The 20.5x Anomaly: Decoding ZEC and NEAR Spot ETF Flows in a Bear Market

Two US-listed spot ETFs printed net creations in the same trading session this week. One holds Zcash. One holds NEAR. The first absorbed $20.52 million. The second absorbed $999,270. That is a ratio of 20.5 to one. Now hold two more numbers from the same dataset. The ratio of the two funds' assets under management is 8.5 to one. The ratio of their lifetime cumulative net inflows is 2.3 to one. Three ratios, one asset pair, three different stories — and the gap between them is the only genuinely new information here. Everything else in the report is a rounding error wearing a headline. Most coverage of small-cap crypto ETF flows makes the same structural mistake. It reads a stock variable as a flow signal, or a flow variable as a demand signal, without ever checking whether the denominator supports the claim. I spent the 2024 ETF cycle building microstructure explainers for institutional readers on precisely this problem, and the discipline that came out of it is unglamorous. The flow number is never the story. The flow number's denominator is the story. So let us do the arithmetic the headline skipped. Start with what these vehicles actually are, because the wrappers are doing more work here than the assets. ZCSH is a US spot ETF holding Zcash — a proof-of-work chain whose distinguishing feature is optional shielded transactions built on zk-SNARKs, and whose supply schedule is a hard 21 million cap in the Bitcoin lineage. NRR is a US spot ETF holding NEAR — a sharded layer-one with inflationary issuance whose market narrative has spent the last two years bolted to AI and high-throughput execution. One is a monetary asset. The other is a productive platform. That distinction will turn out to be the entire argument. The wrapper itself deserves its own paragraph. An ETF share is not a token. It is a security that tracks a token. It cannot be staked, cannot be pledged, cannot vote, cannot be bridged, and cannot serve as collateral in any on-chain venue. It is an exposure certificate with a custodian behind it and an authorized participant network in front of it. This reads like pedantry until you try to reason about what a flow into it actually means. The data comes from SoSoValue, a professional flow-tracking service, and the underlying series are net creations — cash in minus cash out, settled on the fund's own accounting calendar. Anyone comparing these figures across platforms without checking the treatment of pre-market prints, T+1 settlement, and in-kind versus cash creation is comparing incompatible objects. I raise this not to impugn the source but because the methodological seams are exactly where the interesting misreadings live. And the calendar matters more than usual. We are in a drawdown. In a drawdown, price is a noisy and largely uninformative signal, and the reader's real question stops being "what will this be worth" and becomes "is this vehicle structurally sound." That is a question about stock, not flow. It is the question this dataset can actually answer, and almost nobody is asking it. One more piece of context, offered at medium confidence. The ZCSH ticker and its naming convention are consistent with lineage from a private Zcash trust converted into an ETF wrapper. The source gives the ticker and the net asset value but never states the issuer or the launch date, so I mark this as inference rather than fact. If it is right, it changes the interpretation of roughly $500 million of the $707 million in AUM, and I will come back to that. ZCSH holds $707 million in net assets. Its lifetime cumulative net inflow is $204 million. The gap is $503 million. That gap has exactly three legitimate sources: seed capital contributed at launch, in-kind share conversions from a legacy trust, and market appreciation of the underlying. It has no illegitimate source — this is arithmetic, not speculation. But the composition of the gap determines whether the fund is a demand story or an accounting story, and the report does not tell us which. My prior, at medium confidence, is that conversions dominate. Converted trusts carry a legacy float that predates the ETF and never had to be bought in the open market. Why that matters is mechanical. If a large share of AUM is converted legacy stock, then the fund's "flow" series is measuring only the thin marginal layer sitting on top of a static base — and it is that marginal layer which the market reads as institutional demand. A measurement illusion of exactly this kind is what produces headlines about a category "heating up" when the underlying book has not moved in eighteen months. Now run the same subtraction on NRR. Assets under management: $83.39 million. Lifetime cumulative net inflow: $88.26 million. The AUM is smaller than the cumulative inflow, by $4.87 million. Read that again, because it is the single most underreported fact in the entire dataset. For a fund's AUM to sit below its lifetime net inflow, the underlying asset must have depreciated net of fees and net of any distribution since inception. It means the aggregate holder base of NRR is underwater. Not "underperforming." Underwater. That has a mechanical consequence nobody in the coverage has drawn out. An underwater holder base is a redemption overhang. It does not force selling tomorrow. It sets a reference level above which every marginal holder has an incentive to exit through the creation window rather than the secondary market. If NRR ever prints a sustained net outflow, the correct interpretation will not be "sentiment turned bearish." It will be "the trapped float found the door." So we have two funds with opposite structural postures. ZCSH carries a $503 million cushion of non-flow value that makes its AUM robust to outflow. NRR carries a negative cushion that makes its AUM fragile to it. The headline comparison treats these as symmetric. They are not remotely symmetric. Size is the wrong lens. Velocity is the right one. ZCSH absorbed $20.52 million against a $707 million base — 2.9% of AUM turning over in a single session. NRR absorbed roughly $1 million against an $83.39 million base — 1.2%. ZEC's marginal velocity is about 2.4 times NEAR's. This is pure arithmetic and it deserves to be stated plainly, because it survives every objection about absolute scale. Even if both numbers are small, the ratio of a