The 12-Month Cliff: Origin, Math, and When It’s Wrong

Almost every modern token launch puts insider tokens behind a 12-month cliff. Founders write it without thinking. Investors expect it without negotiating. The number isn’t arbitrary — but it isn’t universal either.

Where the standard came from

The 12-month cliff is older than crypto. It was lifted wholesale from Silicon Valley equity vesting, where the convention has been 4-year vest with a 1-year cliff since at least the 1990s. The mechanics match: an employee who quits in month 11 gets nothing; in month 13 they get 25% of grant; thereafter, monthly accrual to 100% at month 48.

Crypto inherited the schedule when it inherited the cap table. A founder with traditional VC backers in 2018 didn’t want to argue about vesting; their backers had already seen this template work in three previous portfolios. The 12-month cliff travelled with the term sheet.

What changed is what the cliff is actually for. In equity, the cliff protects the company against a co-founder leaving in month two with 25% of the cap table. In tokens, the cliff protects retail buyers against insiders dumping on day one. The mechanism is identical; the political function is opposite.

The math the cliff actually does

Consider an allocation of 20% of supply with a 36-month linear vest, no cliff. The unlocked fraction at month t is:

unlocked(t) = min(t / 36, 1)
// month 1: 2.78%
// month 6: 16.67%
// month 12: 33.33%
// month 36: 100%

The same allocation with a 12-month cliff plus 36-month linear vest:

unlocked(t) = 0                           // t < 12
unlocked(t) = min((t - 12) / 36, 1)        // t >= 12
// month 11: 0%
// month 12: 0%      <- cliff date
// month 13: 2.78%
// month 24: 33.33%
// month 48: 100%    <- fully unlocked

Two things to notice. First, the cliff doesn’t reduce total sell pressure — the same 20% supply is unlocked, just shifted. Second, it doesn’t actually spread sell pressure either; on month 12, holders receive zero, and on month 13 they receive a normal monthly tranche. There’s no accumulated overhang that hits all at once. The chart looks smooth.

So if the cliff doesn’t reduce or smooth sell pressure, what does it do? Two things, both political:

• It buys the project a known-quiet first year. The team can build, ship, and miss roadmap milestones without an insider unlock making them look like they exited. Critics can’t accuse a founder of dumping in month 6 if it’s mathematically impossible for them to dump.
• It creates a Schelling point for the market. Everyone knows the unlock cliff date. It becomes the most-watched event of year two — which is its own problem (covered below).

The 12-month cliff doesn’t smooth sell pressure. It buys you a quiet first year + a single, predictable, anxiety-inducing month-13.

Survey: cliffs across 13 real protocols

Here’s the team-allocation cliff across the 13 real public tokenomics encoded in this site, ordered by total team vest length (cliff + linear). Sources link to each project’s public docs.

Two things stand out. The 12 + 36 bundle is the modal choice — five of the most recent eight launches use it exactly. Older or unconventional projects (Uniswap, GMX, MakerDAO) skipped the cliff. Solana DeFi (Jito) compressed the linear portion to 24 months because Solana cycles are shorter — by month 36 most contributors have moved to a different protocol anyway.

When 12 months is wrong

1. Solana DeFi and other fast-cycle ecosystems

Jito uses 12 + 24 (3-year total) for team and investors. The reasoning, drawn from Solana DeFi launches in 2023–2024: Solana product cycles ship faster than EVM equivalents, headcount turnover is higher, and a 4-year lockup risks tokens being held by people who long since stopped contributing. A 36-month total schedule keeps real builders aligned without indenturing them to a project that may have pivoted.

2. Stablecoin issuers and RWA tokens

RWA / stablecoin templates often run 24-month cliffs (longer, not shorter). The reasoning is regulatory optics: an issuer holding 35% of supply for 24 months signals to regulators and institutional partners that the token is not a speculative vehicle for the founders. Compliance reserves on 10-year emission windows are common in this category.

3. Memecoins and fair launches

Memecoins typically have no cliff and no team allocation. The memecoin template puts 80% in liquidity, 10% in marketing, 10% in airdrop, all unlocked at TGE. GMX is similar in spirit: no VCs, no private sale, only a small (~7.5%) team allocation on a 24-month linear vest with no cliff. When a project is selling fairness as its core feature, a cliff becomes a liability — it implies you have something to hide.

4. Founders who already have skin in the game

Uniswap shipped UNI with zero cliff in 2020. Hayden Adams & the a16z-led team had been operating for two years already; a cliff would have bought another year of "trust us" they didn’t need. The protocol had real users, real volume, real revenue. The cliff is most useful for projects that haven’t yet earned that trust through shipped product.

The "cliff date" problem

The single biggest cost of a 12-month cliff is the date itself. Markets price in unlock cliffs aggressively. The most-discussed examples:

• Arbitrum’s March 2024 unlock cliff was the most-watched single date in DeFi that quarter. ARB drifted lower for weeks beforehand on positioning, then chopped sideways for months as actual unlocks arrived smaller than feared.
• Ethena’s April 2025 cliff generated the same dynamic — investor + core unlocks beginning 12 months after April 2024 TGE.

Both projects could have avoided some of this by shipping shorter cliffs (6 months) with linear vesting starting earlier. Daily/weekly micro-unlocks generate almost no narrative even when total unlock volume is identical. Nobody tweets about a 0.13%-of-supply unlock in month 4. Everyone tweets about a 7%-of-supply cliff in month 12.

Daily micro-unlocks generate no narrative. A single cliff date generates a Schelling point for shorts. The best vest has small unlocks often.

What to actually consider

Three questions that should determine your cliff choice:

• How much trust have you already earned? A team with a year of public shipping doesn’t need to buy another year via cliff. A pre-launch team probably does.
• What ecosystem are you in? Solana, NFT-adjacent, and meme launches tolerate (and sometimes prefer) shorter cliffs. EVM DeFi expects 12 + 36 by default; deviating downward signals confidence, deviating upward signals regulatory caution.
• What does the cliff date look like on a chart? If your linear vest starts the day after the cliff, the chart is smooth. If the cliff itself unlocks a chunk (say, 10% upfront after 12 months), you’ve created the worst-of-both: the team is locked out for a year, and then a single date dumps a stockpile.

The default 12 + 36 isn’t wrong. It’s just lazy. If you can articulate why your project deserves something different — shorter for fast-cycle ecosystems, longer for regulatory-sensitive categories, none for fair launches — you’ll signal more thought than the founder who copied the template.