Ponzunomics

A Ponzu project is a closed stack: one token, one sale, one launcher, one distributor. You call the recipe. It clones every contract the project will use and wires them to each other before the transaction ends. Nothing is circulating. Thirty-one percent of supply sits in the launcher, so the pool can exist without being seeded. The rest is sold through an auction or a curve, then claimed once under diamond-hand vesting. Fees and forfeits flow through a distributor weighted by locked LP.

After craft the economics are closed. Vesting, the launcher slice, and the pricing strategy have no setter. What remains before launch is a short list of decisions — when an auction opens, who can buy, where leftover sale allocation goes. At graduation, token distribution no longer depends on further developer actions.

There are nine entities. Everything in the system is a relationship between them.

The mental model: the recipe is the deploy, the bottle is the claim, the distributor is the pot. You cannot change the deploy. You cannot pull the pool. You can leave the claim. Leaving pays whoever stayed.

The flows between them:

• Founder → recipe. One call clones the stack and registers the pool with the hook.
• Recipe → launcher. 31% of supply sits as inventory. Not in the pool. Not in a wallet.
• Buyers → presale. ETH in, a bottle out. The sale discovers a price; the bottle is the claim.
• Launcher → pool. At graduation, initialize and seed are the same transaction.
• Pool, claims, unstakes → distributor. Swap fees, unvested tokens, and early-exit LP land in one pot.
• Distributor → remaining bottles, farm, members. The split is locked LP. Claim tokens and you leave the stream.

You do not deploy a token and then figure out the rest. You call the recipe. It clones the token, the presale, the launcher, the vault, the distributor, the farm, membership, and a pricing strategy, and it wires them before the transaction ends. Either the whole stack exists, or none of it does.

Nothing is circulating at launch.

Supply is minted once. The launcher takes a fixed 31%. The rest is the sale, plus any treasury or team slice you carved out of it. Raising team or treasury reduces the sale, never the launcher. Those tokens sit in contracts. They are not in wallets. New allowances on the project token revert until launch.

• Before launch you can. Schedule when an auction opens. Set who is allowed to buy. Place extra allocations a deferred auction still has room for. Raise the minimum raise — never lower it.
• You cannot. Change vesting, the 31%, or the pricing math. Mint more. Pull the launcher. Grant the treasury to a wallet. Seed or trade the pool before graduation.

At craft, 310,000 tokens go to the launcher. Not to a pool. Not to you. They sit as the inventory that becomes liquidity when the sale graduates. The same transaction registers the pool with the hook. Only that launcher may initialize it. Outside liquidity is rejected.

Until graduation the pool has no seed, and there is nothing circulating to buy, so there is nothing to frontrun. When the sale is ready, the launcher opens the pool at the raise-implied price, deposits every token it holds against the ETH, and locks the LP. There is no withdraw.

The pool can be created. It cannot be frontrun.

The presale is the same machine either way: ETH in, a bottle NFT out, 90% refunds until launch, then diamond-hand vesting. Before launch you cannot dump at a higher price — only refund. The 10% that stays is left for whoever is still in. What you pick is how the price is discovered.

Ponzu Auction

A new token has no price, so the auction does not quote one. A target raise starts high and falls as time runs. Every contribution pulls the finish closer. When the lines cross, everyone settles at the same clearing price.

Being early buys no discount. It buys weight. ETH committed at the open earns up to 5× claim weight, decaying to 1× by the close. Same ETH, earlier, is a larger slice of the allocation when it settles.

Ponzu Curve

The other sale type is a rising price curve. Price starts at $p_0$ and climbs to $p_1$ as tokens sell. Cost for a fill is the area under that line: a trapezoid: the average of the price you enter at and the price you leave at, times tokens bought.

The factory lets you set the run $p_1 / p_0$ from 2× to 100× (2, 5, 10, 20, 50, 100). Unlike exponential curve, which dont have enough area under the curve, and forced to oversell presale, Provided we sell at least 62% of the allocation, Ponzu curves support any multiplier, and the price at graduation will always be higher than the final price. High multipliers have incentive misalignment with vesting, so will cautiously raise multipliers with more price data.

