Last month, a junior developer I mentor forwarded me a research report from a well-known crypto desk. Nine sections, five charts, a definitive verdict in every field. I read it twice, then opened the underlying contracts myself. The token was a governance wrapper whose distribution schedule the authors had clearly never modeled. The TVL figure came from a dashboard that counts self-lending as organic liquidity. The "bullish" thesis rested on an emissions schedule that mints more tokens every day than the protocol collects in a week of genuine fees.
We didn't ask for a framework that would make us feel smart. We asked for one that would keep us honest.
The same week, a friend in institutional research sent me something stranger. Four thousand words, nine analytical dimensions, and every single one terminated in the same two letters: N/A. No verdict. No charts. No confident summary. Just a disciplined refusal to conjure conclusions from missing information. My first reaction was frustration — what good is a report that cannot make a call? My second was recognition. In a market drowning in confidence, the rarest and most valuable sentence an analyst can write is "I don't know."
This is the story of why that empty report, more than any polished dashboard, deserves to be treated as a model for what crypto analysis could become. It is also a story about what technical rigor and human humility look like when they actually share a byline.
The Bear Market's Information Ache
We are in a strange season. The ETF approval of 2024 normalized crypto in the institutional imagination, but the price action since then has been a lesson in gravity. Retail users are scared. The developers who endured the 2022 collapse and the AI-agent mania that followed are exhausted. Everyone, everywhere, keeps asking the same question about their assets: is this safe?
The information ecosystem has answered that fear in the worst possible way — with more certainty, not more honesty. Security shops publish audit summaries with clean verdicts and buried assumptions. Token analysts publish scores with three decimal points of precision and zero notes about the data they were denied. Social media rewards the loudest forecast and quietly buries the person who, two weeks ago, admitted they did not have enough information to forecast at all.
The economics of attention are structurally hostile to epistemic honesty. A nine-dimensional report that returns "insufficient information" in every field does not get clicks. It does not get thread-quoted. It does not soothe anyone's anxiety. It does something harder: it forces the reader to feel the size of the unknown. It does not appear in anyone's quarterly performance review, either, which is precisely why so few people are brave enough to write it.
In open source, we would never trust a pull request that claims to patch a critical vulnerability without showing the diff. In crypto research, we routinely accept verdicts without examining the information diff behind them. That asymmetry is the root problem. And it is why a report populated entirely with N/A fields struck me as the most radical, useful piece of research I had encountered in months.
The report that triggered this reflection was not an outlier. It was the output of a template that has been circulating quietly among research groups in Hangzhou for months — a nine-dimension grid covering technology, tokenomics, market position, ecosystem role, regulatory posture, team governance, risk profile, narrative cycles, and industry-chain transmission. The template is rigorous, which is exactly why its output is unsettling. Applied to a project that refuses to disclose its token distribution, or a protocol whose revenue model cannot be verified, the grid does not produce a confident score. It produces nine honest admissions of ignorance. The author of the report I received chose to publish the admissions rather than bury them.
The Rented Numbers
Let me get concrete about which data gets faked, because this is where the technical life of the article begins. In DeFi, the single most abused metric is total value locked, and it has an inflation problem that no aggregate dashboard will ever solve, because the inflation is structural.
Liquidity mining APY is, in almost every farming product I have audited, the project subsidizing its own TVL with freshly minted tokens. I have watched this cycle replicate itself since 2020. The pattern is identical every time: a treasury announces a reward schedule, LPs rush in to harvest emissions, the dashboard shows a soaring TVL, and then, when the schedule thins or a competitor offers a higher number, the liquidity evaporates. Stop the incentives and the real users vanish.
Take a concrete example, the shape of which I have seen a dozen times. A farming protocol announces a 400% APY. Within two weeks, $300 million in liquidity arrives. A dashboard celebrates the TVL milestone. The protocol's real revenue from trading fees is $400,000 per week. Its token emissions, at the current market price, are worth $1.2 million per week. The emissions-to-revenue ratio is three. The protocol is burning through its own token supply at triple the rate it is earning anything real, and the 400% APY is simply the burn rate printed as a yield. If you model the incentives winding down over the next six months, you will find that the vast majority of that $300 million has no reason to remain. The report that says "TVL: $300M" without that model is not neutral. It is actively participating in the misdirection.
The way you catch this on a first pass is simple arithmetic. Open the token contract and count the daily emissions. Compare that flow against the protocol's actual revenue from fees. If the emissions-to-revenue ratio is above one, then the APY is a marketing expense, not a yield, and the TVL printed above that protocol is rented, not owned. I call it the sticky TVL ratio, and I have never met a research desk that publishes it. I have, however, seen a dozen desks publish the headline number while copying it from a dashboard that was itself telling a story, not reporting a fact.
