Token Unlock Schedules: Modeling Supply Overhang
The 2026 unlock calendar has been unusually concentrated: Celestia released roughly 17.2 percent of market cap in April, Wormhole 6 percent two days later, Pyth 14.2 percent in May, and ZKsync 17.4 percent in June. Studies put roughly 90 percent of unlocks as creating negative price pressure, with the damage arriving before and during the event rather than after — price reaction typically begins around 30 days ahead as participants position for expected supply. That timing is the first clue that the unlock date is not the thing being measured. The schedule is public years in advance; what is not public is how much of the released supply actually reaches an exchange, and when. Modeling supply overhang means modeling behavior, and the model breaks in specific, detectable ways before the price does.
Defining the Metric Precisely
Two distinct quantities get conflated. The unlock is a flow: a scheduled quantity of tokens becoming transferable on a date, fixed by contract and knowable in advance. The overhang is a stock: the total supply already unlocked and sitting in holder wallets that is likely to be sold but has not been, typically because holders are waiting for a better price. Overhang accumulates when unlocks are not absorbed, and it exerts sustained pressure independent of any calendar date because sophisticated participants price expected sell flow in advance. A project can have no unlock for six months and a severe overhang problem the entire time. Modeling the calendar without modeling the residue measures the wrong thing.
Measurement Method and the Primitives
Four inputs build the model. The vesting schedule itself, from project documentation or an aggregator: allocation by category, cliff dates, and unlock cadence, with the structural distinction between a cliff — a lump sum, often 20 to 50 percent of an allocation on a single date — and linear vesting, which drips continuously. Holder cost basis by tranche, because the discount at which an allocation was acquired predicts propensity to sell better than the size of the allocation. Absorption capacity, the parameter that matters most and is estimated rather than observed: daily traded volume and order book depth against the unlock size. And post-unlock destination tracking — whether released tokens move toward exchange deposit addresses, remain dormant, or route into staking.
The denominator is where models go wrong. Expressing an unlock as a percentage of circulating supply is conventional and misleading; the economically relevant ratio is unlock size against daily traded volume. A 1 percent unlock into a deep book is noise, while the same percentage into thin liquidity is the entire market for days.
Distortions and What Breaks First
The model breaks before the price does, and the fragile assumption is absorption. The naive version maps unlocked tokens to sell pressure one-for-one, which is wrong in both directions: recipients frequently hold, and holders who unlocked months earlier frequently sell into an unrelated rally. Detection is tractable — compare the scheduled unlock quantity to observed exchange inflows in the following days, and track the persistent gap. A model whose predicted impact repeatedly exceeds realized impact is over-weighting the calendar; one that repeatedly under-predicts is missing accumulated overhang from prior periods.
Three further distortions matter. The low-float, high-FDV structure lists a token with a small fraction of supply circulating, making the price easy to support at launch while years of scheduled releases stand between that price and a market cap that honestly reflects supply. Anticipation contaminates measurement: if impact is priced 15 to 30 days ahead, a study measuring returns from the unlock date will conclude unlocks do not matter. And schedules are not immutable — in early 2026 some teams delayed unlocks in response to pricing pressure, which helps holders and destroys the predictive value of a schedule treated as fixed.
Other Metrics That Matter
The FDV-to-circulating-market-cap ratio locates the structural dilution ahead, with multiples above roughly five signaling that most of the valuation is still locked. Unlock size as a percentage of daily volume is the absorption estimate. Monthly release rate against circulating supply captures linear-vesting drag that no single date reveals — repeated releases totaling more than about 5 percent of circulating supply per month is a standing headwind. Exchange inflow from known vesting addresses converts the model's central assumption into an observation.
Healthy Trends and Real Usage
Healthy patterns: circulating share rising toward the majority of supply, unlocks absorbed without persistent post-event exchange inflow, demand growth outpacing release rate, and a schedule the team honors. Unhealthy patterns: cliff concentration, a widening overhang where released tokens sit unsold, monthly releases exceeding demand growth, and allocation structures institutional diligence now flags directly — team cliffs shorter than a year, a single round holding more than 30 percent of supply, or more than 15 percent unlocking at launch. For an allocator the usage is position sizing and timing rather than direction: size against the absorption estimate, not the headline percentage, and treat the 30-day pre-unlock window as the execution risk it is. The honest constraint is the one every unlock analysis should carry: supply is half of a price. A token with a punishing schedule and genuine demand growth can rise through every release date, and the strongest projects of every cycle have done exactly that. The constructive development is that this data is now public and standardized — cliff dates, allocation categories, and historical unlock impact are queryable before an allocation rather than discovered after one, which makes supply overhang the rare crypto risk that is entirely knowable in advance.
For informational purposes only. Not an offer to buy or sell any security. Available only to accredited investors who meet regulatory requirements.
