90% of projects that launch a token shouldn't have. A token is a multiplier -- 10x on a great product, 10x on nothing is still nothing. This guide is for the 10% who have a real reason to tokenize and want to get the economics right.
The one question that matters
Remove the token from your product. Does the product still work?
Yes? The token is optional. Launch the product first. No? The token is core infrastructure. Design it carefully.
If you answered yes but still want a token, you need an ironclad value accrual mechanism. "The token will appreciate as the network grows" is not a mechanism. It's a prayer.
Four sustainable tokenomics models
Every working token economy falls into one of these patterns. Most dead tokens tried to be all four at once.
Model 1: fee capture
The protocol charges fees. Token holders receive a share.
Fee sources include trading fees, minting fees, and subscriptions. Distribution happens either directly to stakers or through buyback-and-distribute. The sustainability test: fees must exceed emissions, or you're diluting holders.
Who does this well: Aerodrome charges trading fees on every swap. veAERO lockers receive 100% of trading fees in the tokens being traded, not just more AERO. Revenue is real, tied to actual trading activity, and scales with volume. By early 2026, Aerodrome is generating substantial weekly fees from billions in volume on Base.
Who failed: dozens of "revenue sharing" tokens where the revenue was just other tokens. That's a Ponzi with extra steps.
Model 2: work token
Token required to perform work in the network. More demand for the service means more tokens staked, creating organic demand.
Works for oracle networks, validation, and curation markets. Doesn't work for products where the "work" is artificial. If you're making users stake tokens just to use a dashboard, you've added friction, not utility.
Model 3: access/utility plus burn
Token grants access to features or discounts. Usage burns or locks supply.
The trap: if access can be priced in USD, your token is a gift card with extra steps. The burn must be meaningful relative to supply, and the access must be genuinely valuable.
Model 4: governance plus treasury revenue
Token represents governance power over a productive treasury.
Only works if the treasury generates real revenue from protocol activity. A governance token over an empty treasury is a voting right over nothing.
Emission schedules: the math that kills projects
Your emission schedule is the most consequential decision in your tokenomics. Get it wrong and no amount of marketing saves you.
The wrong way: high initial emissions
Year 1 emitting 40% of supply, Year 2 emitting 30%, Year 3 emitting 20%, Year 4 emitting 10%. This is the default approach. The problem: Year 1 emissions create massive sell pressure before your protocol has enough organic demand to absorb it.
The right way: conservative start, adaptive middle
Year 1 emitting 8-12% (bootstrap carefully), Year 2 emitting 10-15% (increase only if usage justifies it), Year 3 emitting 8-10% (begin tapering), Year 4 onward under 5% annually, approaching zero.
Lower emissions that sustain longer always beat high emissions that burn out. A 5% APY that lasts 4 years builds more loyalty than a 500% APY that lasts 3 months.
Real numbers: Aerodrome's emission schedule
Aerodrome is the best case study of emissions done right on Base.
At launch in August 2023, emissions started at roughly 8M AERO per week. Weekly emissions decay by roughly 0.5-1% per epoch, with veAERO voters directing where emissions go. Emissions don't just spray everywhere -- voters direct them to specific pools. Protocols bribe voters to direct emissions to their pools, creating organic demand for AERO.
By mid-2024, Aerodrome was generating $1-3M per week in real trading fees. Emissions were declining while revenue was increasing -- the crossover point where the protocol becomes self-sustaining. By 2026, with Base TVL at roughly $10B, Aerodrome has become a case study in how ve(3,3) emissions create lasting liquidity infrastructure.
Compare this to a typical launch: 40% of supply in Year 1 with no mechanism to direct it productively. Aerodrome emitted less and got more because every emitted token had a job.
Halving vs. linear vs. adaptive
Halving (Bitcoin-style) is predictable with deflationary pressure but rigid. Best for L1s and long-term infrastructure. Linear decay is smooth and predictable but still rigid. Best for protocols with stable demand. Adaptive or governance-controlled emissions respond to market conditions but require active governance. Best for DeFi protocols with engaged communities. Epoch-based with KPIs ties emissions to real metrics, which is complex but creates natural feedback loops. Best for growth-stage protocols.
The most resilient approach: epoch-based emissions tied to protocol KPIs. If TVL grows 20%, next epoch's emissions increase 10%. If TVL drops, emissions decrease.
Real yield vs. inflationary rewards
This distinction has killed more portfolios than any exploit.
