The Zcash rich list is missing 29% of all Zcash

01 · The list you can look up

And the part of it that was quietly deleted

Type "zcash rich list" into a search box and the results look like every other rich list in crypto. A rank, an address, a balance, a share of supply. The tables disagree with each other by a few percent, which readers put down to stale data, and the top entry usually holds something close to half a million coins.

Zcash is a cryptocurrency launched in 2016 that copied Bitcoin's design and added one thing to it: you can move coins without the amount, the sender or the receiver appearing on the chain. Coins held that way sit in what Zcash calls a shielded pool. There is no address to look up and no balance to read. That is the whole point of the coin.

So a Zcash rich list ranks the coins whose owners chose not to hide them. Our blockchain data lake holds every Zcash block since the first one, so we rebuilt the list from the raw chain, pinned to a single block height. It sees about seven coins in every ten that Zcash has ever issued. The rest is not an indexing gap or a stale cache. It is the coin working as designed, and most of that gap opened up after January 2025.

What is left is worth reading anyway, because the top of it is not what a rich list normally holds.

02 · Where a coin goes when it disappears

Transparent addresses, shielded pools, and the door between them

Zcash runs two ledgers at once. The transparent side works like Bitcoin: coins sit at addresses beginning with t, and anyone can add up what each one holds. The shielded side holds coins in a pool. A shielded balance is a cryptographic commitment, not an entry against a name, and even the amounts are encrypted.

The two sides meet at a door. Send coins from a t-address into the pool and the transaction shows money going in and nothing coming out. Send them back and coins appear at a t-address with no visible source. Almost every coin Zcash has mined arrived at a transparent address first, because that is how block rewards are normally paid. Between August 2020 and October 2022 there was an exception, and miners took the larger part of each reward straight into the pool without it ever having an address. Count both kinds of issuance and the arithmetic closes: coins created, minus coins sitting at addresses, equals coins in the pools.

That subtraction is how anyone measures the shielded pool, and it is the only way. We ran it three separate ways against three different tables and got the same number to the cent.

03 · The year the list stopped working

A fifth of the supply went dark in twelve months

For eight years the pools filled slowly. Across the whole of that stretch they took in less than 2 million ZEC, and the visible share of supply drifted from about 99% down to 88%. Anyone building a rich list back then was missing roughly a tenth of the coins, which is close enough to shrug at.

Then came 2025. In that one year more coins went into the pools than in the previous nine put together, and October alone carried over a million of them, the heaviest month of shielding in Zcash's history. By New Year the list had gone from covering nine coins in ten to covering seven.

The tide turns both ways. June 2026 brought the largest month of unshielding Zcash has recorded, four times deeper than any month before it, and August pushed most of that back in. Anyone quoting a holder ranking without a date attached is quoting a figure that swung by a fifth of the supply inside a year. For the reverse of this story, our audit of how ZEC trades on Hyperliquid covers the paper coins that never touch the chain at all.

04 · What is actually on the list

846,655 addresses, and 19 that matter

Strip out the empty wallets and about 850,000 still hold a balance, though only a tenth of those hold a whole coin. Nineteen hold more than six figures. How that stacks up is normal for a chain of this age, and it is the kind of holder ranking we compute for any chain: the largest ten wallets hold roughly a quarter of what is visible, and the largest hundred hold about three fifths.

Read down the top of the list and the wallets sort into a few habits. Some are cold. One of them took a six-figure payment in November 2025 and has never spent a coin since. Others are exchange plumbing, swallowing hundreds of deposits a week and pushing nearly as much straight back out. Then there is the very top, which behaves like neither.

Here is the top 25, every balance read straight off the chain at block 3,477,382. Addresses link out to a block explorer.

A ranking of addresses flatters whoever splits their coins across the most wallets, so the list above is a starting point rather than an answer. Ranks 1, 4 and 16 turn out to be the same holder, and proving that takes the same technique we used to follow stolen coins in the Coldcard wallet theft: when two addresses sign the same transaction, one key controls both. Bitquery's address APIs expose exactly that join.

05 · The biggest signer also signs for Zcash

47 addresses, 1.07 million ZEC, and three consensus payslips

Follow the shared signatures out from the top wallet and the cluster closes at 47 addresses holding 1.07 million ZEC between them, close to a tenth of every visible coin. The link is not a guess. A hub wallet, t3RuxpWg…GZk, signs with the top wallet two dozen times, and signs with two of the others in spends that pull in two more members at once.

