# moo market ## a market for tokenized equities on base **whitepaper v1.3 ยท august 2026** stocks are onchain. the market isn't. moo. --- ## abstract Tokenized equities have arrived onchain. Wrapped Coinbase stock (wCOIN) trades on Base against stablecoins, and issuance of tokenized stocks is accelerating across chains. What has not arrived is market structure. Tokenized equities today have liquidity pools but no memberships, no income primitives native to the asset, no persistent identities, and no way for the growing population of autonomous trading agents to build portable, provable reputations around these assets. Moo Market introduces the **seat**: a fixed-supply onchain membership, denominated and priced in COIN, that earns yield from covered call premium written on pooled COIN, carries a transferable anonymous reputation via the B20 ring signature standard, and can be leased to human or autonomous operators without leaving its owner's wallet. There are exactly 1,000 seats. Each seat is a bull. This paper specifies the seat mechanism, the self-assessed pricing system, the Milkhouse yield engine, the B20 identity layer, the leasing market, the $MOO token's role, and the phased path from mint to a community-governed protocol. --- ## 1. background ### 1.1 tokenized equities are real now The first generation of onchain stock exposure, synthetics collateralized by crypto, failed on oracle risk and capital inefficiency. The second generation is different: fully-backed 1:1 wrappers issued against custodied shares, redeemable, and increasingly issued or endorsed by the brokerages themselves. Coinbase stock is the emblematic asset. It trades onchain, on the chain Coinbase built, and its price is substantially a bet on the success of onchain markets themselves. The reflexivity is the point. ### 1.2 what is missing An asset is not a market. Compare what COIN has onchain today with what any equity has on a traditional venue: | function | traditional exchange | tokenized equities today | |---|---|---| | trading | matched orderbook | AMM pools | | membership | seats, with rights and fees | none | | income on the asset | dividends, options premium, lending | none for non-dividend stocks | | identity | licensed brokers, track records | anonymous wallets, no memory | | delegation | brokers act for clients | none, or full custody transfer | The trading row is solved. The other four rows are open. Moo Market is a product for those four rows, starting with COIN because it is the deepest and most narratively coherent tokenized equity on Base. ### 1.3 why seats Exchange seats are one of the oldest financial primitives. An NYSE seat was simultaneously an access right, an income stream, a bond of good behavior, and a tradable asset whose price tracked the health of the market itself. Seats were leased more often than sold. Seat prices were front-page indicators of market confidence. The seat model maps cleanly onchain, and it maps especially well to a world where many market participants will be autonomous agents. Agents need identity that is provable but not doxxing, reputation that survives wallet rotation, and a way to rent access instead of raising capital to buy it. Seats provide all three. --- ## 2. system overview Moo Market is four core contracts around one asset: MooSeats (1,000 ERC-721 seats, self price in wCOIN, $MOO-gated mint) connected to the Milkhouse (wCOIN pot, covered calls, premium per epoch), SeatLease (rent in wCOIN, owner keeps the NFT, tenant operates), and the B20 Registry (ring membership, rep per seat, LSAG signatures). Peripheral contracts join around the core: SeatVault (every seat's own bag) and AuctionBarn (the herd's native market) from phase 1, CallVault (options settlement) and StallGate (expansion gates) in later phases (section 9). Everything is denominated in wCOIN. There is no separate reward token, no points system, and no emission schedule. The unit of the market is the stock. The $MOO token exists for exactly one protocol pattern: membership gating (section 7). It is checked at every door the protocol opens, and it is deliberately not the reward asset, not the pricing asset, and never the unit of account. --- ## 3. seats ### 3.1 supply and form 1,000 seats, ERC-721, token IDs 1 through 1000, displayed as seat #0001 through seat #1000. The art is a herd of cow characters. Metadata is revealed post-mint via a baseURI update and then frozen. Supply is fixed in the contract: no owner mint beyond a disclosed team allocation of 25 seats (2.5%) minted at deploy for partnerships, exhibitions, and protocol operations, and no ability to increase supply. ### 3.2 mint Mint opens after the $MOO token launch. Eligibility requires holding at least `minMooBalance` of $MOO at mint time (owner-settable before mint opens, frozen after). Per-wallet cap of 2 seats. The mint gate is a balance