You can claim anytime. You can only claim once. What has linearly vested by that moment is yours. The rest forfeits to whoever is still holding. Claim on day one of a ten-day vest and you keep a tenth. Wait it out and you keep all of yours, plus a share of what early claimers left behind.

Your ETH, any time. Your tokens, once.

Refund 90% before launch. Transfer the bottle after it. Claim early and walk with what you have earned. The one thing you cannot do is dump the unvested remainder onto the pool. At graduation the presale is vesting and the launcher is locked, so the free-floating slice rounds to nothing.

All allocations, including a team vault, airdrips, member seats, come out of the sale, not the 31%. It only exists provided the presale is able to raise enough at a minimum to support a graduation price for the 31% liquidity. Everything is balanced when the token is created: it starts when you craft it. At launch that bottle is sealed.

Team vaults

An auction with allocation to spare can put a slice in a team vault. That slice comes out of the sale, never the 31%, and cannot push the public below the 39% floor. The tokens mint into the vault, not a wallet.

Unlock is not a date. It is a TWAP ladder against launch price: 5×, 10×, 20×, 50×, 100×. Clearing the nth rung entitles the recipient to n/5 of everything the vault has ever received. Nothing moves for the first 30 days, and a rung only clears if the pool actually traded there.

Airdrips

The recipe pre-approves the drop and market contracts to spend the project token. That is the only door. What those contracts receive still clocks off the same launch, and vests on the same terms as the presale. Nobody who received an airdrip can dump before the people who paid can.

Every swap pays a fee. Early claims forfeit tokens. Early farm exits forfeit LP. All of that lands in the distributor. It has no admin. It recomputes the split live from locked LP — veLP: LP that cannot leave, counted with a lock multiplier. Genesis liquidity is permanent, so it weighs as a 10× lock. Farm stake weighs by the lock you chose.

Fees are charged on the input, so the stream is ETH on the way up and tokens on the way down. Holders can pull ETH without selling.

Claiming tokens ends your bottle’s share of the fee stream. What would have kept flowing to you migrates to whoever is still in. The migration is quadratic in claimed weight $C$ against active weight $D$: slow at first, steep at the end, so a few early claims do not empty the presale seat.

The two assets go different ways. ETH that leaves the presale goes to the project (the contributors seat). Tokens that leave go to the farm. The project is not paid in a bag it can dump.

The farm does not mint. It pays from the distributor and from what unstakers left behind. If nobody trades and nobody quits, the farm is quiet.

You pick a lock when you stake. Duration sets your weight: just over 1× at a week, 5× at four years, 10× if permanent. Permanent LP never comes back. Project tokens on a timed card: once. ETH, whenever it accrues. Leave before the lock ends and you get the LP you have earned on the clock; the rest stays with the farm.

Membership is a share pie minted at craft — not tokens, points — sitting in the project until they are granted or sold as cards. Those points map onto the contributors veLP seat: the 31% of genesis LP that is not the presale. Hold the card and you earn that stream for as long as the market trades.

The pie exists from conception. An auction can sit dormant and fill the member table before the public sale. After launch the shares keep earning. They do not vest into a dump. Contributors are never paid the project token through the distributor.

Launch parameters are not a vote. Operator control can be relinquished to futarchy — opt-in at craft. After that, the only decisions left are what to do with assets the project already holds, and whether to rotate the governor. A proposal opens a market. Traders price pass against fail. The TWAP decides.

A proposal nobody trades loses. Funds are escrowed when you propose, so settlement cannot bounce on an empty treasury. The governor never holds the LP; the vault does. You can transfer operator control. You cannot change the recipe.

The stack was designed as one system. Each phase hands pressure to the next: refunds keep the sale honest, an empty pool makes graduation un-frontrunnable, vesting keeps the open from being dumped, fees and forfeits pay whoever is still in, and the farm extends that past the vest. Members were in from craft. Futarchy, if opted in, can move what the project already holds — not the sale, not the 31%.

They do one thing, together: give the token time. The contracts are the source of truth. This page is the map.