Three wallets in that cluster stand out. They are paid out of the coinbase, the freshly minted coins in each block, and they are paid in blocks the cluster plainly did not mine. We took a window of several thousand blocks running up to the snapshot and found a cluster wallet paid in every single one of them. No miner wins every block. Only the chain itself can pay that reliably, and Zcash does, because a slice of each block reward is written into the consensus rules.

The amounts confirm which slice. Two of the wallets were paid in a ratio of exactly eight to five between the Canopy upgrade and the 2024 halving, which matches the split in ZIP-1014, which set the dev fund at a fifth of the block reward and divided it between the Electric Coin Company, the Zcash Foundation and Major Grants. The third wallet took over at the halving and has been paid the same fixed amount in every block since, which works out to exactly the share ZIP-1015 routes to the Financial Privacy Foundation for Zcash Community Grants. The rest of the dev fund now goes to a protocol lockbox, which has no address at all, and that is why the cluster's share of new coins fell at the halving.

Two things follow. The dev fund has already shielded most of what it was paid, so the coins parked in the cluster today came in from somewhere else and are treasury money held under the same keys. And the largest visible holder on a privacy coin uses the privacy. To see the same sum done on a chain where hiding is a bolt-on rather than a feature, our Tornado Cash audit counts seven years of it on Ethereum.

06 · The wallet that appeared on one afternoon

Six addresses, $872 million, and a transaction that has happened five times

Four of the top twenty first moved coins on the same day, 2 March 2026, within four hours of each other. With two smaller siblings they now hold close to 700,000 ZEC. Six months ago none of them existed.

They are one holder, and the proof is a habit rather than a signature. Filter every Zcash spend ever made for one very specific shape: a single input worth exactly one coin, two outputs, and an odd fee that no standard wallet rounds to. Across the entire history of the chain there are five matches. All five happened on 2 March 2026, and four of them sit in the same block, paying the same destination from four separate wallets.

That is one person testing a wallet four times over in a single block. Ninety minutes earlier, a block carrying only a handful of spends had given birth to five brand new wallets, four of which went on to hold real money and two of which now hold six figures.

The handover is visible too. An older wallet, t1Z2JqyY…mcj, had run the same one-coin test over and over through the winter, paid for the new wallet's very first receipt, then fed it six figures of ZEC across hundreds of spends and reached a zero balance three weeks later. It had shared keys with thousands of sibling wallets since 2023. The new wallet shares keys with nothing, and never merges coins across its own addresses, which is why the usual grouping method finds nothing here and the fee habit finds everything.

Not one of these six wallets has ever touched the shielded pool, across more than 12,000 spends. Whoever runs this is on Zcash for a reason other than privacy, and behaves like an exchange: close to 10,000 separate wallets have paid the main one. Tracing that kind of flow across a chain is what Coinpath and our investigation services are built for.

07 · What a rich list cannot tell you

Three traps, and the honest limits of the answer

Three things in this data will produce a wrong answer if you trust the obvious query, and all three bit us first.

The first is the coinbase input. Zcash's records carry an input row for every coinbase, one per block ever mined, standing in for a spend that never happened. Subtract those from wallet balances and the visible supply comes out negative, which is impossible, and which lands on exactly the negative of the true shielded pool. The second trap was our own summary table. It double-counted receipts and produced a top twenty that was mostly fiction, crediting one wallet with twice what it had ever been paid, when the chain shows a single receipt that was spent in full and a balance of nothing.

The third is the one that moved this article's headline. For two years from August 2020, miners could take their reward straight into the shielded pool, and 60,163 blocks did. Zcash's own mined-value field counts only the transparent part of those rewards, so a count built on it understates total supply by 151,917 ZEC and understates the pool by the same amount. Reading the shielded side of the coinbase back in moves the hidden share from 28% to 29%.

Both were caught by the same check. Add up every positive balance and the total has to equal mined supply minus the pool. Ours agrees to the cent, with exactly one wallet negative by a hundredth of a coin. We then read the ten largest balances back from a public block explorer and matched every one of them down to the last decimal place.

A Zcash rich list ranks the coins whose owners did not care to hide them, on a chain built for hiding. Every version of that table has the same hole in it, and the hole grew by a fifth of the supply in a single year. What sits at the top of what is left is the fund that pays for the software, and it has already shielded most of what it was paid. Where those coins went next, the chain does not say.

You can rebuild all of this. The same index sits behind Moneyflow for following coins between wallets and the address labels API where a chain has labels to give. For the same rebuild-from-raw-blocks approach applied to the oldest coins in crypto, see our count of what Satoshi Nakamoto actually holds.