Hook: An Eight-Decimal Tell
The number 85,705.32563297 is the tell.
Nobody negotiates a position to eight decimal places. That value is not a negotiated allocation. It is a wallet balance swept into a fund and reported down to the last satoshi the ledger would surrender. It appears in the disclosure of an in-kind transaction in which DCG International Investments contributed that quantity of ZEC to the Grayscale Zcash ETF in exchange for roughly $100 million of fund shares.
Two other numbers sit beside it.
One billion dollars: where the fund's net asset value landed in the weeks after it began trading on NYSE Arca on August 25.
Three hundred six million, one hundred twenty thousand dollars: cumulative net inflows, described in the same material as less than 30% of total assets.
Line those figures up and the headline disassembles itself. A fund that reached a billion dollars on $306 million of net inflows did not get there because institutions arrived in force. Something else did the work. The composition of that something else is the entire risk profile, and it did not make the headline.
I have spent most of my working life taking the round number off the billboard and reconciling it against the ledger underneath. It is almost never flattering. This is not the exception.
Context: What Is Actually Being Wrapped
Start with the quality of the numbers
Before any analysis, a disclosure about the disclosure. The material this rests on names no originating newsroom. Several data points carry no source at all. Others trace back to Grayscale or to DCG, which is the sponsor and the parent, respectively. That does not make the numbers false. It makes them claims rather than measurements, and the difference should change how much weight you assign to each.
The structural facts are the safest ones. A NYSE Arca listing on August 25, a trust renamed into an ETF wrapper, a 1-for-3 split with a September 28 record date and a September 30 effective date. Those are the kinds of details that surface in filings and exchange notices. Assume they are roughly correct.
The interpretive numbers are softer. A billion dollars of net asset value. $306.12 million of cumulative net inflows. Roughly $100 million contributed in kind. Those figures are assembled by, or adjacent to, the party that benefits from their framing. Everything below treats them accordingly, and where a number cannot be reconciled, I say so rather than smoothing it.
What Zcash is, technically
Zcash launched in October 2016 as a fork of the Bitcoin Core codebase with a single, enormous substitution. It kept the UTXO accounting model, the 21 million supply cap, the halving schedule, and proof-of-work. It replaced the fully transparent ledger with a dual-state one.
Every ZEC exists in one of two pools. In the transparent pool, balances, senders, receivers, and amounts are visible exactly as they are on Bitcoin. In the shielded pool, a transaction is proven valid by a zero-knowledge succinct non-interactive argument of knowledge, a zk-SNARK, without revealing sender, receiver, or amount. The same coin can move between pools. Privacy is a mode, not a property of the asset.
The original Sprout construction required a multi-party trusted setup ceremony, a ceremony whose transcript had to be destroyed for the system to be sound. Sapling in 2018 cut proving costs by orders of magnitude and made shielded transactions practical on commodity hardware, though it retained a setup step. Orchard, delivered with the Halo 2 proving system in 2022, removed the need for a trusted setup entirely and introduced a new shielded pool with a different note commitment tree.
Three shielded pools have now existed over the network's life, each with its own anonymity set, and each migration has reset the size of the crowd a coin hides in. That detail matters later.
Proof-of-work ran on Equihash, chosen for memory hardness and for some resistance to the ASIC arms race that Bitcoin's SHA-256 invited, though that resistance eroded. Block rewards halve roughly every four years, most recently in November 2024. And for most of the network's history, a defined share of issuance was earmarked for development entities: the Electric Coin Company, the Zcash Foundation, and later grant committees. Those allocations were not metaphorical. They landed in addresses, and those addresses are public.
Why the dual ledger is the whole investment case
Zcash is not a privacy coin in the way Monero is a privacy coin. It is a coin with a privacy mode. That is a distinction most retail holders treat as cosmetic and most institutions treat as decisive.
A fund that holds an asset on behalf of regulated investors must be able to prove what it holds, to whom it belongs, and where it came from. Monero's design makes that structurally impossible without trust in the custodian. Zcash's design makes it possible without touching the protocol at all. The custodian simply holds in the transparent pool.
Understand that and the reason this asset, and not its larger privacy competitor, ended up inside a US-listed wrapper stops being a mystery. It also becomes the source of the deepest tension in the product, which I come to later.
