, payroll and payables for verified companies — the layer that brings its own users, and the one that most needs confidential transfers.
The Romanian corridor
4.6 million people abroad sending more than €6.5 billion home each year at five to seven percent. There is no authorised leu e-money token in existence.
Four EU currencies — forint, leu, krona and krone — have no native e-money token at all, and the koruna has one closed-network token. Under our licence each new currency is a whitepaper and a notification. On our own chain each one is also a fee currency, a settlement pair and a corridor on day one.
Four phases to a rail of our own.
Every date below is a target. Nothing here is built, and the first phase is a decision rather than a deployment.
Devnet and the case.
A running chain and a specification complete enough to commit against. No public network, no partner conversations, no token.
- Client fork with fee abstraction, a single sequencer, PLNY and EURY as fee currencies
- Confidential transfer prototype with the auditor key enforced in the circuit
- Issuer mint and burn against the Ethereum contracts on a testnet
- Legal opinion on settlement finality and a designation route under Danish law
- Economic design for $OLBRA settled, and put to counsel before it is published
Permissioned testnet with design partners.
Three to five institutions running validators against real integration work, under contract and under DORA terms. This is where the architecture meets a risk committee.
- Named validator set, each with an Article 30 contract and audit rights
- Confidential transfers in the hands of a supervisor and an external auditor
- Payment lane, sponsorship and the fee-currency directory under load
- Public specification, explorer, faucet and the first external deployments
- Third-party security audit of the client fork and the proving system
Mainnet, with money on it from the first block.
The chain opens under flows that already exist. PLNY, EURY and USDY are native at launch, and the group's own products settle here.
- Genesis with contracted European validators staking $OLBRA
- Ethereum supply ledger unified: mint here, burn there, one attestation
- Savings shelf, lending and business accounts settling on-chain
- Qpexa, WhiteBIT and Kanga wired to the new chain alongside mainnet
- Grant programme open, paid in EMTs on milestones
Currency expansion and progressive decentralisation.
The validator set widens beyond the founding institutions and the fee-currency list widens beyond three currencies.
- Leu, koruna, forint, krona and krone as EMTs, fee currencies and settlement pairs
- Validator admission opened to any institution meeting the published standard
- Payment-versus-payment settlement between EMTs as a supervised service
- Designation as a settlement system under the chosen national law
- Confidential transfers extended from amounts to the counterparty graph