How it starts

Nobody puts money in at the beginning. What day one actually looks like, what the first transaction is, why anybody sends it, and how the protocol grows from an empty state.

Spec v0.9.1, reviewed 2026-09-08

Most protocols open with a treasury behind them: a team deposits into the pools, funds a reserve, provides the first liquidity, and the numbers on the screen are theirs until users arrive. Fyber has no treasury and no company, so it cannot do that, and it does not.

Guarantee

At deployment the fyUSD supply is zero, every Stability Pool is empty, the peg module holds nothing, the reserve is at zero and no fyUSD pool exists. Neither the founders nor any entity puts money into any of it. The deployment script transfers twenty millionths of a token, one per branch, for a self-test contract, and nothing else.

Every number these pages describe is therefore a number users will have created. On day one they are all zero.

Nobody can borrow yet

SPY and AAPL are open from the moment the contracts land: their price sources are alive and their branches are active. But a branch's debt ceiling is the smallest of four terms, and one of them is its own Stability Pool divided by 0.4× on tier 1 or 0.8× on a single stock. An empty pool makes that term zero, so the ceiling is zero, and nobody can borrow a cent.

That is deliberate. The pool is what absorbs a liquidation, and a branch lending against a pool that could not absorb one is exactly the failure the requirement exists to prevent. It applies at every size, including this one.

Rule R-5.4.1

The first transaction

Anyone can send it, and it is one call:

Opening a branch

PSM.swapInToSP(branch, USDG)

USDG in → fyUSD minted at the peg → deposited into that branch's Stability Pool

It converts and deposits in a single transaction, so the sender never holds fyUSD in between and never touches a market. The moment it lands, borrowing opens on that branch, at 1.25× times the pool divided by the tier's requirement, under whatever debt step the branch has reached.

There is no allowlist, no minimum beyond the pool's own rules, and nothing anybody has to approve. The first person to send it is whoever sends it.

Rules R-8.3.3, R-9.1.1

Why the peg module can mint at all

Thirty per cent of nothing is nothing. The module's cap is a share of the fyUSD supply, and on day one there is no supply, so the cap would forbid the first mint and the protocol could never leave zero.

One constant resolves it. The module may always have minted up to 1,000,000 fyUSD, whatever the supply is:

How it is calculated

minted by the module + incoming USDG ≤ max( 1,000,000 fyUSD , 30% × (supply + incoming USDG) )

Past a supply of about 3,333,333 fyUSD the share cap binds again and the franchise stops mattering.

Risk

While the protocol is small, that means up to 1,000,000 fyUSD of fyUSD can be backed entirely by USDG rather than by tokenized stock, and therefore by the balance sheet of one issuer. It is bounded in absolute terms, it shrinks as a share of the supply as the protocol grows, and the depeg guard still refuses to mint while USDG prints under 0.985. It is a real concentration for the first months, and it is the price of starting from zero without anybody's money.

Rules R-8.1.1, R-8.6.2

Why anybody goes first

Two reasons, one of which is unusual and finite.

Season zero. The FBR counters run from the deployment transaction, not from the first season. 2,000,000 FBR is set aside for everything done before the first season opens, 35% of it by fyUSD-days deposited and the rest by interest paid — the season zero split is 55 and 45 rather than the four-class split of later seasons. Whoever is early is alone on that counter, and the counter does not care how large the protocol is. It cares how many fyUSD-days there were and how few people accumulated them.

The interest. Once borrowing opens, depositors receive 90% of what borrowers pay, on a pool that is small relative to the debt it backs. A pool holding 40% of a branch's debt at an average rate of 4% is receiving about 8.5% a year. That is the mechanism described on Earn, and the launch range there is high for exactly this reason: the pool is small, not the protocol rich.

Neither of those is a promise. The first is a counter that may be worth nothing, and the second moves with what borrowers choose to pay.

Rules R-19.8.2, R-19.8.3, R-10.2

The shape of the first six months

Each step is a debt ceiling opening on a date, and each one needs a pool behind it. The figures below are what the constants imply, not a plan anybody is executing.

When	What is possible	What it needs
Day 1	$100k of USDG becomes $100k in the SPY pool, and a $100k borrowing ceiling at the first step	Ten borrowers, since the cap is $50,000 per address for the first 90 days, or a tenth of the ceiling, whichever is smaller
Day 14	The step rises to $1M	$400k in the pool, because the ceiling is also the pool divided by 0.4×
Day 90	The step rises to $2M	$800k in the pool, and no bad debt ever recorded
Day 180, the first FBR season	Something like $5M of debt against $4M of pool	Every dollar of it deposited or borrowed by somebody

Nothing in that table is guaranteed and nothing in it is scheduled. The dates are the dates the ceilings open; whether anything sits under them is a matter for the people who show up.

Rules R-5.4.3, R-5.4.4

The reserve fills last

The fyUSD reserve that absorbs bad debt before anything else also starts at zero, and it has no other source than revenue: five points of the routed flow while it is under target, the FBR share once the first season opens, and half the vault fees.

At the pace the revenue model implies, it reaches its target of 2% of the debt somewhere between the end of the second year and the end of the third. Two things follow, and both are stated here rather than discovered later.

Risk

For the first two years or more, the first loss layer is thin. Bad debt that appears before the reserve is full falls through to the staked FBR and then to the branch's own borrowers, and the layer between them is smaller than the target suggests.

Because the reserve is filled before anything else, no FBR is bought back at all until it is full. The buy-back that page FBR and sFBR describes is close to nothing in the first year, and does not overtake what the seasons distribute before the fifth.

Rules R-10.8, R-19.1.2

What none of this is

There is no launch programme, no allocation for early users beyond the season counters every user earns, no private round, no partner deposit and no address that receives anything for being first. The protocol has no way to give any of those, because it has no discretionary flow and nobody to decide one.

What it has instead is an empty set of contracts, a door anybody can walk through, and a counter that started at deployment.