What Grayscale built
The Grayscale Zcash Trust was established in 2017 as a private placement, quoted over the counter, and later renamed and converted into an exchange-traded fund. The wrapper changed. The underlying did not. There has been no Zcash protocol upgrade, no consensus change, no cryptographic milestone attached to this event. It is a distribution event, not an engineering one. Anyone reading a billion-dollar headline as evidence of technical progress is reading the wrong document.
The split mechanics, stated once and plainly
On September 30 the fund executes a 1-for-3 share split. Record date September 28. NAV per share falls to roughly one third. Share count triples. The dollar value of any holder's position is identical before and after, minus friction.
That is the entire mechanical content of the event. I will return to what it is for.
The regulatory perimeter around privacy assets
The venue set for privacy assets has been contracting for years, not expanding. Binance removed Monero from its spot markets in early 2024. European frameworks have been moving toward restricting credit institutions from holding or transmitting anonymizing assets, with implementation timelines that push into the late 2020s. Japanese and South Korean venues have been the most restrictive of any major markets for the longest. The net effect is that privacy assets trade on a shrinking list of exchanges, and shrinking exchange lists mean thinner books and weaker price discovery.
Against that backdrop, a US-listed wrapper is the single most valuable piece of real estate a privacy asset can occupy. It is also the only piece of real estate that comes with conditions.
Framing for a bear market
In a market like this one, the question is not what a product can do for you. It is what can be taken from you. Readers holding ZEC, holding the fund, or considering either want one thing: a straight answer about whether the asset behind the ticker is as safe as the ticker suggests.
This is not a price call. I do not make those. It is a decomposition of what a one-billion-dollar figure is actually made of, and an enumeration of the ways that composition can hurt a holder who mistakes the headline for the mechanism.
Core: The Decomposition
One: Assets under management is not a demand metric
Start with the arithmetic nobody in the promotional material wants to run.
AUM equals shares outstanding multiplied by net asset value per share. Net inflows equal creations minus redemptions, measured in dollars at the NAV on the date each creation or redemption settled. Both numbers are denominated in dollars. Neither measures the same thing.
If the underlying asset appreciates, AUM rises with zero new money. If the underlying asset falls, AUM falls even while money is pouring in. Comparing the two figures over a window in which the underlying nearly doubled is not a comparison. It is a category error in one direction and, depending on how it is presented, a concealment in the other.
So run it. The fund crossed $500 million within two weeks of the August 25 listing. It reached a billion dollars by late September. Over that stretch, ZEC's price nearly doubled.
Now do the multiplication. A base of roughly half a billion dollars of assets, held while the underlying appreciates approximately 100%, becomes roughly a billion dollars without a single incremental dollar of external capital. The $306.12 million of cumulative net inflows is therefore not merely consistent with a billion-dollar fund. It is arithmetically unremarkable. It is what you would expect.
This matters because the criticism and the praise are both being aimed at the wrong target. The framing of less than 30% net inflows is a fair attack on the headline, which implies broad institutional demand. It is not a fair attack on the fund, which did exactly what a fund tracking an asset that doubled should do. Precision about which claim fails is the difference between analysis and commentary.
Then the number gets worse, because it is contaminated.
The roughly $100 million in-kind contribution from DCG is a creation. In-kind creations are creations. They are booked as net inflows by every standard convention in the fund industry. They are also not external capital. They are the parent company moving an existing asset from one balance sheet line to another and receiving fund shares in return.
Strip it out. Roughly $206 million of the $306.12 million came from outside the corporate family. Against a billion-dollar NAV, that is approximately 20%. Against the roughly $500 million of assets that existed before the appreciation ran, it is closer to 40%, which is a materially healthier number and a materially smaller one than the headline.
Three different figures, three different stories, all true. One billion in assets. Three hundred six million in net inflows. Two hundred six million from outside the family. The editorial choice of which to lead with determines what a reader concludes, and the source material led with the largest.
There is one more wrinkle worth naming before moving on. Net inflows are struck at the NAV on the date of each creation, which means a fund that received most of its money early, before an asset doubled, shows the same dollar inflow figure as a fund that received it late. The number tells you what was paid, not what it is worth. Anyone using cumulative net inflows as a proxy for current demand is measuring history against a present-tense balance sheet.
