The Economic Model

How the money math works: goal-equals-supply pricing, the three-way revenue split, the lazy yield index and claim formula, the Diamond-Hand multiplier, rollover, compensation, solvency, and the no-retroactive-yield theorem.

This is the rigorous version of how Gally moves money. All of the math below is fixed-point integer arithmetic — there are no floats on-chain — using a scaling constant $SCALE = 10^{9}$ , computed in 128-bit integers, multiplying before dividing, and always flooring in the protocol's favor.

Pricing: the goal is the supply

Gally never needs a price oracle. A project's funding goal, measured in USDC, is also its exact total deed supply:

totalshares = fundinggoal, 1 share = 1 USDCofprincipal

Because pricing is fixed at one-to-one, there is nothing to quote, manipulate, or feed from outside. This single decision removes an entire class of oracle risk.

The three-way revenue split

When revenue arrives at an operational project, the contract splits every deposit atomically — the entity never has to be relied on to "send the investors' share." For gross revenue $G$ , a protocol fee rate $f$ , and an investor split $s$ (both in basis points):

fee = \frac{G \times f}{10000}, investor = \frac{(G - fee) \times s}{10000}, entity = G - fee - investor

The fee goes to the protocol treasury, the investor portion enters the yield index, and the remainder returns to the entity — in one transaction.

The lazy yield index

Rather than pushing payments to every holder (which would cost gas proportional to the number of holders), Gally moves a single global index and lets holders pull. When an investor portion $P$ is deposited and the current unwrapped supply is $u$ :

Δ index = \frac{P \times SCALE}{u}, index + = Δ index

Every deed stores the index value at its last claim (its personal index). What a holder is owed is the gap between the global index and their personal index, times their share count:

payout = \frac{(globalindex - personalindex) \times sharecount}{SCALE}

On claim, the deed's personal index is set equal to the global index, so the same yield can never be claimed twice. Both operations are O(1) — independent of how many holders exist.

The Diamond-Hand multiplier

The denominator $u$ above is unwrapped supply only. Wrapped coins are excluded. So as holders wrap their deeds for liquidity, $u$ shrinks and each remaining unwrapped deed earns a larger share of every deposit:

yieldperunwrappeddeed \propto \frac{1}{u} ↑ aswrappedsupply ↑

In the extreme, if everyone but you wraps, you receive essentially the entire investor portion of each deposit. This amplification is emergent — there is no multiplier parameter to set or game. It is simply the reward for staying unwrapped, funded by the yield that wrapped holders forfeit. For what the wrapped coin is for — DEX trading, lending collateral, and why it's a trustworthy claim — see Wrapping, Liquidity & Collateral.

Rollover: revenue while nobody is unwrapped

If revenue arrives when unwrapped supply is zero (everyone has wrapped), dividing by $u$ is undefined. Instead the investor portion parks in a rollover reserve. The moment anyone unwraps, the reserve is swept into the index and distributed — and the first unwrapper, now counted in $u$ , shares in it. No revenue is ever lost to the 100%-wrapped case.

Compensation distribution

Slashed and seized funds (from disputes or defaults) are distributed through the same index machinery as ordinary revenue, after a grace window. Because the index only reaches unwrapped holders, the protocol first freezes wrapping and opens a grace window so everyone can unwrap, then sweeps the compensation pool through the index. The scaling factor is mandatory here too — without it, a restitution smaller than the unwrapped supply would floor to zero and strand.

Solvency: the pool always covers what it owes

Define what the protocol owes as the sum, over all unwrapped deeds, of each deed's claimable payout. The core solvency invariant is:

rewardpool \geq \sum_{unwrappeddeeds} owed

The argument is short: every deposit adds $P$ to the reward pool but increases total owed by $Δ index \times u / SCALE \leq P$ (because $Δ index$ floors). Every claim decreases both by the same payout. Wrapping force-claims first (zeroing that deed's term); unwrapping adds a term worth exactly zero. So the pool is always at least what is owed, and the tiny gap is exactly the accumulated rounding dust — which stays in the pool as a safety buffer and is only reclaimable once the project has fully closed.

No retroactive yield for wrapped time

When you unwrap, the new deed's personal index is set to the current global index. So for any wrap-then-unwrap round trip between times $t_{1}$ and $t_{2}$ :

claimableover [t_{1}, t_{2}] = (index (t_{2}) - index (t_{2})) \times count = 0

You earn exactly zero yield for the time your deed spent wrapped — by construction. This holds regardless of transaction ordering, block boundaries, or flash-loan tricks. A short cooldown between wrapping and unwrapping is layered on top as extra defense against oscillating supply around a known large deposit, but the zero-retroactive-yield guarantee does not depend on it.

A stated capacity limit

All index and payout math is done in 128-bit integers with an explicit overflow check on each payout, so amounts can never silently wrap around. There is one honestly-disclosed bound: an asset whose lifetime revenue per share grows astronomically large could in theory exceed the safe arithmetic range. For any realistic project this is never approached, and it is documented as a capacity limit rather than hidden.