A report that cites TVL without modeling the token flow behind it is not doing analysis. It is performing a form of party-trick arithmetic. The honest version would show two numbers side by side: stated TVL, and "TVL that would remain if emissions dropped to zero tomorrow." The distance between those two numbers is the most important fact about the protocol's strategy. It tells you whether the team is renting attention or earning it.
This is why I keep defending the N/A report. When a project's public data does not permit an analyst to compute the sticky TVL ratio, the honest analyst has two options: find a better source, or write a large, unmissable N/A. That absence is not a failure of analysis. It is itself the finding. It means the protocol is not yet accountable enough to be evaluated, and that unaccountability is an evaluation in its own right.
The Blob Countdown
The infrastructure layer offers a cleaner case, because the uncertainty there is temporal rather than fraudulent. Consider the post-Dencun data landscape. EIP-4844 gave rollups a cheap blob garage, and the entire L2 ecosystem responded by parking more cars in it than the garage was designed to hold.
The blob base fee is no longer the calm, negligible line it was in the first months after the upgrade. The quarterly averages hide a growing variance. Based on the adoption curves of the major rollups, and on the blob-posting traffic I have been tracking since the Dencun hard fork, I believe the available space will reach saturation within roughly two years. When it does, the cost structure of every mainstream rollup shifts upward at once. Blob gas prices will climb, the comfortable L2 experience users have enjoyed since 2024 will erode, and all rollup gas fees will double again.
Before Dencun, rollups paid calldata gas for every transaction batch, and the cost floor was high enough to make L2 usage feel like a premium product. The blob upgrade changed the pricing model by giving rollups a dedicated, cheap data lane. But a lane is still a shared road. When every optimistic and zero-knowledge rollup posts to blobs, the base fee behaves exactly as it was designed to behave: it rises. The question has never been whether saturation happens; the question is whether the industry will treat the warning as a design constraint or as an unexpected tax. The analysts who return N/A on L2 cost forecasts are the ones who understand that the answer is still being written.
Some teams will migrate to alternative data availability layers — Celestia, EigenDA, or newer projects that are not yet mature enough to evaluate. Most will not migrate in time, because migration itself carries engineering risk and because a cheap solution today rarely feels urgent until it is expensive. The analytical implication is blunt: any L2 performance study that treats today's blob cost as a permanent baseline is automatically N/A. Not because the analyst made an arithmetic error, but because the report fails to notice that the protocol's cost curve is a function of a shared, congestible public resource.
I find this case clarifying because it demonstrates honest uncertainty working in a domain where nobody is lying. The data exists. The trajectory is inferable. But the precise month of saturation depends on adoption decisions, upgrade schedules, and the migration behavior of dozens of teams. A truthful analyst in this bear market does not write "L2 fees will stay low." A truthful analyst writes "low fees are borrowing against future blob capacity, and I cannot verify when the bill arrives." That is a forecast with integrity — the financial engineering equivalent of flagging a duration mismatch instead of pretending it does not exist.
The Human Dimension No Scorecard Captures
The most important missing information, though, is not on-chain at all. No analysis framework I have seen includes a dimension for the moral and emotional state of the core team, and yet I have watched that variable determine project outcomes more reliably than any token metric.
In 2022, when the market collapsed, I partnered with three open-source foundations to build a survival guide for developers and early adopters. The anxiety was not abstract. It arrived as insomnia, burnout, and the quiet decision to leave the industry permanently. I personally mentored fifteen junior engineers through that period, helping several move from speculative trading into infrastructure roles.
I remember one engineer in particular. He had built a derivatives protocol that survived the 2022 liquidation cascade, only to watch his co-founder leave in January 2023. The protocol's code was sound. Its treasury was adequate. But he was alone, and the alone-ness was not visible in any dashboard. Over six months of weekly calls, we rebuilt his support network before we ever touched his token model. The protocol survived because a human being was cared for, not because a metric was optimized. I do not know how to put that into a scoring grid, and I have stopped pretending otherwise.
No nine-dimensional framework captures that. The honest report, the one full of N/A fields, is at least structurally truthful about its blindness on this front. The polished report is worse, because its confidence implies that a team's commitment, integrity, and collective endurance can be inferred from commit counts and engagement metrics. They cannot.
The Theater of Regulatory Clarity
Regulatory analysis has the same problem, dressed in a more formal costume. The framework that produced all those N/A fields includes a Howey test assessment — whether a token involves an investment of money in a common enterprise with an expectation of profit derived from the efforts of others. It is a reasonable checklist. In practice, it is also a stage prop.
We didn't enter this industry to escape the law. We entered it to escape opacity — to build a system where everyone can see the reasoning behind every rule. Replicating the opacity in our analysis would be a quiet defeat. And yet every project claims it is not a security. Very few publish the internal token economics decks, the investor communications, or the promotional statements that would allow an external analyst to apply the test meaningfully. And in the post-ETF world, the boundary has become even more contextual. A token that functions as a security in a retail distribution may play a completely different role in an institutional settlement system. An analyst who issues a "regulatory compliant" verdict without access to the actual distribution mechanics is issuing a verdict-shaped object rather than a verdict.
