Last quarter a nine-dimension analytical framework came back across my desk with a unanimous verdict. Technical positioning: N/A. Token supply structure: N/A. Market cycle assessment: N/A. Ecosystem dependencies: N/A. Regulatory posture: N/A. Team and governance: N/A. Risk matrix: N/A. Narrative sustainability: N/A. Supply-chain transmission: N/A.
Not "insufficient data," which at least implies an attempt to gather. Not "inconclusive," which implies a hypothesis that failed a test. Every cell resolved to the same null value — including the confidence intervals, which is a category error, because you cannot assign a confidence level to an absence.
The document had a table of contents, three appendices, and a disclaimer section. It was structured correctly. A framework that returns N/A across every dimension has not failed; it has executed exactly as designed, and that is the problem.
I have been doing this work since 2017. That was not the most alarming report I have read. It was the most honest.
The crypto research industry has spent four years industrializing the template and de-industrializing the input pipeline. Nine dimensions is the standard shape now. Technical: what the code does, what it assumes, what it structurally cannot do. Token economics: emission schedule, unlock cliffs, treasury composition, and whether the advertised yield is revenue or transfer. Market: liquidity depth, venue concentration, funding rates. Ecosystem: who depends on whom, and what breaks first. Regulatory: securities posture, KYC surface, jurisdictional footprint. Team: shipping history rather than slide decks. Risk: the matrix nobody reads. Narrative: what the market believes versus what was delivered. Transmission: what happens to adjacent sectors when this thing moves.
Each dimension exists because someone lost money on its absence. The unlock schedule exists because insiders sold into retail liquidity. The treasury dimension exists because "revenue" turned out to be a token emission loop. The transmission dimension exists because Terra's failure did not stay inside Terra.
In a bull market, these templates are marketing collateral. A fund publishes a sixty-page report to demonstrate rigor, nobody audits the inputs, because everything is up and checkable inputs are expensive. In a bear market the same template becomes survival equipment. It is supposed to answer one question: is this asset safe to hold through a drawdown?
That is when desks discover the template never had inputs. It had headings. The headings were doing the work that data was supposed to do.
There are three species of N/A, and the industry deliberately conflates them.
The first is not disclosed. The protocol holds the data and does not publish it — insider vesting schedules, treasury wallet attribution, side agreements with market makers. This is an adversarial N/A. It is information in itself, because something is being withheld and the withholding has a cost that someone decided was worth paying.
The second is not verifiable. The claim exists but cannot be tested against a primary source: "audited by a top-tier firm" without a published report, "backed by Tier 1 VCs" without a cap table, "partnership with a major bank" without a contract. This N/A is a claim laundered into a fact by repetition.
The third is not assessed. The analyst had access to public data and did not look. This is the only species that belongs to the researcher, and the only one that carries professional liability.
Most published frameworks blend all three into a single gray cell. That blending is the failure. A withheld unlock schedule and an unchecked unlock schedule have opposite implications: the first means reduce exposure now, the second means open a block explorer and spend the afternoon.
I learned the taxonomy the expensive way. In 2017 I reviewed contract logic for an ERC-20 launch and found three arithmetic overflow vulnerabilities in the voting mechanism, using nothing more elaborate than Python scripts and a forked testnet. The team's whitepaper had a risk section — well-formatted, two pages, listing "smart contract risk" as a category without naming a single function. When I delivered the findings, the token was up 400% and my analysis was described as too theoretical. Three months later the project collapsed through the exact overflow paths I had flagged. Code compiles, but context reveals the exploit. Documents compile too, and they fail the same way.