Inflationary yield: you stake TOKEN, you earn more TOKEN. The yield comes from new token emissions. You're being diluted and don't realize it because the number goes up.
Real yield: you stake TOKEN, you earn ETH or USDC from actual protocol revenue. The yield comes from users paying for a service.
The real yield checklist
Can you calculate protocol revenue in USD terms? Does that revenue exist without token emissions? Is yield paid in a non-native asset? Would the yield still be attractive at 5-10% APY? Can the protocol sustain this yield in a bear market?
If you answered no to any of these, your "real yield" is inflationary yield wearing a disguise.
Case study -- Morpho on Base: Morpho operates as a lending protocol where the yield comes from real borrowing demand. In February 2026, Apollo (the Wall Street asset management giant) announced the acquisition of Morpho tokens -- major institutional validation that Morpho's model generates real, sustainable revenue. When a TradFi firm with hundreds of billions under management buys into your token, the tokenomics have to be real. This is the clearest signal yet that DeFi on Base has moved beyond speculative farming into genuine financial infrastructure.
Anti-case study -- OlympusDAO forks: offered 7,000%+ APY through rebasing (printing new tokens). Worked while new money entered. The moment inflows slowed, the math collapsed. Most forks went to zero.
Vesting design: don't create your own sell pressure
Bad vesting creates predictable dumps. Good vesting aligns incentives over years.
Team and investor vesting
Minimum credible vesting: 12-month cliff, 36-month linear vest (total 48 months). No more than 20% to team at launch (optics matter enormously). Investors should have equal or longer vesting than team.
What investors actually check: team vesting shorter than investors is a red flag. Large unlock events mean they'll front-run it. No cliff means they assume you'll dump.
Community and farming vesting
Farming rewards with zero vesting create instant sell pressure. Every token emitted to farmers with daily unlocks is a market sell order waiting to happen.
Better approaches: the veToken model locks tokens for 1-4 years to earn boosted rewards and governance power, with longer locks meaning more weight. This is what Curve pioneered and Aerodrome adapted for Base. Epoch-based partial vesting releases 30% liquid with 70% vesting over 6 months. Bond mechanics let users buy discounted tokens with LP positions, receiving tokens over 5-7 days.
Unlock schedule template
Month 0-12: team locked, investors locked, community receives 10% of allocation. Month 12: team cliff releases 10%, investors cliff releases 15%. Month 12-48: team linear monthly unlock. Month 12-36: investor linear monthly unlock. Month 0-48: community emissions per epoch (governance-controlled). Month 48: full circulation.
Track unlock events with tools like Sonarbot -- set alerts for major unlocks before they hit the market.
Death spiral prevention
Every token project needs to understand exactly how death spirals work, because if you don't actively prevent one, the default outcome is one.
The death spiral sequence
Launch with high emissions so farming rewards look juicy. Mercenary capital floods in and TVL looks great on paper. Farmers sell rewards immediately, creating sell pressure that exceeds buy pressure. Token price drops, yields drop in USD terms, farmers leave. TVL collapses, remaining users panic sell. Ghost protocol.
Five circuit breakers
First, minimize emissions from day one. You cannot market your way out of bad math. If you're emitting 5% of supply monthly, you need 5% monthly buy pressure just to stay flat.
Second, lock periods on rewards. Vesting farming rewards over months (not days) smooths sell pressure and filters for long-term participants.
Third, revenue-backed rewards over inflationary. Distribute protocol revenue, not new tokens. If there's no revenue, your yield is fake.
Fourth, utility before speculation. Launch the product before the token. If users use your product without a token, they'll stay when the token exists.
Fifth, usage-driven burns. Not artificial burns. Burns that happen because people use the product -- transaction fees burned, access fees burned, failed-commitment penalties burned.
Case studies: what worked, what didn't
Worked: Aerodrome's ve(3,3) model on Base
Forced long-term alignment through vote-locking. Want maximum rewards? Lock for years. Bribing wars create organic demand for the token as protocols compete for liquidity direction. Emissions serve a purpose (directing liquidity to where traders need it) rather than just attracting TVL. By 2026, Aerodrome has proven this model works at scale on Base.
Worked: Virtuals Protocol
Virtuals built an AI agent launchpad on Base that has processed over $477M through autonomous agents. Their tokenomics tie into "Agentic GDP" -- a metric that captures value flowing through AI agents on the platform. The token accrues value because agents need it to operate. This is Model 2 (work token) done right: real demand from real usage, not artificial staking requirements.
