How to read a token’s supply schedule: circulating supply, FDV, unlocks and dilution

Who this is for: Analysts, allocators, founders and serious retail holders who want to read a token’s supply page, vesting schedule and unlock calendar the way a credit analyst reads a debt maturity table, and turn it into a dilution and sell-pressure estimate they can defend.

Most people price a token by one number: the market cap on a price aggregator. That is price multiplied by circulating supply, and circulating supply is the figure that changes most over a token’s first three years. A project that launches with 15% of its tokens trading is not a $1bn asset because its market cap says so; it is an asset whose supply will roughly sextuple, on a published timetable, into whatever demand exists at the time.

The years 2025 and 2026 have made this impossible to ignore. Tokenomist tallied $97.43bn in token emissions in 2025, and in 2026 the market has watched the first team unlocks at Hyperliquid, a $160m cliff at Plasma, a Pump.fun cliff equal to a fifth of its float and a DoubleZero release equal to nearly half of its circulating supply. At the same time, protocols with real revenue have begun buying back and burning their own tokens at a scale that did not exist two years ago.

This guide explains the four supply figures, allocation and vesting mechanics, unlock data and what the studies say about it, inflation, burns and buybacks, how market makers distort float, and the red flags. It ends with a worked 24-month dilution model of a real token, a checklist and a glossary.

Token supply by the numbers

The four supply numbers and what each one hides

Every token page shows four figures that sound interchangeable and are not.

Max supply

The hard ceiling written into the protocol or token contract. Bitcoin’s is 21 million; Hyperliquid’s HYPE has a fixed 1 billion. Some tokens have no cap: Plasma’s XPL has a 10 billion initial supply but an unlimited max supply because validator rewards will add tokens indefinitely once staking is live. A missing cap is not automatically bad (Ethereum has none), but it means dilution never fully ends and you must model the inflation rate instead.

Total supply

Tokens that exist on chain today, minus any verifiably burned. This includes tokens sitting in locked vesting contracts, foundation treasuries and unclaimed airdrop pools. On October 6, 2026, CoinGecko listed HYPE’s total supply at about 955 million against a 1 billion max.

Circulating supply

The tokens that can actually be sold today: the number that matters most and the one with the softest definition. CoinGecko’s methodology says circulating supply is obtained from token teams and verified by its staff; for smart-contract tokens it subtracts known locked balances (foundation, investor and team wallets supplied by the project) from total supply, and leaves the field blank where it cannot verify. Circulating supply is therefore a self-reported figure an aggregator checks, not an on-chain fact, and two aggregators can disagree on the same day.

Market cap versus FDV

Market cap is price multiplied by circulating supply. FDV is price multiplied by max supply (or total supply where there is no max). The ratio between them tells you how much of the eventual supply the current price has yet to absorb. HYPE on October 6, 2026 traded near $93.48 with a $20.8bn market cap and an $89.3bn FDV, a ratio of about 23%. Plasma’s XPL the same day showed a $433m market cap against a $956m FDV, about 45%, because its one-year cliff had just released a large tranche. Neither number is “right”: market cap understates the claims on the project’s future, FDV overstates what the market would pay for all tokens at once. Hold both and ask how fast one converges on the other.

Allocation categories and typical splits

Supply is carved into buckets at launch, and the labels are standard even when the percentages are not. Benchmark data compiled by 8Blocks from Liquifi, Binance Research and Tokenomist’s 2024 annual report (378 tokens) gives these ranges for recent launches.

Two real schedules show the template applied. Hyperliquid’s HYPE, per Tokenomist, allocates 38.9% to future emissions and community rewards, 31% to the November 2024 genesis airdrop, 23.8% to core contributors and 6% to the Hyper Foundation, with no investor bucket because it raised no venture money. Plasma’s XPL allocates 40% to ecosystem and growth, 25% to team, 25% to investors and 10% to the public sale, a conventional venture-backed structure.

