Barter is a top-3 Ethereum router — the solver system behind $35B+ of executed order flow on CoW Swap, 1inch Fusion, UniswapX, Bebop and Velora. Its token is a disciplined claim on that working business: metered on-chain, settled in USDC, and engineered so that nothing breaks when markets do.
01 · The business
A token can only be as honest as the business underneath it. Barter runs three complementary product lines — each earns differently, and they peak in different market weather.
Barter competes in execution auctions across six venues, winning the right to fill orders at the best price. It earns execution margin on every fill — and earns more when markets are volatile. Trading revenue is long volatility.
A self-custody market-making layer: idle balances quote swaps directly from users’ wallets and earn fees on every fill — no deposits, no pooled custody. Autopilot inventory also sharpens the solver’s edge.
A peer-to-peer orderbook for fixed-rate lending. Origination fees are collected in-code at match time. Credit earns in calm markets — the counterweight to the trading engine.
02 · The Conduit
A solver’s P&L lives off-chain, in spreads and inventory. Most projects would ask you to trust quarterly promises. We built a meter instead.
An immutable registry lists every Barter settlement contract, the Autopilot layer, and the Lend engine.
Every executed fill is a public, machine-readable event. Anyone can reconcile the meter against venue data.
The Conduit computes what the company owes and the company pays it publicly, in USDC, within 7 days.
Received revenue flows through a fixed, code-enforced waterfall. No discretion, no votes.
In-path protocol fees, taken atomically in settlement: a few hundredths of a percent on own-app and API fills, 10% of each Autopilot LP fee, and a small origination fee on Lend. Where no auction can be lost, an in-path fee costs nothing.
Zero in-path fees on auction flow — ever. A fee inside an auction bid is a worse bid: the market taught this lesson publicly when a venue-level fee change repriced solver economics overnight. Auction fills are metered and invoiced instead — the toll never touches the bid.
03 · The waterfall
After reserves are filled (12 months of operating cost, 120% of all outstanding service-credit liabilities, the insurance schedule), each remaining dollar splits four ways — enforced by code, denominated in percentages, never in promises:
Buybacks execute daily on trailing audited receipts — sized to run 30 months through a bear market without pausing. The endowment’s principal is unspendable by code, forever. Reinvestment pays Autopilot liquidity providers and integrators — in USDC, never in emissions.
Why not put everything into the burn? Because the data says holders deserve better. Research across dozens of revenue-paying tokens (the Effective Revenue Multiplier framework by Valueverse) shows a consistent pattern: tokens that only buy back and burn deliver far less tangible cashflow per dollar of eligible market cap than tokens that also distribute directly — burn value partly accrues to sellers on their way out, not to holders who stay. So the Barter Token does both: a continuous burn and quarterly USDC distributions to committed holders.
04 · Holding it
The daily buyback retires supply out of real revenue. Your share of the network compounds without you doing anything at all.
Bonded, 12-month-locked tokens form the eligible class: quarterly USDC distributions from the 20% leg, yield from the endowment, a 20% rebate on Barter service fees, and priority routing. Real dollars from real fills — never printed yield.
Bonded Autopilot wallets earn USDC from the Interchange Pool for providing liquidity. Locked tokens carry network capacity rights — delegable to third-party LPs for a market-priced share of their yield.
05 · Enforcement
The hard question for any revenue-backed token: why would the company actually keep paying? Our answer is a stack of mechanisms — each one checkable, none of them a promise.
Every team and investor claim requires invoiceDebt == 0. One unpaid invoice and all insider vesting halts — no council, no vote, no negotiation. The team is mechanically paid after holders, always.
A compliance bond — tokens plus $250K in USDC — slashable by an independent 3-of-5 council of named individuals if invoices go 30+ days delinquent or flow bypasses the registry.
The operating company sits under a foundation whose only value-capturing asset is the token. No second pocket, no equity path that quietly siphons what the token was promised.
A determined bad-faith team could still starve the meter. The design makes that visible and expensive — not impossible. You should price that residual. We’d rather tell you than have you find out.
The launch gate: the token does not launch until the meter has run publicly for 12 consecutive weeks — real monthly USDC invoices, paid on time, verifiable by anyone — and two independent audits come back clean. If either slips, launch slips. Revenue first. Token second.