flow to its own stock is scale-invariant. That is why I trust it more than the headline figure. The second ratio in the report is the one I trust least. It lists an "ETF net asset ratio" of 3.38% for ZCSH and 1.32% for NRR. If that metric means ETF net assets as a share of the underlying asset's total market capitalization — a penetration rate — then ZEC's institutional penetration is roughly 2.6 times NEAR's, and the conclusion is that ZEC's marginal price discovery has migrated further into the ETF channel than NEAR's has. That is a double-edged finding. The more price discovery lives in one channel, the faster it moves in both directions. But the source never defines the denominator, and I have been burned by undefined ratios before. Treat 3.38% as directional, not as a penetration figure, until you can rebuild it from a live float and a live market cap. Filtering the noise to find the art means knowing which numbers are load-bearing and which are decorative. What would falsify the velocity reading is straightforward: if five consecutive sessions show ZCSH's flow-to-AUM ratio holding above 2% while NRR's stays below 1%, the gap is structural. If it collapses to a single print, it was an order ticket. Here is where the flow comparison stops being a curiosity and starts being a thesis. A passive ETF wrapper captures exactly one thing: the price of the underlying. It captures nothing about what the underlying does. For an asset whose entire value proposition is monetary — fixed supply, credible neutrality, optional privacy — the wrapper is a near-perfect fit. Holding ZEC and holding ZCSH are, for economic exposure purposes, the same act. There is no yield to miss, no governance to forfeit, no productive capacity left on the table. For NEAR, the wrapper is a category error. NEAR's value proposition is productive: blockspace, fee capture, developer activity, an application layer that compounds. A passive security captures none of that. It captures the token price and only the token price. The gap between what NEAR does and what NRR holds is the gap between a business and a share of its float. I learned this distinction the hard way in 2021, when I ran social-graph analysis on the Bored Ape cohort and argued that NFT value was decoupling from art and re-anchoring on community status signaling. The readers who misread that report were the ones who assumed the wrapper and the asset were the same object. They were not. A status good, a jpeg, and an ERC-721 token are three different things with three different price mechanisms that happen to share a chart. The same structure applies here. ZEC and NEAR are not two small-cap assets competing for institutional attention. They are two different kinds of object that happen to both have a ticker and a wrapper. Ranking their "institutional appeal" by comparing their flows is like comparing the volume of a gold ETF to the volume of a mining equipment manufacturer and concluding gold is more popular. There is a candidate explanation on the ZEC side that the data cannot support and that I will therefore state at low confidence: if Zcash sits anywhere near a supply-reduction cycle, the interaction between issuance contraction and wrapper absorption could be one contributor to its relative flow advantage. One contributor, not a cause. The report contains no evidence for it, and I would not build a position on a low-confidence inference dressed up as a finding. And this is where I push back hardest on the framing. If the data shows ZEC ahead by 20.5 times on the session, the honest reading is not that privacy is winning. It is that the wrapper suits a monetary asset, and the monetary asset is therefore easier to wrap. The code does not lie, but it is incomplete — and the incompleteness here is that the ETF structure has no way to express what NEAR actually is. The transmission analysis is where most readers will get this wrong, so let me be blunt about what an ETF flow does not do. It does not improve on-chain liquidity. ETF shares are securities; they settle in the legacy system and cannot be pledged into a lending pool, used as collateral, or supplied to an automated market maker. Compare that with a stablecoin inflow, which lands directly in the on-chain order book and deepens it within a block. A $20 million stablecoin mint improves DeFi depth measurably. A $20 million ETF creation improves nothing on-chain at all. It does not reduce float in any way that touches the shielded pool. ZEC's privacy feature — the thing that makes it ZEC — is orthogonal to how many shares a custodian holds. A custodian holding transparent addresses on behalf of a fund increases the fund's assets and does nothing for shielded transaction volume. If you want to know whether privacy demand is growing, you do not watch ETF flows. You watch shielded pool size. It does not confer governance. ETF holders carry economic exposure and zero protocol voice. For NEAR, which has an active validator set and a real governance surface, that means the ETF holder base is structurally disenfranchised. Ownership and control are separated by design. The only thing an ETF flow does is relocate price discovery. That is not nothing. But it is far less than the coverage implies, and it is close to the opposite of ecosystem health. Arbitrage is the market's way of correcting itself, and the arbitrage here runs between two pricing venues — the ETF and the spot book — not between a protocol and its users. One detail in the report deserves more scrutiny than it received: each category recorded net inflow into exactly one product. That sentence has two readings and they point in opposite directions. Reading one: the category contains exactly one product, so "only one" is a statement about supply scarcity. Reading two: the category contains several products and only one printed a positive number, so "only one" is a statement about dispersion — the rest were flat or negative. The first supports a bullish thesis about first-mover capture. The second supports a bearish thesis about a category where the marginal dollar picks a single winner and abandons the field. The report does not disambiguate, and the difference is not academic. What I can say with more confidence is that issuer distribution depth explains the size gap better than demand does. Bitwise, the issuer associated with NRR, is a specialist crypto asset manager — competent, credible, and not BlackRock. The distribution network behind a specialist is a fraction of the wirehouse and RIA footprint a megacap manager commands. When a fund's AUM is small, the first hypothesis to test is not "nobody wants this." It is "nobody can reach it." If ZCSH's $707 million partly reflects a converted legacy trust, the same logic runs from the other direction: its distribution is inherited, not built. Inherited distribution is real. It is not evidence of new demand. It is evidence of an old book. Scale check, because the bear market makes this essential. The two funds combined absorbed $21.5 million in a session. The large-cap crypto ETFs move multiples of that in an hour. On any absolute measure this is a corner-of-the-tape number, and it has no independent capacity to set price. But scale and information are different axes. A small number with clean internal structure tells you more than a large number with a muddy one. A 20.5-to-one flow ratio sitting against an 8.5-to-one AUM ratio is a clean structural fact: ZEC's wrapper saw genuine marginal acceleration relative to its own size. That is worth knowing even though the dollars are trivial. Then there is the regulatory layer, which the report treats as background and which I consider the actual story. A US-listed spot ETF on a privacy asset is a stronger signal than any flow number in the dataset. For years the regulatory posture toward privacy-enhancing technology ranged from suspicion to hostility, and the Tornado Cash precedent — where publishing code was treated as potentially sanctionable conduct — cast a long shadow over every developer working on shielding. A spot ETF does not resolve that. But it establishes that a compliant, auditable exposure vehicle for a privacy asset can exist in the US market. That is a meaningful crack in the wall, and it matters more than $20 million of creations. It also raises a question the report never touches: how does a custodian reconcile auditability with shielded addresses? A fund must be able to prove its holdings. Shielded transactions are, by design, opaque. My working assumption at medium confidence is that custody is restricted to transparent-address assets, or that shielded holdings require special disclosure and are capped. Either arrangement introduces tracking error and operating cost that a plain transparent-asset ETF does not carry. It also means the fund's exposure is to ZEC-the-asset rather than ZEC-the-privacy-technology — a subtle but decisive difference for anyone buying it as a privacy bet. And the risk here is structural, not directional. If policy toward privacy-enhancing technology tightens, the vehicle does not fall in price. It faces restriction, forced remediation, or delisting. That is a different risk class from volatility, and it does not appear in any flow chart. The consensus read of this dataset is that institutions are quietly accumulating privacy exposure while the smart-contract trade stalls. I think that read is backwards on timing and wrong on mechanism. The timing is wrong. In small-cap wrappers, creations are lagging indicators. Authorized participants hedge after the fact; the creation order is the settlement of a price move that already happened, not a forecast of the next one. A single session's inflow into a thin fund is far more likely to be the mechanical residue of last week's tape than a directional conviction signal about next month. Anyone building a thesis on one session is extrapolating from n=1, and the base rate for that specific error in this market is close to 100%. The mechanism is wrong too. When the float available to a wrapper is small, small dollars produce enormous ratios. A 20.5-to-one flow ratio is not 20.5 units of conviction. It is 20.5 units of a small denominator. The number that should anchor the analysis is $21.5 million in absolute terms, and at that scale the most parsimonious explanation is one or two large creation orders from a handful of participants — not a broad institutional reallocation. And the finding everyone skipped is the negative one. NRR's AUM sitting below its cumulative net inflow is a template, not an anomaly. Every small-cap crypto ETF launched over the next eighteen months will meet the same structure. The wrapper does not create demand. It relocates demand into a venue with a thinner order book, a wider spread, and fewer participants than the spot market it left. Relocation is not accretion, and a bear market is precisely when the difference becomes visible, because there is no rising tide to disguise the relocation as growth. The deeper cost is this: efficiency is the enemy of the outlier. The entire historical case for small-cap crypto exposure was that the segment was inefficient — mispriced, under-covered, and arbitrageable by anyone willing to do the work. Wrapping it in a 40-basis-point security with a market-maker obligation is the most efficient available method of stripping that premium out of the market. The wrapper is not a door for new capital. It is a machine for compressing the dispersion that made the asset interesting in the first place. So what do you actually watch from here? Watch the five-to-ten-day rolling net flow, not the single session. One print is noise; five consecutive prints in the same direction is a signal, and the half-life of a flow-driven narrative is short enough that you will know within two weeks whether this was a trend or a ticket. Watch the second issuer. If a 19b-4 filing appears for a competing ZEC or NEAR product, the first-mover capture story gets tested immediately, and fee compression will tell you more about institutional seriousness than any inflow figure will. Watch the language. Any FinCEN or SEC statement on privacy-enhancing technology is a structural event for ZCSH specifically, and it will not announce itself through price. And watch whether NRR ever prints a sustained outflow. That is the tell for whether the wrapper model works for productive assets at all, or whether it only works for monetary ones. The next narrative is not ZEC versus NEAR. It is which assets can survive the wrapper — which value propositions a passive, non-stakable, non-governance-bearing security can actually carry without lying about what it holds. Storytelling is the new consensus mechanism, but a wrapper only knows how to tell one story, and it is always about price.