check, not a payment. $MOO is not spent, transferred, or locked by the mint. This keeps the token's role clean: it is a coordination filter, not a toll. Minting also plants a **seed**: `seedAmount` of wCOIN, pulled from the minter and deposited into the new seat's own vault (section 3.4). The protocol keeps none of it. There is no mint fee, no treasury cut, and no split: everything a minter pays lands in the seat that minter just received. The seed is paid directly in wCOIN rather than swapped from ETH at mint time, because a swap in the mint path would be sandwichable on thin liquidity and would put a price dependency where the protocol has none. The seed is permanent seat capital: it can never be withdrawn, and it is realized by selling the seat, since a buyer pays for the seat including its bag. A refundable seed would make minting free in every economic sense and every bag would be empty within a day. `seedAmount` is published before mint opens, frozen when it does, and may be zero. The 25 team seats are seeded identically from team funds. ### 3.3 self price Every seat carries a `selfPrice` denominated in wCOIN, set by its owner, visible to everyone. In phase 1 this is informational: an always-on, owner-signed statement of what the seat is worth, aggregated into a protocol-wide seat price index that serves the same signaling role NYSE seat prices once did. In phase 2 the self price becomes binding under a Harberger-lite regime: - the owner pays a continuous tax of `taxRate` per year on the self price, paid in wCOIN, streamed to the Milkhouse pot - anyone may buy any seat at its self price at any time; 97.5% of the payment in wCOIN goes to the owner and 2.5% to the pot, ownership transfers atomically, activation state resets - protections: a 7-day grace period after a takeover during which the seat cannot be taken again (an ordinary barn sale does not grant grace, so a seat cannot farm immunity by trading with itself), a default self price of 1.5x the last takeover price if an owner never sets one, and a minimum self price floor set by governance The tax rate is the key parameter. It launches low (1 to 2% annually) so the regime feels like a listing fee rather than a threat. It is adjustable downward by the multisig and upward only by seat-holder vote (section 8). Design intent: the tax converts idle speculation into pot revenue, the takeover right guarantees a liquid exit and entry at all times without an AMM vault holding inventory, and the self price index becomes the market's public heartbeat. ### 3.4 the seat vault Every seat carries its own onchain account: a token-bound vault (ERC-6551 addressing, restricted implementation) that holds wCOIN and travels with the seat through every sale, takeover, and transfer. The owner's share of milk lands in the vault by default, so an unattended seat compounds its own bag. The mint seed sits in the vault as a permanent earmark, and phase 2's activation lock as a temporary one; free balance is always the total minus every earmark, and only free balance is ever withdrawable. The activation lock returns to the departing owner's wallet on any transfer; the seed never leaves the seat. Withdrawing from a vault takes a 24-hour timelock, and any transfer of the seat cancels pending withdrawals: what a buyer sees in the bag is what the buyer gets. The vault cannot make arbitrary calls or grant allowances, so no previous owner can leave a hole in it. A seat's rational self price gains a visible floor, the bag plus the market's price for the seat's rep and membership; the protocol does not enforce that floor, visibility does the work. ### 3.5 the auction barn Seats trade on their own market from day one. The AuctionBarn takes listings and escrowed offers, priced in wCOIN only, and settles atomically: 97.5% to the seller, 2.5% to the Milkhouse pot, seat and vault to the buyer. A barn sale writes the sale price into the seat's `selfPrice`, so the public price index reflects real trades, not just declarations. When phase 2's Harberger regime activates, fixed-price listings retire (a live takeover price makes them redundant) and the barn remains for below-ask offers; the barn fee and the takeover fee are the same 2.5%, one number for both paths, both feeding the pot. Full mechanics and invariants: `docs/specs/seat-vaults-and-barn.md`. ### 3.6 what a seat is not A seat is not equity in a company, not a claim on Moo Market revenues beyond the mechanisms described here, not governance over anything except the parameters in section 8, and not a promise of profit. A seat is a membership that participates in a rules-based distribution of option premium generated from pooled assets, plus an identity and delegation primitive. Section 10 discusses regulatory posture. --- ## 4. the milkhouse ### 4.1 the problem: COIN pays no dividend Coinbase stock pays no dividend. A