$OLBRA finally has a job.
Gas is the money. Security is the stake. The token is the tier — the thing a user holds on purpose, because holding it makes everything else cheaper.
Nobody ever needs $OLBRA to transact.
Security and money are two different assets. Conflating them is what broke every chain in the field: a user who must buy the security token to send a euro is a user who churns at the first step.
What $OLBRA does is price everything else. Stake it and the fee falls, the exchange spread narrows, the card pays more, the savings shelf allocates first and the lending limit rises. It is the account tier, held openly, visible in the app, with a published table of what each level buys. Validators stake the same asset and are slashed against it, so the token that discounts a user's fee is the token that secures their transfer.
The tier key
One staked balance sets the fee, the exchange spread, the cashback rate, the lending limit and the priority on a new savings Series. The user sees a tier, and the tier is the token.
Paid back in euro
Staking rewards settle in EURY, from sequencing and gas-spread revenue that was earned. A token that pays a measurable cash yield can be valued. A token that pays in itself can only be sold.
Bought, never minted
Every reward, rebate and cashback is funded from reserve income and interchange, and the token is bought on the open market to pay it. Supply is fixed, and it is never minted against.
Validator collateral
A seat requires a stake with a long unbonding period. Institutional demand is a supply sink that does not sell on a drawdown, which is what retail float cannot give a token.
Partner bonds
Resellers minting through the partner API post a bond in $OLBRA against their mint line. It scales with distribution and it locks for as long as the relationship runs.
A treasury that can buy
Reserve income on assets we hold is real cash, audited and ours. The Foundation may apply treasury resources to acquire $OLBRA from a published address anyone can check. No other chain's token has a reserve behind it — and no rate, schedule or entitlement is promised against it.
What a stake actually buys.
Six levels, one staked balance, one published table, and a seven-day queue to unstake. The exchange fee is the anchor: the app charges 0.35% of the amount exchanged, stated on the quote before you confirm. A stake lowers that number permanently.
| Level | Staked $OLBRA | Exchange fee | Card cashback | Borrowing limit | Savings shelf |
|---|---|---|---|---|---|
| Base | 0 | 0.35% | 0.5% | Standard | Open subscription |
| One | 1,000 | 0.30% | 1.0% | Standard | Open subscription |
| Two | 10,000 | 0.25% | 1.5% | +5 points of loan-to-value | 48 hours early |
| Three | 50,000 | 0.18% | 2.0% | +8 points of loan-to-value | One week early |
| Four | 250,000 | 0.10% | 3.0% | +10 points of loan-to-value | Guaranteed allocation |
| Five | 1,000,000 | 0.05% | 4.0% | +12 points of loan-to-value | Guaranteed allocation |
Illustrative, and the shape of the argument rather than a priced product. Staking rewards sit outside the ladder: they are paid in EURY, pro-rata to the stake, from a fixed share of sequencing and gas-spread revenue, so a larger stake earns more without needing a tier to say so. Cashback is quoted against the card, which does not exist yet — the rate is a commitment at card launch, and the page says so rather than implying a live benefit.
Never mint to pay a reward. Plasma paid for adoption in XPL, XPL is down 94.8% from its high, and a cliff of roughly 1.8 billion tokens — about 18% of supply — releases on 25 September 2026. Cashback funded by issuance is a spend-down wearing a rewards programme's clothes. Reserve income is the only thing that has ever paid for free payments: Circle's FY2025 was $2,747 million of revenue and a $70 million net loss, and it is the largest stablecoin business on earth.
Gas is the smallest line on the page.
A payment chain that tries to earn from fees ends up charging like Tron. The chain earns by keeping balances, and by selling the things a regulated rail can sell that an open one cannot.
Start with the size of the prize. Tron collects more fee revenue than any stablecoin chain in existence — roughly $354 million a year — and it does it by charging about a thousand times what its competitors charge. At a tenth of a cent a transfer, a billion transfers a year is a million euro. Gas is not a business.
Float is. Every euro that settles and rests on our chain is a euro whose reserve we already hold, earning the reserve rate. At three percent, a hundred million euro of retained balances is three million euro a year — more than a billion transfers of gas, from one number that the chain exists to move. Read every line below against that.
- Float retention
- The reason to build it. Balances that live here stay in our reserve rather than leaving for a venue we do not own. Reserve income at 2.0 to 4.0 percent, on assets under our own licence.
- Validator seats
- An annual seat fee plus a staked bond, sold to institutions that want the operational position and the commercial relationship. Canton proved institutions pay for a seat on a rail they need.
- Exchange between our own currencies
- Payment-versus-payment settlement between PLNY, EURY and USDY at a published spread. Basis points on volume, and the volume is already ours.
- The confidential tier
- Private transfers with supervised reporting, sold to verified businesses as a subscription with a per-transaction component. Priced against the treasury team's alternative, which is a bank.
- Reserved capacity
- A guaranteed share of the payment lane, contracted ahead for a year. An exchange or a payroll processor buys throughput the way it buys a database instance.
- Issuance as a service
- Another licensed issuer runs its own token on our chain and pays for the rail, the fee abstraction and the confidential layer. The margin on someone else's licence, with none of its obligations.
- Gas spread and sequencing
- Real, denominated in our own money, and the smallest of the seven. It funds the validators rather than the group.
Every competitor in the table earns from one side of the trade. Circle issues and rents rails. Tempo builds the rail and issues nothing. Celo runs the chain and Mento issues the money. We would be the only operator holding the licence, the reserve, the token, the chain and the distribution at once — so a euro that arrives through Qpexa, settles on our chain, sits in our savings shelf and is spent on our card earns us something at every step, and leaks at none.