The useful question is not the percentage but the behaviour. Keyrock’s study found team unlocks produced average drops near 25% because employees sell on their own timetables, investor unlocks were more controlled because funds hedge and sell over the counter, and ecosystem unlocks averaged a small gain (plus 1.18%). Reading an allocation table means estimating who will sell, how fast and through which venue.

Cliffs, linear vesting and the shape of the curve

Vesting describes when locked tokens become transferable. Three shapes cover almost every schedule.

1. Cliff. Nothing unlocks until a set date, then a block unlocks at once. A one-year cliff on a team allocation is the norm. Plasma’s team and investor tranches each released one third of their 2.5bn tokens on September 25, 2026, exactly one year after mainnet beta.
2. Linear. Tokens unlock continuously or in equal monthly instalments. After its cliff, Plasma’s remaining two thirds of team and investor tokens release pro rata monthly until September 25, 2028; its ecosystem bucket released 800m at launch and the other 3.2bn over 36 monthly instalments of about 88.9m.
3. Cliff plus linear. The combination above, and the most common design for insiders. The cliff is the headline, but the monthly drip that follows usually represents more total supply.

Some schedules are claim-driven rather than calendar-driven. Hyperliquid’s core contributor tokens were locked for a year after the November 2024 genesis and then, from late November 2025, began vesting per person; Tokenomist noted that only about $38m of a $675m tranche scheduled for June 2026 was claimed that month, and that on October 6, 2026 a 3.75m HYPE tranche worth about $336.75m was sold over the counter to an institutional buyer rather than released to market. Where unlocks depend on individual claims, the calendar gives the maximum, not the realised amount.

Cliffs matter more for price than linear releases of the same size because the market cannot absorb them gradually and everyone sees them coming. Keyrock found price suppression typically begins about 30 days before a large unlock, with stabilisation roughly two weeks after, and larger events produced declines up to 2.4 times steeper. A 2% monthly release is a tax; a 20% cliff is an event.

Unlock calendars and the 2025 to 2026 unlock wave

Two providers dominate the unlock calendar business: Tokenomist (formerly TokenUnlocks) and CryptoRank. Both reconstruct schedules from project documentation and on-chain vesting contracts and publish daily, weekly and monthly totals, split between one-time (cliff) and linear unlocks. Tokenomist also tracks 49 market-making firms and their wallets, a point we return to below.

The scale has grown with the 2024 to 2025 launch cohort. Binance Research estimated in May 2024 that roughly $155bn of tokens would unlock between 2024 and 2030; Keyrock’s December 2024 study put the running rate near $600m a week; Tokenomist recorded $97.43bn of emissions in 2025. For the month from early October 2026, Tokenomist data cited by Wu Blockchain on October 3 counted more than $1.913bn of unlocks above a $10m threshold, with one-time events for HYPE, ENA, ZRO, ARB and SUI and linear releases for SOL, TRUMP, WLD, PUMP, AVAX, TAO and NEAR.

A handful of 2026 events shows the range.

The lesson is that share of float is the number to read first. A $337m HYPE unlock against $1.09bn of daily volume is a different event from a $114m DoubleZero unlock that nearly doubles the sellable supply. Dollar headlines flatter large caps and understate the risk in small ones.

Inflation, burns and buybacks: the other side of the ledger

Vesting is not the only way supply changes. Three other mechanisms matter.

Inflation

Proof-of-stake chains mint new tokens to pay validators. Plasma’s documentation sets validator rewards at 5% annual inflation at activation, falling 0.5 percentage points a year to a 3% floor, with locked team and investor tokens ineligible for rewards. Inflation is dilution that never appears on an unlock calendar, which is why the CCS staking guide treats staking yield partly as compensation for dilution rather than pure income.

Burns

Burns permanently destroy tokens, reducing total supply. Fee burns (Ethereum’s EIP-1559 model, which Plasma also adopts) scale with usage. Discretionary burns are one-off decisions: on April 29, 2026 Pump.fun burned every PUMP token it had repurchased, about $370m worth and roughly 36% of circulating supply, after nine months of directing 100% of revenue to buybacks had failed to lift the token above its launch valuation. CoinDesk reported a 6.9% rise in the following 24 hours. The team’s own explanation, a perceived lack of trust in the business’s longevity, is a reminder that supply mechanics cannot substitute for demand.