pot of wCOIN sitting still produces nothing. But COIN is one of the most volatile large-cap equities in the world, and volatility is sellable. The Milkhouse manufactures a dividend from that volatility by writing covered calls on the pot. ### 4.2 mechanism The Milkhouse is a vault holding wCOIN, funded by: - Harberger tax streams (phase 2) - auction barn fees: 2.5% of every seat sale (section 3.5), live from phase 1 - lease protocol fees: 5% of rent (section 6, phase 2) - seat takeover fees: 2.5% of every takeover price (phase 2) - graze fees: outside projects pay a flat published fee for an attested herd list (section 5.5, phase 1) - direct contributions (anyone may fund the pot) Each epoch (default 14 days), the Milkhouse writes European covered calls against a bounded fraction of the pot (`writableFraction`, default 50%, governance-adjustable within [25%, 75%]): - **strike selection.** default 110% of the epoch-open TWAP of COIN/USDC from a manipulation-resistant oracle; phase 2 allows seat holders to vote strikes per epoch through ring-signed ballots - **sale.** counterparties buy the calls by paying premium in wCOIN or USDC. phase 2 launches with fixed-price sales to whitelisted market makers, moving to sealed-bid auctions - **settlement.** cash-settled in wCOIN against the oracle TWAP at expiry. if the call expires in the money, the pot pays the difference from the covered portion. the pot never sells more optionality than it holds, so it cannot be liquidated Premium collected in an epoch, plus residuals, less the milking tip (section 4.4), defines `epochMilk`. Each seat can claim: ``` milk(seat) = epochMilk x weight(seat) / sum(weights) ``` ### 4.3 weights and activation Base weight is 1.00 per seat. Activation (phase 2) lets a holder boost weight by locking wCOIN in the seat's name for the epoch: locked amounts up to caps map weights from 1.00 to 1.50. Unlike burn-based tier systems, the lock is returned at epoch end; the boost compensates the seat for its lock's contribution to writable pot size. Activation resets on transfer or takeover. ### 4.4 milking hour and claiming Epoch settlement is permissionless: once an epoch's duration has elapsed, anyone may ring it in, and the first caller earns a milking tip of 0.25% of that epoch's distribution, capped at 5 wCOIN. Calling early reverts, there is no randomness anywhere in settlement, and the tip is a keeper incentive sized to cover gas, never a yield product. If the team disappears, the herd keeps milking. Claims are pull-based per seat per epoch, callable by the seat owner or its current tenant according to the lease split (section 6). The owner's share lands in the seat's vault by default (section 3.4); the tenant's share pays the tenant directly. Unclaimed milk rolls into the next epoch's pot after a 6-epoch claim window rather than accumulating as liabilities. ### 4.5 honest accounting of the yield Covered calls cap upside: in epochs where COIN rallies through the strike, the pot underperforms simply holding. The strategy harvests volatility risk premium, which is positive in expectation over long horizons for most equities but arrives unevenly. The protocol commits to publishing per epoch: premium collected, strikes, settlement outcomes, and the counterfactual buy-and-hold comparison. No APY marketing. The milk is what the milk is. --- ## 5. b20 identity ### 5.1 the standard B20 is a ring-signature-based reputation standard. Members of a fixed ring can sign messages proving that some member signed, without revealing which one, using linkable spontaneous anonymous group (LSAG) signatures. Linkability means two signatures by the same hidden member in the same scope are detectably same-signer, which enables per-member reputation without identity disclosure. ### 5.2 seats as ring members The 1,000 seats form the Moo Market ring. Each seat controls a signing key slot, held by the owner or by the tenant during a lease. Ring-signed actions include: - epoch strike votes for the Milkhouse - market calls: directional predictions on COIN posted to the registry, timestamped, and scored against realized prices - attestations within the broader B20 ecosystem, where a Moo Market seat signature carries the weight of a priced, taxed, yield-bearing membership rather than a free wallet ### 5.3 rep Each seat accrues a rep score from scored actions: accuracy of market calls, participation streaks, tenure. Rep is computed from linkable signature history, stored against the seat ID, and transfers with the seat. A takeover or sale acquires the track record. Rep is therefore a capital asset: a seat with a long profitable calling history should command a higher self price, and the tax regime makes that premium legible. Rep can decrease. Wrong calls score negative. Inactivity decays rep toward a neutral baseline at 5% per epoch. Rep cannot be bought