My own regulatory predictions have been wrong before, and they will be wrong again. I thought, in 2024, that ETF approval would either crush decentralization or be absorbed without consequence. It was neither. It became a messy, contradictory third thing — precisely the outcome that binary analysis cannot accommodate. The only fully honest regulatory report in a bear market says: I know what the law was last quarter, I can read the current enforcement signals, and I cannot tell you how a court will rule on a question that has not yet been litigated. That is not a dodge. It is the accurate state of knowledge.
The 2017 Lesson, Still Unlearned
I have practiced this discipline long enough to remember what happens when it works. In late 2017, during the ICO mania, I led a volunteer audit team for a prominent Ethereum-based utility token project. I spent forty hours reviewing the whitepaper's economic model. What I found was a token distribution that materially favored insiders — not illegal, but corrosive to the project's decentralization claims. I published a detailed, empathetic critique on Medium. It reached fifty thousand readers. The team revised their allocation.
That experience taught me that honest analysis is not a refusal to act. It is a more precise form of action. Saying "the information is missing" does not abstain from judgment. It intervenes in the reader's decision process by forcing a choice: either locate the missing piece, or accept a risk that has not actually been measured. That is the service the N/A report performed for me. It did not tell me what to think. It did something rarer. It showed me the outline of what could not yet be known.
Assumptions Are Where Exploits Hide
Formal verification has a phrase for this. Every smart contract audit is a judgment that ends with a list of assumptions. Most readers skim that list. The auditors know that the assumptions are where the next exploit will be found, because code tends to be correct where it is most constrained and wrong in the space the auditor could not see.
Financial analysis is exactly the same. A claim is only as strong as the information it is permitted to ignore. The assumption that an information gap is an opinion gap is how money gets lost. The analyst who disagrees with a peer's forecast is working from the same facts. The analyst who never records the absence of facts is not disagreeing; they are preemptively forfeiting. They are guessing in a costume, and the costume is what the market pays for.
What the N/A Report Does Not Do
Now the contrarian angle, because I do not want to romanticize the blank page. There is a version of epistemic honesty that becomes an escape hatch — the analyst who says "I don't know" and then stops, as though the sentence itself were a deliverable rather than the starting point of a better inquiry.
We didn't lose money in 2022 because our metrics were wrong. We lost it because we treated confidence as a substitute for verification. But we will not save ourselves in this bear market by treating N/A as a substitute for investigation. The empty report is a precondition for good analysis, not a replacement for it. Its value is that it forces the reader to carry the weight of the absence. If it makes you nervous, it is working. If it lets the analyst off the hook — if "I don't know" becomes the end of the inquiry instead of the reason the inquiry must be pursued — then the framework has been absorbed into the very theater it was meant to expose.
There is also a subtler failure mode that I have to name, because I have felt its pull myself. The framework itself can become a crutch. A nine-dimension grid offers the comforting illusion that if we just keep collecting categories of information, certainty will assemble itself out of the pieces. The N/A report resists that illusion, and that is good. But the same grid, in less disciplined hands, becomes a factory for false precision — nine boxes to check, nine opportunities to fill a blank with a guess. A framework is not a source of information. It is a way of organizing the absence of information. The moment we forget that distinction, the grid becomes gossip dressed in a lab coat.
So the standard I am fighting for is not a blanket N/A. It is N/A plus a path to resolution. A report that says "the token distribution is unauditable" should also say "here is the data we would need, and here is the request we sent to the team." A report that identifies liquidity subsidies should also compute the break-even fee revenue required for sustainability. An honest caveat is not a shrug. It is a demand.
I am proposing, in other words, an Epistemic Disclosure for every piece of crypto research: a short section stating what was not known, what was assumed, and what the analyst did to shrink the gap. In open source, we have LICENSE files that define the terms under which code can be used. In financial research, we need something analogous for certainty. We need to know what we are licensed to trust.
A Badge of Honor
The market will not reward this immediately. The N/A report will continue to lose followers. The confident forecast will continue to capture engagement, and its author will continue to be resurrected by the next bull run, even if every prior prediction was an accidental consequence of a rising market.
But I am not writing for the algorithm. I am writing for the junior developers, like the one who forwarded me that polished report, who are still deciding what kind of analysts they want to become. I want them to know that the most useful thing a careful researcher can offer is not a price target. It is a map of the unmapped.
We didn't build open source so that a few people could hold all the answers. We built it so that every contributor could check the answers together — and inspect the questions, too. This industry was founded on the principle that verification is a public good. It is time for that principle to govern our analysis as strictly as it governs our code.
What if "N/A" became a badge of honor — a mark that an analyst has decided not to lie to you? What if, the next time a report tells you it cannot see, you read that as an invitation to look with them?
I would read that report first.