Buybacks

Buybacks use protocol revenue to purchase tokens on the open market, and they have scaled fast. Cointelegraph Magazine, republished by TechFlow in September 2026, put project buybacks at about $640m from January to early September 2026, up 17% year on year and far above the $366,000 spent in all of 2024, with Hyperliquid and Pump.fun accounting for roughly 90%. The main programmes differ in design.

The sceptical case deserves equal weight. DL News reported in October 2025 that Messari found no clear evidence the market rewards buyback programmes, and that Aave’s pilot was then sitting on a realised loss of about $1.7m. Pump.fun’s April 2026 reversal is the clearest data point: buying back 36% of float did not, on its own, support the price. A buyback is only as durable as the revenue behind it, which is why CCS’s coverage of Multicoin Capital’s Hyperliquid thesis frames the valuation as a race between earnings growth and dilution. Net supply change is a four-line sum: unlocks, plus inflation, minus burns, minus buybacks. For HYPE in 2026 that sum has at times been negative; for XPL before staking activates it is simply the vesting schedule.

The “low float, high FDV” debate

In May 2024 Binance Research published “Low Float and High FDV: How Did We Get Here?”, documenting that early 2024 launches debuted with floats between 6% and 20% and an average market-cap-to-FDV ratio of 12.3%, while their combined FDVs already approached the total for all of 2023. It estimated $155bn of unlocks between 2024 and 2030 and argued roughly $80bn of new demand would be needed just to hold prices flat.

The critique had three strands. Private rounds at rising valuations meant insiders held tokens at a fraction of the listing price, so any unlock was profitable to sell regardless of fundamentals. Thin floats made launch prices easy to support and to manipulate, leaving public buyers as exit liquidity. And FDV became “a meme”, in the words of Binance Research’s own thread: a number nobody expected to be realised but everybody used for ranking.

The response through 2025 and 2026 has been uneven. Some launches raised initial float, shortened vesting, or (as at Hyperliquid) skipped venture rounds so there was no investor bucket to unlock. Others moved the problem: an airdrop-heavy launch can still face a 47% cliff a year later, as DoubleZero did in September 2026. By end-2024 the 2024 cohort’s market-cap-to-FDV ratio had climbed toward 35%, partly because unlocks arrived and partly because prices fell. Both routes close the gap.

How market makers and token loans distort float

Published circulating supply assumes unlocked tokens are held or sold. A third category is quietly large: tokens lent to market makers under option agreements so they can quote two-sided liquidity. Those tokens count as circulating while sitting in inventory, and the agreements can give the firm the right to sell into the market.

Two 2025 episodes made this concrete. In late April 2025 Movement Labs disclosed that an intermediary arrangement had placed roughly 66m MOVE, around 5% of supply, with a market-making entity that sold tens of millions of dollars of it the day after listing; Binance removed the market maker and Movement suspended a co-founder in early May 2025. On April 13, 2025 Mantra’s OM fell roughly 90% in hours, which the team attributed to forced liquidations of collateral on exchanges, while critics pointed to how little supply was in public hands. In both cases the aggregator’s circulating supply said nothing about who controlled the sellable tokens. (These figures are from contemporaneous press coverage; primary filings were not retrieved for this guide.)

The same logic applies to staked and fund-held tokens. Hyperliquid’s Assistance Fund holdings are unlocked but not for sale, so some analysts exclude them from float; CoinGecko’s 222m HYPE circulating figure on October 6, 2026 sits well below the 271m other sources cited in December 2025, reflecting different treatment of fund and staked balances. Tokenomist tracks 49 market-making firms because the published float is a starting point, not an answer. When a float is small, ask what share is on loan.

Where to find the data

All of this data is public, but it lives in five places and none is sufficient alone.