Hyperliquid had to negotiate for this. We already own it.
The money on a stablecoin chain is not gas and it is not lending fees. It is reserve yield — and we are the issuer.
Start with what a chain actually earns. Plasma carries $1.2 billion of stablecoins and earned $0.3 million of chain revenue in a year, because stablecoin transfers are free by design. For scale, a full year of chain revenue: Ethereum $65.7m, Base $56.9m, Solana $28.3m, Hyperliquid's own L1 $11.2m. A sovereign chain for regulated euro money will earn approximately nothing in gas, and the honest thing is to plan for that rather than model around it.
So look at what the best-performing token in the category actually does. Hyperliquid's assistance fund converts trading fees to HYPE inside block execution — 99% of fees, running at about $795 million a year, 4% of its market capitalisation. And buried in its own specification is a mechanism called Aligned Quote Assets, under which stablecoin issuers deploying on Hyperliquid share roughly 90% of their reserve yield with the protocol, enforced by a slashable bond of a million HYPE.
Hyperliquid built a protocol rule, a bond and a slashing regime to capture somebody else's reserve yield. We are that somebody. The revenue it had to engineer for, we book.
- €20m
- A year of reserve yield at €1bn of float, at two percent net
- €0.3m
- What Plasma's chain earned on $1.2bn of stablecoins, in a year
- 17×
- Reserve yield against curator fees, at the same balance-sheet size
- Reserve yield
- The flow, and nobody else in the comparison has one. Balances that settle and rest here sit in our own reserve. At two percent net: €500m of float is €10m a year, €1bn is €20m, €3bn is €60m. Circle, Tempo, Plasma and Celo can none of them do this — three hold no reserve and the fourth does not own the chain.
- Exchange between our own currencies
- Payment-versus-payment across PLNY, EURY and USDY at a published spread, on volume that is already ours. It grows with the chain rather than with a market cycle.
- Gas and sequencing
- Real, in our own money, and smaller than the two above by an order of magnitude. It funds the validators. It is also the only one of these that any other chain token has.
- Lending, honestly sized
- Worth naming because the instinct is right and the arithmetic is not. The entire global curator industry earns about $17m a year, and SparkDAO — $579m of assets at a ten percent fee — booked $2,331 in August 2026. No protocol anywhere routes a curator fee into a token buyback. It is a real line and a small one, and it is not what pays for anything here.
The Foundation may apply treasury resources to acquire $OLBRA, from a published address anyone can audit, against audited revenue. No rate, no schedule, no commitment, and no entitlement for any holder — because the moment a buyback becomes a promised share of revenue it stops being a treasury decision and starts being a profit-participation instrument. Aave promised a million a week and has paid nothing for eighty-six days. Ethena's fee switch passed unanimously and pays zero because supply sits below its first threshold. A licensed issuer breaking a public financial commitment is a different order of problem than a protocol doing it.
Four places the float goes and does not come back.
A buyback against a supply nobody holds is a transfer to sellers. These are the reasons to hold, and the reason is never a promised return.
Slashable bonds on every privileged role
The strongest sink found anywhere, and it carries no security character at all — it is collateral, not income. Hyperliquid's perp-dex bond locks roughly $441m across ten deployers; Canton requires five million coins per featured party and twenty-five million from an asset issuer. Validators, issuers, market makers, ramps and vault operators each post one, slashable by validator vote, and slashed stake burns.
The account tier
One staked balance sets the exchange fee, the cashback, the borrowing limit and the savings priority. Six levels, a real discount, and a seven-day unstaking queue — which is the part the evidence says actually retains stake through a drawdown.
Partner bonds
A reseller minting through the partner API posts a bond against its mint line. It scales with distribution and it locks for the life of the relationship, which is the holder profile a drawdown does not shake out.
Nothing is ever force-bought
No user needs $OLBRA to transact. Gas is the money. A token nobody is compelled to hold is a token whose price is information rather than a toll — and it is why the tier reads as a benefit instead of a tax.
The tier keys off the staked $OLBRA balance and nothing else — never off how much EURY you hold, never off how long you have held it, and never off a formula combining the two. A benefit that tracks an e-money balance over time is interest whoever pays it. Keeping the trigger on the stake is what makes the whole ladder a fee discount rather than a yield, and a fee discount is expressly the thing European guidance says a utility token may do.
What the evidence actually says about mechanisms
Twenty tokens with live accrual mechanisms, measured over twenty months. Two went up. dYdX pays one hundred percent of fees to stakers in USDC, on-chain, every block — and is down 92.6%. pump.fun burned a third of its float and fell 27%. Mechanism quality is not the binding constraint.
Float is
FTT died because one holder's book was larger than the circulating value of the token. BNB survived because its burn consumed the entire non-circulating overhang — market cap and fully diluted value are the same number. Publish the whole vesting schedule at launch and keep the float ratio high. Surprise is the damage, not supply.
Plasma paid for adoption in XPL and XPL is down 94.7 percent. Berachain paid apps in emissions until its incentive market reached zero and the mechanism was deleted. Blast paid fourteen percent of supply in Gold and killed the programme after seven months. Every work token in the survey pays between three and forty-four cents of real revenue for every dollar it emits. Every one of them minted to pay a reward. A fixed supply, never minted against, is the property this design rests on. Protecting it is the whole job.