Failed: LUNA/UST -- anatomy of a $40B death spiral
The most important case study in tokenomics history. To mint 1 UST (stablecoin), you burned $1 of LUNA. To redeem 1 UST, you minted $1 of LUNA. By April 2022, UST had roughly $18B in circulation with Anchor offering 19.5% APY. In May 2022, large UST withdrawals caused UST to slip to $0.98. Holders panic-redeemed for LUNA. Redeeming UST minted new LUNA, which crashed LUNA's price, which meant each UST redeemed for more LUNA tokens. LUNA went from roughly 350M tokens to 6.5 trillion in 5 days. Price went from $80 to $0.0001. $40B+ in combined value: gone.
The lesson: never design a mechanism where declining price increases token supply.
Failed: friend.tech
Launched on Base in 2023 with a bonding curve model for "social tokens." The tokenomics were entirely speculative -- keys derived value from perceived social capital with no underlying revenue or utility. When attention moved on, there was no floor. The project shut down entirely, becoming one of the most visible failures on Base. The lesson: social mechanics need real value accrual, not just reflexive speculation.
Pre-launch tokenomics checklist
Before you deploy that token contract, verify every item:
- Token has a specific behavior it incentivizes (not just "holding")
- Sybil resistance plan for farming incentives
- Identified at least one organic buy pressure mechanism
- Stress-tested the model in a 2-year bear market scenario
- Protocol can generate revenue without token appreciation
- Emission schedule approaches zero within 4-5 years
- Team vesting is 12+ month cliff, 36+ month vest
- No single unlock event releases more than 5% of supply
- Governance has meaningful lock requirements
- You've honestly asked: "Would a points system work instead?"
Tokenomics health check
Run your model through these stress tests before launch and quarterly after.
Revenue and sustainability
What is your protocol's monthly revenue in USD? If zero, all yield is inflationary. At current emission rates, how many months until revenue exceeds emissions? If the answer is "never" or "we need 100x growth," redesign. Remove all token incentives -- do users still use the product?
Supply and sell pressure
What percentage of circulating supply unlocks in the next 90 days? Over 10% is a red flag. Over 20% is a crisis. Who are your top 10 holders, and what's their cost basis? If they're 10x+ in profit with no lock, expect selling. What's your daily emission rate as a percentage of circulating supply? Over 0.5% per day is extremely aggressive.
Demand and buy pressure
Name three reasons someone buys your token that aren't "number go up." What happens to token demand if your product doubles users? Is there a lock/stake mechanism with meaningful participation?
Death spiral test
If your token drops 80%, does any mechanism accelerate the decline? If all farming rewards stopped tomorrow, what percentage of TVL leaves? Model a scenario where your largest holder dumps their entire position.
Token launch paths on Base in 2026
The meta has shifted. You have several paths to launch:
Bankr is the fastest social-first launch. Deploy directly from X by tagging @bankr. Under the hood, Bankr uses Doppler -- an onchain protocol for price discovery auctions with configurable bonding curves. Doppler supports multicurve auctions with market cap ranges, supply curves, vesting schedules, and treasury allocation. You can also use Doppler's SDK (@whetstone-research/doppler-sdk) directly for more control. Clanker is a competing launcher with lower fees. Best for tokens where community momentum matters more than precision engineering.
Uniswap V3 is for serious DeFi launches with concentrated liquidity control. Best when you need precise price management.
Aerodrome is for tokens that need deep, sustainable liquidity through the ve(3,3) model. Best for protocols planning to build lasting infrastructure.
Each path has different tokenomics implications. Bankr/Doppler launches favor rapid distribution with built-in bonding curve price discovery. Uniswap V3 launches favor controlled initial pricing. Aerodrome launches favor long-term liquidity sustainability.
For the tactical side of liquidity, see the liquidity strategies guide. For the broader DeFi integration picture, read building DeFi on Base.
Monitor or die
Tokenomics don't run on autopilot. You need real-time visibility into emission rate vs. buy pressure, wallet concentration, vesting unlock events approaching, LP depth and slippage, and governance participation rates.
Use Sonarbot to track whale movements, liquidity changes, and price alerts on your token and competitors. Checkr provides real-time token insights that complement your monitoring setup. The projects that survive are the ones that see problems before they become crises.
The test
Can you explain your tokenomics to someone who's never heard of blockchain -- and have it still make sense?
If yes, you might have something. If you need a 30-page paper and a PhD in game theory, your users won't understand it either.
Tie rewards to real revenue. Vest aggressively. Emit conservatively. Build the product first, add the token only when it makes the product work better.