Total ZEC held in transparent addresses, broken down by how long each coin has been unspent. Each transparent output (UTXO) is bucketed by age since creation: <30 days (fresh), 30-180 days, 180 days to 1 year, 1-3 years, and 3+ years (dormant).
The Y-axis is ZEC, not UTXO count. The 3y+ cohort likely includes lost coins and long-term holders who never shielded. Built from a full UTXO set index, replayed block-by-block from genesis.
This is the transparent side of Zcash only. Every coin in a shielded pool is absent from it — not omitted by choice but genuinely uncountable, since a shielded note carries no publicly readable age. So this is not a picture of how long Zcash holders have held. It is a picture of how long the transparent minority has sat still.
A cohort shrinking does not mean coins were sold. When transparent ZEC is shielded it leaves this chart entirely, and from here that is indistinguishable from a spend. Check the shielding rate before reading a drop as distribution — on this chain the two look identical and mean opposite things.
The Y-axis is ZEC, not outputs. One dormant exchange-era output can outweigh thousands of small fresh ones, so the bands track value at rest rather than users at rest.
The 3-year band is not all conviction. It includes coins whose keys are gone, which never move again and so ratchet upward forever. Read it as an upper bound on long-term holding. For the same data as a decay curve see survival; for the single-number version, mean coin age.
Where this comes from: a full set of unspent transaction outputs (UTXOs) replayed block by block from the very first block, so an output’s age is its real creation height rather than an inference from balances.