Two: The in-kind swap and the implied unit price
Now to the eight-decimal number.
Divide $100 million by 85,705.32563297 ZEC. The implied unit valuation is approximately $1,166.79 per coin.
That figure is either the most important number in the entire disclosure or the least reliable one, and the material does not tell you which. Three hypotheses fit.
First, the valuation date differs from the reference date. In-kind contributions to a fund are typically priced at a valuation struck on a specific date, sometimes a valuation date that precedes the transfer settlement by days or weeks. If ZEC moved sharply between the pricing date and the public reference date, the implied multiple will look wrong without anything being wrong.
Second, the ZEC quantity is partial, or the $100 million includes consideration beyond the coins. If the disclosed coin count represents only the first tranche of a larger contribution, or if the dollar figure bundles in a fee waiver, a seed commitment, or a backstop, the division is meaningless.
Third, the swap was struck at an agreed valuation that is not the market price. This is the one that matters, and it is the one nobody outside the filings can rule out.
Here is why it matters, and it is not a technicality.
The price at which an affiliate contributes assets to a fund determines how many shares of that fund the affiliate receives. If the contribution is valued above the prevailing market price of the underlying, the affiliate receives more shares than the market value of what it handed over. The difference is transferred from existing holders to the affiliate, immediately and silently. If the contribution is valued below market, the affiliate is subsidizing the fund and its holders get a gift.
In a cash creation, this problem does not exist, because a dollar is a dollar on every date and in every venue. In an in-kind creation of a volatile, thinly traded asset conducted by the fund's own parent, the problem exists on every single settlement.
I spent four months inside the 0x v2 exchange protocol in 2018, manually auditing the maker fee calculation logic while the team was preparing for mainnet. I found an integer overflow that could have drained liquidity pools, filed seven issues, and the launch slipped two months while the patch went in. The lesson was not that the code was bad. The lesson was that an eight-decimal discrepancy in an accounting path is never neutral. It is either a rounding artifact or a redistribution of value.
The disclosure here does not tell us which. A fund that publishes a net asset value every single trading day has not published the valuation date it used to price an affiliate contribution representing roughly a tenth of its float.
Code does not lie; people do. And the code here may actually be able to settle the question. DCG's historical ZEC holdings, if held transparently, are traceable on-chain to the block. Anyone with a node and patience can reconstruct when the coins moved, at what market print, and what the implied discount or premium to spot was on that date.
Unless the coins moved through a shielded pool. In which case the audit stops at the pool boundary, and the same cryptographic feature that justifies the product's existence becomes the reason its most consequential transaction cannot be independently priced.
That is a test of the privacy feature as a disclosure instrument, and the disclosure did not run the test.
Three: The split does not create value, and the timing is not random
On September 30, the fund splits one share into three. Record date September 28. NAV per share drops to approximately one third. Share count triples. Every holder's dollar position is unchanged.
There is no version of this transaction in which a holder gains value. There is no version in which the fund gains assets. There is no version in which the underlying ZEC position changes by a single coin. It is a denominator operation performed on a spreadsheet.
I have reviewed enough fund restructurings to know that splits are done for a small, boring set of reasons, and none of them is value creation.
The first is per-share price optics. A share trading at a low absolute number is treated as accessible by retail brokerage interfaces and by investors who read price level as a proxy for affordability. This is irrational and it is extremely well documented.
The second is granularity. Smaller notional per share means finer position sizing for smaller accounts.
The third, and the most substantive, is derivatives. Options exchanges quote strikes in fixed increments. A share price that is too high relative to the strike grid produces a coarse and illiquid option chain. A share price brought into a more quotable range makes a listed option plausible. If ZCSH options ever appear, the fund stops being an exposure product and becomes a market-making product, which is a genuinely different and more durable business.
The fourth is signaling to distribution channels. A split is what a sponsor does when it wants the flow to continue after the founding story has been told and the AUM milestone has been printed.
Notice the sequencing. The fund crosses a billion dollars. One month later, it splits three for one. The milestone gets the product into screens. The split gets it into the hands of people who buy on the screens.
None of this is improper. All of it is marketing, and marketing is not a fundamental.