1. Aggregator supply pages. CoinGecko and CoinMarketCap show circulating, total and max supply, market cap and FDV; treat circulating supply as a verified self-report. The CCS prices page surfaces the same figures.
2. Project documentation. The tokenomics page (Plasma’s spells out every bucket, cliff date and monthly release) is the primary source. Vague vesting docs are themselves a finding.
3. Unlock trackers. Tokenomist and CryptoRank turn documentation into calendars with dollar values and share-of-float figures, and flag when realised unlocks diverge from schedule.
4. Block explorers. Etherscan, Solscan and equivalents verify total supply from the contract and let you watch vesting contracts and team wallets. Team tokens moving to exchanges is the earliest sell signal.
5. Governance forums. Buyback and burn programmes (Aave’s, Jupiter’s) are set by proposals whose numbers appear in forum posts before aggregators reflect them.

Airdropped supply deserves its own check, since recipients have no cost basis and tend to sell early; the CCS airdrop guide covers how allocations are sized and claimed, and the DAO governance guide covers how supply concentration affects voting.

How we got here: a timeline

Worked example: modelling Plasma’s XPL dilution over 24 months

How to evaluate a supply schedule: a checklist

• What is the market-cap-to-FDV ratio today? Below 20% means more than four fifths of supply is still to come; HYPE sat near 23% and XPL near 45% in October 2026.
• Is there a max supply, and if not, what is the inflation rate? An uncapped token needs an explicit issuance schedule (Plasma’s 5% falling to 3%) and a burn mechanism to be modelled.
• When is the next cliff and what share of float is it? Share of float, not dollars, predicts impact. DoubleZero’s 47% cliff is the kind to model explicitly.
• Who receives the next 12 months of unlocks? Team-heavy schedules have the worst record in Keyrock’s data, ecosystem-heavy ones the best. Read the recipient split, not the total.
• Is the schedule calendar-driven or claim-driven? Hyperliquid’s per-person claims turned a $675m scheduled tranche in June 2026 into $38m of realised unlocks. The calendar is a ceiling where claims are discretionary.
• What share of float is on loan to market makers or posted as collateral? If the project will not disclose, assume several percent of supply can hit the market without appearing on any calendar.
• Does revenue fund buybacks, and is the programme governed or discretionary? Aave’s and Jupiter’s run on passed proposals; Pump.fun’s changed by team decision in April 2026. Neither guarantees price support.
• Can you verify supply on chain? Contract total supply, vesting-contract balances and team wallets should reconcile with the aggregator’s figure. A gap is a question for the project.
• What is monthly sell pressure as a share of monthly volume? Below 2% to 3% is usually absorbable; above 10% is a liquidity problem whatever the fundamentals. Recompute when volume changes, because volume falls faster than schedules do.

Risks and open questions

The biggest analytical risk is treating published figures as facts. Circulating supply is self-reported, unlock calendars assume every scheduled token is claimed, and volume on many venues is inflated. Each error understates sell pressure relative to real liquidity, the direction that hurts holders.

The second is extrapolating the buyback era. The $640m of 2026 buybacks is concentrated in two protocols whose revenue depends on speculative trading: Hyperliquid’s perpetual futures and Pump.fun’s memecoin launches. A downturn in either shrinks the buyback while unlocks continue, and Pump.fun’s April 2026 reversal shows buyback policy can change faster than vesting policy. The CCS guide to perpetual futures explains why that revenue is cyclical.

The third is regulatory. Insider vesting, market-maker agreements and fee-funded buybacks all touch on securities questions still being allocated between the SEC and CFTC (see the CCS guide to SEC and CFTC jurisdiction). Supply disclosure remains voluntary in most jurisdictions. Open questions include whether aggregators will standardise treatment of staked, lent and treasury tokens, and whether buyback-and-lock (Jupiter) or buyback-and-burn (Pump.fun) proves more durable for holders.

What to watch next

Glossary

Circulating supply

Tokens that can be sold today. Reported by projects, verified by aggregators, and the base for market cap.

Total supply

All tokens in existence on chain, including locked ones, minus verifiably burned tokens.

Max supply

The protocol’s hard cap on tokens that can ever exist. Some tokens have none.