directly; it can only be earned by a signer or acquired by buying the seat that earned it. ### 5.4 bonded calls A seat may escrow a flat, refundable $MOO call bond to badge its market calls as bonded (section 7.2). The bond never changes rep arithmetic and returns in full regardless of outcomes; it exists to make a track record costlier to fake and is slashed only on mechanically provable abuse. Bond tiers are fixed protocol-wide so bond size never fingerprints a signer inside the ring. ### 5.5 graze fees Any project may pay a flat published fee in wCOIN to the pot and receive an **attested herd list**: the seat IDs and current owners meeting a rep threshold of their choosing, signed by the registry at a specific block. The price is public and identical for everyone, with no curation, no allowlist of buyers, and no governance decision about who may purchase. The data is derived entirely from public onchain state, so nothing private is collected or revealed and ring activity is never deanonymized: this is an attestation over facts anyone can already read. Every other revenue line in the protocol comes from our own churn or from selling volatility. This one grows with the wider ecosystem: the more projects on Base want distribution that is not sybils, the more the pot earns, and the more a seat is worth to someone who does not own one yet. The protocol promises nothing about which projects buy a list or what any seat holder receives from them; a purchased list is not an airdrop guarantee. ### 5.6 privacy properties Observers learn that a seat in the ring acted, and can track a persistent pseudonymous actor across actions in the same scope, but cannot map actions to a wallet and cannot link a seat's activity across scopes with different linkability tags. Signing keys rotate on transfer, mandatorily on takeover. --- ## 6. leasing ### 6.1 mechanism A seat owner lists the seat for lease: term length, rent in wCOIN paid upfront, and milk split (the fraction of epoch milk flowing to the tenant during the term). A tenant accepts by paying rent. For the term: - the tenant holds the seat's B20 signing slot and accrues rep to the seat - milk is split per the listed ratio, enforced by the Milkhouse at claim time - the NFT never moves. sales and takeovers transfer the seat subject to the lease, rent unaffected - at expiry the slot reverts to the owner; renewal requires a new agreement - operating the slot requires the tenant to keep an operator pass: a $MOO TWAB above `minGrazeBalance` for the term (section 7.1). if the average lapses, the signing slot suspends until restored; rent, the milk split, and the lease itself are unaffected Protocol fee: 5% of rent, to the Milkhouse. The fee is payable in wCOIN only. ### 6.2 why this matters Leasing is the on-ramp for operators who should not need the full capital cost of a seat to participate, and it is the natural interface for autonomous agents. An agent with a strategy but no capital rents a seat, performs under its ring identity, builds the seat's rep, and creates a track record that raises both future rents and the seat's self price. Owners become landlords of trading identities. The protocol becomes a labor market for market participation, with the stock as its unit of account. --- ## 7. the $MOO token $MOO has one protocol pattern: it is the membership filter at every door. It began as the mint gate, and it extends, on the same technical and legal shape, to every future moment where the protocol admits a member, an operator, or a new market. It is checked, never spent. It is never the reward, never the unit of account, and never a vote. ### 7.1 the gate pattern Every $MOO gate is the same primitive: a balance check. Nothing is transferred, locked, or burned by a gate; holding is the only way through it. The mint gate reads the balance at mint time; the later gates, which arrive phases after the token is liquid, read a 30-day time-weighted average (TWAB) so they reward sustained holding rather than just-in-time buying. - **mint gate (phase 1).** minting a seat requires holding at least `minMooBalance` at mint time, two seats per wallet (section 3.2). - **operator pass (phase 2).** operating a leased seat's signing slot requires the tenant to keep a TWAB above `minGrazeBalance` for the term. the threshold is flat (no rep tiers), launches low, is grandfathered per lease at signing, and moves only by a published formula inside seat-voted bounds. a lapsed pass suspends the slot until restored and touches nothing else. - **stall pass (phase 3).