06 · The stress test
Every token design should answer one question before launch: if the token gapped down 95% overnight, what breaks?
Autopilot funds sit in users’ own wallets behind revocable approvals — zero linkage. Lend positions are externally collateralized; the token is not loan collateral in year one. Service credits are 120% reserved in stables. The insurance fund holds only USDC — never the token. Auctions are won and lost exactly as before. The buyback simply retires 20× more supply per dollar.
History’s worst token failures — Terra, ICHI — shared one trait: the token’s price was load-bearing for someone’s principal, so a price crash became a system crash. Here the only casualty of a −95% gap is holder P&L. The business, user funds and credit markets don’t notice.
07 · Discipline
Each rule below exists because someone else paid for the lesson. The precedent is named — look it up.
08 · The scoreboard
The token is built to be measured. Its valuation dashboard follows the Effective Revenue Multiplier methodology (Valueverse / Vasily Sumanov): count only revenue-eligible supply, count only cash actually delivered.
| Metric | What it tells you |
|---|---|
| Receipts / 1,000 tokens | The fundamental: trailing audited revenue per unit of supply. The number that should compound. |
| ERM (buyback leg) | Eligible market cap ÷ annualized buyback spend. How the market prices $1 of burn-delivered revenue. |
| Claimable ERM | Locked-class market cap ÷ annualized USDC actually distributed. Our target: priced like a fee-sharing token, not a buyback-only one. |
| Net protocol profit | Receipts minus the dollar value of all tokens entering the float. With zero token incentives, positive by construction — a line no emissions-funded protocol can print. |
| Depth, both sides | Bid and ask depth at ±2 / 10 / 25% — published daily. Liquidity health is never measured in TVL. |
| Locked-float share | Time-averaged bonded-and-locked supply, with a lock-duration histogram. |
| Stress-window report | In any 7-day window with a >15% market drawdown: fee growth vs. price. Solver revenue is long volatility — we publish the proof each time. |
“Protocol revenue is not holder revenue. A token should be measured by what its holders actually receive — and be structured so that number is worth measuring.”The design principle behind the scoreboard
09 · Supply
The contract contains no mint function — anyone can verify this in the deployed bytecode. Every allocation is published; there are no hidden senior claims.
| Allocation | Share | The condition that matters |
|---|---|---|
| Community & points | 20% | Distributed by a formula published before launch and executed verbatim — earned by real usage, not engagement farming. Streaming gated by actual revenue. |
| Team | 20% | Unlocks only at cumulative revenue milestones — and only while every holder invoice is paid. No revenue, no team tokens. |
| Investors | 10% | 12-month cliff, 24-month vest, and the same revenue milestones — whichever comes later. |
| Ecosystem reserve | 15% | Emission capped by revenue already earned — faking volume to mint is negative-EV at any price. |
| Permanent endowment | 10% | Principal unspendable by code. Forever. |
| Treasury | 17% | Operations and insurance seeding; large sales require 90 days’ public notice. |
| Market liquidity | 8% | Laddered across the price curve — never concentrated into anything resembling a floor. |
10 · Questions
You shouldn’t — you should check them. Fills are public events; invoices are public transfers; the dashboard reconciles to the meter with a published error bound; an independent firm attests the accounting annually; and a standing bounty pays anyone who finds a discrepancy. Belief is not part of the design.
Revenue falls — and the machine keeps running. Reserves are sized for 12 months of operations before a single dollar reaches the waterfall, the buyback throttle is calibrated to run 30 months on zero new revenue, and the trading engine historically earns more in volatile sell-offs. What never happens: emergency emissions, un-burns, or quiet rule changes.
The design was built with regulatory caution throughout: distributions flow through code from protocol revenue, the structure is being established with specialist counsel, and availability may be restricted in some jurisdictions. This page describes mechanism design; it is not an offer.
Because the design punishes overpricing: above a fixed multiple of trailing receipts, the buyback automatically diverts to the endowment until the price cools. We would rather compound from an honest base than defend a headline. Slow, funded, and unkillable beats fast and fragile.
After 12 consecutive weeks of publicly paid invoices and two clean audits. That is the only calendar we publish.