The reason to dwell on it is that splits are one of the few corporate actions that reliably generate a predictable class of misreading. Holders open their accounts, see three times the share count, and experience a moment of apparent gain. The moment passes in a second for anyone who checks the total value column. For anyone who does not, it is the seed of a story about a fund that is growing.
High yield is a warning, not a welcome. In fund flow terms, the analogue is breakneck AUM growth: it is a warning, not a welcome. Growth that comes from the price of the underlying is beta with a marketing department attached. Growth that comes from an affiliate's balance sheet is a related-party transaction. Growth that comes from external cash is demand. Only the third one tells you anything about whether anybody actually wants this.
Four: The custody contradiction
This is the structural problem, and it is larger than the flow arithmetic.
A US-listed fund must be auditable. Auditing a ZEC position held for shareholders can be done two ways. The custodian can hold in the transparent pool, in which case the entire position is verifiable by anyone with a block explorer and a known address, and the auditor's job is trivial. Or the custodian can hold in the shielded pool and provide an attestation, in which case the verification rests entirely on the custodian's honesty and on the auditor's willingness to accept a cryptographic claim they cannot independently check.
The industry spent 2022 learning what attestation without independent verification is worth. The answer was, in several well-documented cases, nothing at all.
So the market will demand the transparent option, and the custodian will supply it, because it is the only defensible position.
Follow that to its conclusion. A fund whose entire marketing premise is exposure to a privacy asset must hold that asset in its least private state. The wrapper's compliance requirement selects for the exact property the product is supposedly about not needing.
That would be an irony if it were merely an irony. It is worse than that, for three reasons.
First, concentration. A billion dollars of ZEC held in transparent addresses controlled by a single known corporate entity is a large, identified, continuously observable position on a ledger whose selling proposition is that positions are not observable. Every future movement of size is visible to the market in real time, which means the fund's creation and redemption activity leaks ahead of the tape in a way that does not happen with a cash-settled exposure.
Second, degradation of the anonymity set. Shielded pools only work if there are enough coins in them. A privacy guarantee is a network effect. Zcash has already migrated its shielded state across three pools over nine years, each migration resetting the crowd a coin hides in. If institutional adoption pulls a growing share of supply into transparent custody, the share of supply in shielded pools falls, the anonymity set shrinks, and the privacy property that justified the asset's valuation erodes as a direct function of the asset's institutional success. Adoption and utility are in tension here. That tension is not disclosed in any marketing document I have seen.
Third, and this is the practical one, a known transparent address is a target in a way an unknown one is not. Whales get found. Custodians with public balances get front-run by anyone with a mempool watcher. The security posture of the fund is now partly a function of how visible the fund chooses to be.
I did this analysis once before, in 2024, on the spot Bitcoin ETF custody arrangements. I found conflicts of interest inside the segregated custody structures of three large issuers, and I caught considerable backlash from people who read the report as an attack on institutional adoption. It was not. It was an observation that the more layers you insert between an asset and its holder, the more places there are for the interests of the intermediary and the interests of the holder to diverge without anyone technically doing anything wrong.
The Zcash case is simpler and considerably worse. The conflict is not between two custodians. It is between the product's thesis and the product's own auditability. They cannot both be satisfied at the same time, and the fund has implicitly chosen.
Five: NAV, the price index, and the liquidity that is not there
The fund's net asset value is struck from a price index. That index is assembled from the prices available on a set of venues. For a privacy asset, the venue set is constrained by the delistings and restrictions described earlier, which means fewer sources, thinner books, and a reference price that is derived from a smaller and more fragile universe than the reference price for any other major digital asset.
Now consider the mechanism that is supposed to keep the fund's market price tethered to its NAV. It is creation and redemption arbitrage. An authorized participant observes a premium, buys the underlying, delivers it to the fund, receives shares, sells them, and pockets the difference, or runs the sequence in reverse against a discount. The activity compresses the spread and is the reason an ETF can be trusted to track its index.
That arbitrage requires a market maker to buy or sell the underlying in size, repeatedly, against a price index. The question is whether the capacity exists.
A billion-dollar notional cannot be hedged on order books that clear a few million dollars a day without the hedging itself moving the print. If the depth is not there, the tether is a hypothesis rather than a mechanism, and the fund's market price becomes free to wander from NAV during exactly the moments when holders care most.