FDV (fully diluted valuation)

Price multiplied by max (or total) supply: what the whole eventual supply would be worth at today’s price.

Float

Informal term for the circulating, freely tradable supply; “low float” means a small share of total supply is trading.

Cliff

A date before which none of an allocation unlocks and at which a block unlocks at once.

Linear vesting

Continuous or monthly release of an allocation in equal portions over a fixed period.

Dilution

The reduction in a holder’s proportional share of a network as new tokens enter circulation.

Inflation

New token issuance, typically to validators or stakers, that adds to supply outside the vesting schedule.

Burn

Permanent destruction of tokens, reducing total supply; can be automatic (fee burns) or discretionary.

Buyback

A protocol using revenue to purchase its own token on the market; purchased tokens may be burned, locked or held.

Market-maker loan

Tokens lent by a project to a trading firm to provide exchange liquidity, usually counted as circulating even while held in inventory.

Why it matters

Token supply schedules are the closest thing crypto has to a published capital structure, and they are more transparent than most equity cap tables. The problem is not a lack of information but a failure to read it: market cap gets quoted, FDV gets dismissed as a meme, and the release table in the docs goes unread until the cliff arrives. The 2025 to 2026 unlock wave, with nearly $100bn of emissions in a single year, has made the cost of that habit visible across hundreds of tokens.

Revenue-funded buybacks at Hyperliquid, Aave, Jupiter and Pump.fun change the arithmetic without changing the discipline. A token’s net supply path is unlocks plus inflation minus burns minus buybacks, and the last two terms depend on business performance that can reverse. The useful habit is the one the worked example shows: rebuild the schedule from primary documents, convert it to share of float and share of volume, state your selling assumptions, and update when the inputs change.

Sources

1. Plasma: XPL Tokenomics documentation, accessed October 6, 2026
2. CoinGecko: Methodology (circulating supply and market cap), accessed October 6, 2026
3. CoinGecko: Hyperliquid (HYPE) price, supply and volume, October 6, 2026
4. CoinGecko: Plasma (XPL) price, supply and volume, October 6, 2026
5. Tokenomist: Hyperliquid (HYPE) allocation and unlock schedule, October 6, 2026
6. Tokenomist: Plasma (XPL) allocation and unlock schedule, October 6, 2026
7. Tokenomist Research: Weekly unlock digests and 2025 emissions data, October 5, 2026
8. Binance Research: Low Float and High FDV: How Did We Get Here?, May 2024
9. BeInCrypto: 90% of Token Unlocks Drive Prices Down: Keyrock Research, December 6, 2024
10. Streamflow: Do Token Unlocks Cause Price Drops? What the Data Shows, 2026
11. CoinDesk: Pump.fun burns 36% of PUMP supply, locks 50% of revenue into ongoing buybacks, April 29, 2026
12. Cointelegraph Magazine via TechFlow and KuCoin News: Crypto token buybacks surge to $640M in 2026, September 7, 2026
13. DL News: Why Aave eyes permanent $50m buyback programme despite scepticism, October 23, 2025
14. Crypto Briefing: Aave founder Stani Kulechov considers token burn for Aavenomics 3.0, September 29, 2026
15. Crypto Briefing: Jupiter announces JUP buyback policy starting Monday, February 13, 2025
16. KuCoin News: Jupiter Litterbox Trust adds 177,570 JUP, total holdings reach $31.4M, June 28, 2026
17. KuCoin News: Hyperliquid Assistance Fund crossed $1B in HYPE buybacks, March 29, 2026
18. Netcoins: Hyperliquid’s first major HYPE unlock begins, December 1, 2025
19. BeInCrypto: 3 Token Unlocks to Watch in the Fourth Week of September 2026, September 21, 2026
20. Hokanews: Crypto tokens worth more than $1.913 billion face major unlocks over next month, October 3, 2026
21. 8Blocks: Token Vesting and Allocation Benchmarks (2026), 2026
22. Jupiter Research forum: Increase the Litterbox buyback allocation from 50% to 70%, 2026

Disclosure: This guide is for education only and is not investment, legal or tax advice.