** every new parallel pot (a "stall") admits seats whose controller held a TWAB above the stall gate at stall-in. the gate is a fixed fraction of the original mint threshold, lowerable only by seat vote. stall-in is per seat, per pot, and stays open after launch under the same check. a granted stamp is permanent and travels with the seat; the founding COIN pot is grandfathered to all 1,000 seats. every stall opens on a bell that anyone may ring at the scheduled boundary, and seats that stall in during the first window are marked founding members of that pot: a permanent marker that is cosmetic and reputational, paying nothing and weighting nothing. an early-member bonus was considered and rejected, because the milk would have to come from later members or from emissions, and milk is sold volatility or it is nothing. ### 7.2 the two exceptions to "never spent" Two mechanisms move $MOO. They are named here precisely because nothing else does: - **call bonds (phase 2).** a seat may escrow a flat, refundable $MOO bond to badge its b20 market calls as bonded (section 5.4). the bond returns in full regardless of call outcomes and is slashed, burned to the zero address, only on mechanically provable abuse: duplicate-flood posting, oracle-window griefing, ballot spam. accuracy moves rep; the bond only makes the record costlier to fake. an outcome-linked variant (bond burned on wrong calls) is specified as an optional additive scoring module that only a seat vote can enable, and only after independent legal review. - **stall nominations (phase 3).** proposing a new stall burns a $MOO application fee, paid by the proposing community, never by the treasury or any protocol revenue. the burn buys agenda space only; seats hold the only binding vote, and stall admission ballots count owner-held votes exclusively. admission ballots are plain owner-signed votes, not ring-signed: stall admission is an ownership decision, so it runs outside the b20 ring, never touches a leased signing slot, and is excluded from any lease's voting rights. the fee is anchored to the real per-stall cost of oracle setup, audit, and operations, so its characterization as an application fee is economically true. ### 7.3 non-roles Deliberate non-roles, unchanged from v1.0 and load-bearing: - $MOO is not the reward asset. milk is wCOIN. - $MOO is not the pricing asset. self prices, rents, taxes, fees, and takeovers are wCOIN, and protocol fees are payable in wCOIN only. - $MOO earns nothing. no mechanism pays $MOO, pays for holding it, or grants milk weight for it. Milkhouse activation locks remain wCOIN only (section 4.3). - $MOO carries no governance rights. governance sits with seats (section 8). ### 7.4 the zero test Every $MOO mechanism must pass one design law before it ships: **if $MOO trades at zero, the protocol must not notice.** Gates become trivially satisfiable; existing stamps, leases, bonds, and rep are untouched; no pot revenue, no settlement, and no seat mechanic depends on a $MOO price, and there is no $MOO price oracle anywhere in the system. Mechanisms that fail this test, including revenue sharing, fee discounts, activation locks in $MOO, and price-linked burns, were considered during design and rejected categorically. The zero test itself is never governable. ### 7.5 launch parameters Launch parameters (venue, supply, distribution) will be published before the token launch and are outside this paper's scope. --- ## 8. governance Phase 1 is admin-operated through a public multisig with timelocked parameter changes. Phase 2 moves the listed parameters to seat-holder voting via ring-signed ballots, one seat one vote, with the tenant voting when a lease includes voting rights: - `taxRate`: upward moves only by seat vote; downward by multisig - `writableFraction` within [25%, 75%] - strike policy per epoch - epoch length within [7, 28] days; claim window length - treasury spends above 5% of the pot - the minimum self price floor (section 3.3) - the vault withdrawal delay within [12h, 72h] and the barn fee within [1%, 5%] - the milking tip within [0%, 1%] and its cap within [0, 25] wCOIN, and the graze fee - the operator pass formula bounds and the stall gate fraction (downward only) - the call bond tier schedule and its mechanical slashing list (additive modules only) - stall nomination fee anchoring, per listing Never governable: seat supply, the wCOIN denomination, the mint gate after opening, `seedAmount` once mint opens, the permanence of planted seeds, the core rep scoring algorithm (only additive scoring modules can be voted in), and the zero test (section 7.4). --- ## 9. contract architecture Solidity ^0.8.24, Foundry, Base mainnet (8453), staged on Base Sepolia (84532). | contract | responsibility | phase | |---|---|---| | `MooSeats` | ERC-721, mint gate, seed collection into the new seat's vault, self price storage; takeover and tax streaming in phase 2 | 1 -> 2 | | `Milkhouse` | pot custody, permissionless epoch settlement with the milking tip, claims, lease-aware splits, vault routing | 1 (funded distributor) -> 2 (options-fed) | | `SeatVault` | each seat's restricted token-bound account: deposits, timelocked withdrawals, permanent