In 2020, during the DeFi summer, I built the model that became a fifteen-page risk assessment titled The Illusion of Arbitrage. It looked at the interaction between staked ETH and Compound lending markets and concluded that the implied yield spread was unsustainable because it depended on a price reference that behaved correctly only while liquidity was adequate and stopped behaving correctly at precisely the moment it was needed. The finding was not that the strategy was fraudulent. It was that the people collecting the spread were being paid to hold tail risk they had not priced.
The same question applies here, pointed at the oracle layer. Which venues feed the index. What aggregation method is used. Whether the method is a median, a volume-weighted mean, or something with a fallback path that a single venue outage can hijack. What happens during a de-risk event when a major venue halts withdrawals and its price print freezes while everyone else's moves.
Oracle feed latency and composition are the least discussed and most consequential structural features of every wrapped asset product, and the DeFi industry's habit of solving a decentralization problem with a small set of permissioned node operators has not made anything better. It has made the attack surface more concentrated and the accountability more diffuse at the same time.
For a holder of ZCSH, the practical implication is narrow and actionable. The informative number is not the AUM. It is the premium or discount to NAV. It is the only figure a sponsor cannot inflate through price appreciation or in-kind window dressing, because it is set by other people's willingness to arbitrage the spread. If it stays wide, the mechanism is not working. If it stays tight through a drawdown, the mechanism is real and the product deserves more credit than I am giving it here.
Watch that number. Nothing else in the disclosure tells you as much.
Six: Follow the addresses, not the manifesto
Zcash's issuance historically diverted a defined share of block rewards to identifiable development entities. Those addresses are public. Anyone with a node can compute the cumulative inflows to each over the network's life, and anyone can see when they move. This is not an accusation of impropriety. It is a statement about what decentralization means in practice: a governance narrative sitting on top of a ledger that records exactly who got paid, when, and how much.
Apply the same lens to the fund. ZCSH holders have no on-chain governance rights. They do not vote on the split. They do not vote on the custody arrangement. They do not vote on the management fee. They do not vote on whether the sponsor launches eleven more single-asset trusts in the same format. The split was announced as a fact, not proposed as a decision, because the trust structure exists precisely to convert an opaque governance arrangement into a tradable share that does not carry a vote.
That is the actual product. Not privacy. Not decentralization. A share class.
I have held this position for years and it has cost me readers. In this industry, the organizations that describe themselves as decentralized are, with rare exceptions, the ones with the most concentrated control and the least transparent accountability, because the vocabulary of decentralization functions as a compliance shield: it lets a foundation holding a treasury argue that it is not an issuer, and it lets a parent company contributing assets to its own subsidiary's fund argue that it is not a related-party transaction with governance implications.
Forensics does not care about the label on the entity. Follow the addresses. Audit the flow. If the flow says the parent seeded roughly a tenth of the fund's float in a transaction whose pricing date is undisclosed, then that is what happened, regardless of how the press release describes the arrangement.
Contrarian: Where the Bulls Are Right
I have spent this entire article dismantling the demand narrative. Now the part that most bearish analyses skip, which is where the bulls are correct and I am not.
First, and most importantly: optional privacy is not a compromise on the architecture. It is the only architecture that could have passed.
If you want a privacy asset inside a regulated wrapper, you need an asset whose holdings and counterparties can be proven on demand. Mandatory-privacy designs make that structurally impossible; the custodian can only attest, and attestation is exactly what the market stopped accepting after 2022. Zcash's dual-state ledger means a custodian can satisfy anti-money-laundering obligations, prove reserves, and execute a clean audit without modifying the protocol, without a fork, and without a governance vote.
The market is not arbitrarily preferring Zcash over Monero. It is selecting for the only candidate that satisfies custody. That is a structural advantage, it is durable for as long as the rules look the way they currently look, and anyone who dismisses it as a rounding error has not thought through what a custodian physically needs to do.
Second: in-kind seeding is standard practice, and treating it as inherently sinister is lazy.
