Twelve token disclosure pages. Forty-one fields each. Four hundred and ninety-two data points. Eighty-three returned N/A.
I ran the scrape on a Sunday, mostly out of habit. The twelve projects shared two traits: each had closed a round north of $40 million within the last two quarters, and each shipped a documentation site that rendered cleanly on mobile. Clean rendering is not disclosure. Of the 492 fields, 83 read blank, or TBD, or to be announced. Seventeen percent of the dataset did not exist. I have run this scrape every quarter for three years. The N/A rate has never fallen.
Aggregate fully diluted valuation across the twelve: roughly $6.4 billion. The market did not reprice a single one on the basis of the missing seventeen percent. That is the anomaly.
In a bull market, blanks are not read as absence. They are read as optionality.
The template improved. The completion rate did not.
Crypto disclosure has moved through four formats in eight years, and every step traded auditability for aesthetics. The 2017 whitepaper was at least a fixed artifact โ thirty pages, timestamped, frequently dishonest, but immutable once published. The 2019 litepaper shortened it. The 2022 docs portal made it versioned and editable. The current standard, a Notion page behind a marketing domain, makes it ephemeral and unversioned.
The operational consequence is specific. A whitepaper token distribution table was difficult to fake because the rows had to sum to 100%. A docs site distribution table is a grid of placeholders that can be updated the week before a listing. Format improved. Rigor degraded.
Funding does not require disclosure. In this cycle, nine-figure rounds have closed on a deck, a founder's prior exit, and a Discord server. Allocators are not demanding the fields; they are competing for allocation, which is an entirely different incentive. When a round is oversubscribed, the buyer's leverage to request a vesting table disappears. That is why the N/A count tends to rise with the size of the round, not with the obscurity of the project.
In 2018 I audited fifteen ERC-20 contracts ahead of an XDAI testnet migration. In the standard implementation behind Project Alpha, I found a critical integer overflow โ the class of bug that lets a transfer wrap and manufacture balance from nothing. Exposure at the time was roughly $40,000. I wrote the report. The founders called it too aggressive and pushed it back. I published the diff on GitHub instead. Three security researchers cited it inside a month.
The lesson was not that I was correct. The lesson was that the field labeled audit functioned as a marketing input rather than a technical output. Everything downstream of that distinction is the game.
Pricing the blanks
Now, order flow. When a field reads N/A, someone is buying it at par. Map the buyer.
Vesting. The most expensive blank on any page. Consider a $200 million FDV token with 14% float and a team allocation reading subject to a vesting schedule to be finalized. If that allocation is 20% and it carries a six-month cliff instead of a four-year linear unlock, the float model is wrong by a factor, not a percentage. Forty million dollars of supply the tape has not priced. Run the sensitivity: at a twelve-month linear release, monthly sell pressure lands near $3.3 million against perhaps $800,000 of organic fee revenue. The blank is not a neutral gap. It is a $2.5 million monthly deficit wearing a polite mask.
Audit status. "Audit in progress" is the second-most-expensive blank. An audit is a snapshot, not a warranty; it reports that the reviewed bytecode was not obviously broken on the review date. It says nothing about the upgrade path, the admin keys, or the deployer's intent. Audit the code, then audit the intent. A completed audit is weak positive evidence. A withheld audit is strong negative evidence, because the only rational reason to sit on a finished report is that it reads badly. Two of the twelve projects I scraped carried a live admin-mint function and an audit field reading TBD at the same time.
Revenue composition. APR listed as dynamic. In 2020, during the gas spike to 500 gwei, I ran a pre-coded rebalancing script that unwound positions automatically and preserved 92% of capital while yield-chasers lost 40% to slippage. The lesson generalizes: a yield sourced from emissions is a liability with a schedule, not revenue. If a protocol will not decompose its APR into fees versus incentives, assume the fee share is under ten percent. That assumption has held in every dataset I have pulled.
Bridge and security model. "To be published." Every added interoperability protocol introduces a solver set, a message-passing layer, and a liquidity pool that must be funded from the same finite capital base. More chains do not concentrate liquidity; they fragment it. A blank bridge disclosure is not missing information. It is a structural admission that the security budget is still being negotiated with the people who profit from its absence. The same logic governs the Layer 2 field. The published difference between the OP Stack and the ZK Stack is architectural. The operational difference is which team can convince more projects to deploy sovereign chains first, because every new chain adds a bridge, and every bridge is a fresh audit field waiting to read TBD. Distribution beats cryptography in the disclosure column, every cycle.
There is a precedent for permanence. The Lightning Network routing success rate has been the industry's longest-running N/A. Seven years of nodes, channels, and liquidity provisioning, and the field remains effectively blank at any scale that matters, because channel management friction and routing failure do not improve with volume; they compound with it. Some blanks are temporary. Some are structural. Separating the two is the analyst's actual job.
Follow the launchpool. Exchanges list tokens whose distribution tables are thinnest, because thin float produces volatility, and volatility produces volume. The venue has no incentive to demand the vesting field; the venue profits from the blank. When a token lists on three venues in the same week with no published unlock schedule, you are not watching price discovery. You are watching distribution, executed against buyers who cannot read the supply table.
In 2022, running a desk for a small fintech, I had pre-committed a circuit breaker that halted algorithmic stablecoin trading thirty seconds ahead of the main crash. It was not a prediction. It was a rule that removed discretion at the exact moment discretion becomes most expensive. The firm stayed solvent; peers who reasoned live did not. The same discipline applies to disclosure: you set the threshold when the page is calm, not when the page is comfortable.
None of this requires malice. It requires only that the blank field be cheaper to publish than the filled one. Producing a completed vesting table invites questions from the community, the market makers, and potentially a regulator. Leaving it blank invites none. The incentive gradient points one direction, always.
The short put nobody thinks they sold
The standard defense of a blank field is that it functions as a free call option: zero cost, unbounded upside, no downside if nothing ships. That framing inverts the payoff. A blank field is a short put written by the buyer. Disclosed risk gets priced into the float and stops moving the tape. Undisclosed risk gets deferred, and deferred risk settles at par, in one direction, without notice.
In 2021 I held a $120,000 NFT floor position across CryptoPunks and Bored Apes. When the bid thinned, I triggered a stop-loss at a 15% drawdown and sold 60% inside an hour. My peers held, because the pages they were reading still said blue chip, which is a narrative field, not a liquidity field. The decision preserved $70,000. Ledger books, not feelings, settle the debt.
Retail reads N/A as "not yet disclosed." A desk reads N/A as "disclosed somewhere else." The information exists; it simply has not been published to you. When a founder tells me a vesting table is still being finalized, what they have actually said is that it is finalized and unfavorable.
Takeaway
Build the screen. Scrape the docs sites. Count N/A per hundred fields. Flag anything above twelve. Then isolate the three blanks that move float: team vesting, audit status, revenue composition. If all three read empty, the position size is not small. It is zero.
The market is currently paying a premium for pages it cannot read. Liquidity dries up when confidence breaks, and confidence breaks first at the fields nobody filled in. Anyone can read a green candle. Very few will read a blank field and treat it as a price. The next unlock calendar prints in six weeks. Count the blanks before you count the candles.