seed earmark, activation lock | 1 | | `AuctionBarn` | native seat market: listings, offers, 2.5% fee to the pot, self price sync | 1 | | `CallVault` | covered call issuance, sale, oracle settlement | 2 | | `SeatLease` | listings, rent, slot delegation, fee routing | 2 | | `B20Registry` | ring state, key slots, signature verification, rep, attested herd lists, founding markers | 1 (registry) -> 2 (LSAG verify, scoring) | | `StallGate` | per-pot stall-in TWAB checks, stamps, nomination burns | 3 | | `MooToken` | ERC-20 with balance checkpointing (ERC20Votes-style) so the phase-2 and phase-3 gates can read 30-day averages; launched separately | pre-mint | Operator passes live inside `SeatLease`; call bonds live inside `B20Registry`. All gates read balances only; none of them custody $MOO. Security posture: pull payments everywhere, reentrancy guards, no upgradeable proxy on MooSeats (immutable core; peripheral contracts replaceable by governance), external audit before phase 2 contracts hold meaningful pot value, public bug bounty from mint day. Oracle: COIN/USD via TWAP over the deepest wCOIN pool, cross-checked against a Chainlink COIN feed if and when one exists on Base. Settlement pauses rather than settles if sources diverge beyond a threshold. Fail closed, always. --- ## 10. risk and regulatory posture **market risk.** the pot holds a single volatile equity. covered calls cap upside and do not protect downside. milk can be small or zero in some epochs. seat prices can fall. **smart contract risk.** despite audits and testing, contracts can fail. phase gating keeps pot value low until the options stack is audited. seat vaults add custody surface (mitigated by restricted accounts and withdrawal timelocks), and vault withdrawals take 24 hours by design. **oracle and settlement risk.** TWAP manipulation and feed divergence are mitigated, not eliminated. settlement pauses fail closed. **wrapper risk.** wCOIN is an issued wrapper with issuer, custody, and redemption dependencies outside this protocol's control. a depeg or halt in the wrapper propagates to the pot and every denomination in the system. **regulatory.** seats are memberships in a rules-based distribution system, not shares. milk is option premium from pooled assets, not a dividend. the protocol does not custody user equities, does not intermediate stock trades, and does not solicit. tokenized equities are jurisdiction-restricted instruments, typically unavailable to US persons under issuer terms, and the interface will mirror the issuer's geoblocking. none of this is legal advice, the design may change to meet regulation, and users are responsible for their own compliance. **token risk.** $MOO is a membership filter (section 7) with no cash flow rights. it should be expected to be volatile and may lose all value without impairing seat mechanics. --- ## 11. roadmap **phase 0 - herd formation (now).** brand, community, $MOO launch parameters published. **phase 1 - the pen opens.** $MOO launches. mint opens under the holding gate. seat vaults live: every seat is minted carrying its seed, and milk lands in the same bag. the auction barn opens with 2.5% of every sale feeding the pot. milking hour is permissionless from the first epoch, and graze fees open as soon as rep has enough history to filter on. self prices live (informational, synced by barn sales). Milkhouse v1 distributes funded wCOIN by epoch. B20 registry live with seat key slots. site: mint, milk, herd gallery, the barn. **phase 2 - the market wakes.** CallVault live: real covered call epochs with oracle settlement. Harberger regime activates with grace protections; barn listings retire into standing takeover prices. SeatLease live with operator passes gating tenancy. LSAG verification, scored market calls, and call bonds. governance to seats. **phase 3 - more stalls.** additional tokenized equities as parallel pots under the same 1,000 seats, admitted through stall nominations, opened on a bell, and joined through stall passes, seat-directed allocation between pots, and external B20 integrations making Moo Market rep portable collateral for agents across Base. --- ## 12. closing Every cycle rebuilds the same discovery: assets are easy, markets are hard. Tokenized equities have completed the asset step. The market step needs membership, income, identity, and delegation, in forms that both humans and agents can hold. Moo Market's answer is deliberately small: one thousand seats, priced in the stock they serve, milked from its volatility, remembering everything they do. stocks are onchain. the market isn't. moo. --- *moo market - built on base - x.com/moomarketHQ* *this document describes intended protocol mechanics and is not an offer, solicitation, or financial advice. mechanics may change before and after launch. parameters published at deployment supersede this paper.*