New funds need scale to function. A fund with fifty million dollars of assets has wide spreads, poor market-making, and no chance of platform inclusion. Sponsors seed funds with in-kind contributions routinely, precisely because forcing the sponsor to go buy the underlying asset on the open market would create market impact, cash drag, and a worse outcome for every subsequent holder. The DCG contribution moved the fund toward viable scale, and that is a legitimate purpose. The bulls are right not to scream fraud on sight.
What I would say in response is narrower: the disclosure standard should scale with the relationship. An in-kind contribution from an unaffiliated institutional investor at arm's length is one thing. An in-kind contribution from the parent of the sponsor, priced by the sponsor, on a date the sponsor does not publish, is another. My complaint is not that the transaction happened. My complaint is that a fund capable of publishing a daily NAV has not published the valuation date on the single largest creation in its history.
Third: the split genuinely does precede option listings, and options genuinely change a product.
If ZCSH options get listed, the fund acquires a second life. Market makers take positions. Hedging flows appear. Liquidity begets liquidity. A product that was a passive exposure vehicle becomes an instrument other instruments are built on top of, and instruments that other instruments are built on top of are extremely difficult to kill. Bulls pointing at the split and predicting an option chain are reading the mechanics correctly, and it is a real, not a rhetorical, thesis.
Fourth: crossing a billion dollars changes distribution in ways that have nothing to do with price.
Platform inclusion. Screener appearance. Model portfolio eligibility. Advisor approval workflows that require minimum AUM thresholds before a product is even reviewed. These are the unglamorous mechanics of asset gathering, and they are more durable than any inflow number. A privacy asset clearing those thresholds is a larger structural event than any single quarter of flows, because thresholds, once cleared, do not usually get re-litigated downward in a hurry.
So here is the honest version. The bulls are right about the mechanism and wrong about the meaning.
They read a billion dollars as evidence of demand. It is distributed across three sources: price appreciation of an underlying that nearly doubled, an affiliate's balance sheet, and roughly two hundred million dollars of external money. Only the third is a signal about anyone's desire to own ZEC through a wrapper. The first is beta. The second is a related-party transaction. The third is demand, and it is smaller than the headline by an order of magnitude in terms of what it implies.
And there is a blind spot on the bullish side that neither camp discusses, because it is uncomfortable for both.
If the fund's compliance requirements push ZEC into transparent custody, and if institutional adoption pulls an increasing share of supply out of shielded pools, then the privacy property that gives the asset its premium erodes as a function of institutional success. Both sides of this debate are arguing about whether Zcash will be adopted. Neither is asking whether adoption destroys the thing being adopted. That is the question I would put to a bull who is genuinely confident rather than merely long.
Takeaway: What to Watch, and What the Wrapper Cannot Hide
Six signals will settle this over the next two quarters, and none of them is the AUM figure.
Weekly creation and redemption data, tracked separately from price. If external net inflows accelerate while ZEC is flat or falling, demand is real and the headline was merely early. If inflows dry up the moment price stops rising, the billion was beta.
The premium or discount to NAV, observed through a drawdown rather than a rally. A tight spread under stress means the arbitrage mechanism functions. A wide one means the tether is theatrical.
The valuation date of the in-kind contribution, if it is ever disclosed, and the ZEC market print on that date. If DCG contributed at a premium to spot, the transfer of value is quantifiable and the disclosure question becomes a governance question.
DCG's subsequent holdings, as they appear in whatever periodic filings apply. A parent that seeds a fund and then reduces is executing a different strategy than one that seeds and holds.
The ratio of shielded to transparent ZEC, tracked across the migration of supply into institutional custody. If ETF adoption shrinks the shielded share, the product is consuming its own premise.
The venue composition of the price index, and whether it survives the next delisting or venue outage intact.
None of these requires access to anything a retail holder does not have. All of them are observable. That is the point of doing this kind of work in public.
The forward question is not whether a privacy coin can be placed inside a billion-dollar wrapper. It clearly can. The question is whether the wrapper can hold anything the wrapper's own compliance regime requires it to see. So far the answer has been a transparent address, a published NAV, an unpublished pricing date, an affiliate's coins, a split that creates nothing, and a governance structure in which the holder has no vote on any of it.
That is not a scandal. It is a disclosure gap, and gaps close in one of two directions: with information, or with a repricing.
Audit the promise, not the poster. The poster says privacy. The promise, for now, is a ticker.