What the N/A report actually documents is the collapse of the input layer. Here is the ranking that matters in a bear market — which dimensions resolve with public data, and which never will:
| Dimension | Primary data source | Resolvable without cooperation | Bear-market weight | |---|---|---|---| | Token supply / unlocks | Vesting contracts, treasury wallets | Yes | High | | Liquidity depth and LP concentration | DEX pool state, holder distribution | Yes | High | | Revenue vs. incentive spend | On-chain inflows, emission contracts | Yes | High | | Technical risk | Bytecode, audit reports | Partially | Medium | | Governance concentration | Proposal history, voter addresses | Yes | Medium | | Team quality | Mainnet commits, incident response | No | Low | | Narrative sustainability | Nothing | No | Negative |
The bottom two rows are where every empty framework spends its word count and produces nothing of value. Team quality is not measurable; it is inferred from behavior under stress, which means you only know it after you needed to know it. Narrative is not measurable at all; it is a social quantity that re-prices without a data source. Both dimensions feel rigorous, because both invite long prose, and both are unfalsifiable. A framework that weights them equally alongside unlock schedules is not nine dimensions. It is two dimensions and seven paragraphs of vibes.
The measurable rows are where I have actually caught things. In 2020 I built a SQL dashboard tracking daily liquidity-mining APYs against a protocol's treasury reserves. The APY field was promotional. The reserve field was public. The ratio between them was a solvency curve, and it was flattening weeks before the incentive program was paused. The report was ridiculed by accounts that were reading the APY field. The measurable field is always the one nobody markets.
The same discipline applies to volume. In 2021 I traced roughly 15% of weekly NFT volume to wash-trading clusters connected to a single governance wallet, inflating apparent market cap by a minimum of $40 million. Art value is not measurable. Wallet clustering is. Every report I have produced since carries a Wash Trading Index row, and I now apply the same index to research itself: what share of published protocol analysis contains a primary-source denominator? In my sampling it sits below one in five. The research market has its own wash volume, and it is not small.
Comparative work is the cheapest source of information gain available, and it requires zero cooperation from the projects involved. After May 2022 I ran a fifty-page assessment comparing algorithmic and partially-collateralized stablecoin designs. The consensus framed Terra's collapse as a design flaw unique to one mechanism. The data said something narrower and more uncomfortable: both models price stability off the same variable — market confidence — and neither holds a hard floor beneath it. That conclusion only existed because there were two datasets to place side by side. A single-project report would have produced a narrative instead of a finding.
In 2025 I led a MiCA compliance audit for a Portuguese crypto asset service provider. Their transaction-monitoring system had fields marked "not applicable" that should have been marked "not tested." Mapping those fields against the new data requirements exposed a gap priced at roughly €10 million in fines. The fix was not better software. The fix was refusing to accept N/A as a terminal state. We built rule-based testing that converted every unassessed field into a pass or fail, and the firm secured its license while two competitors did not.
That is the operative distinction. In compliance, an N/A field is an unresolved liability. In investment research, it is a hidden short position.
Here is what the optimists get right, and it is not nothing. An honest N/A report is more valuable than a confident one. The research market is saturated with filled-in templates where every dimension has a number sourced from the project's own dashboard, and the number is wrong. A report that says unresolved nine times is a stop signal, and a stop signal in a bear market preserves capital more reliably than a buy signal generates it. Most losses I have reviewed post-mortem were not caused by missing data. They were caused by fabricated data that arrived on schedule.
The second thing they get right: the input layer has genuinely improved. Public RPC archives, decoded contract repositories, query layers, and mandatory disclosure regimes mean that a dimension returning N/A in 2026 is a choice rather than a constraint. In 2017 I reverse-engineered bytecode by hand. Today the same question costs an afternoon of SQL. Code compiles, but context reveals the exploit — and context is now purchasable.
That reframes the critique. The problem is not that the template is empty. It is that someone decided to keep it empty and ship it anyway.
MiCA's enforcement wave and its equivalents will not read your framework. They will read your fields, and an unassessed cell is now a documented exposure with a price attached to it. The question for every desk running a nine-dimension template this quarter is not whether the framework is complete. It is whether any single cell in it can survive contact with a primary source. If the answer is no, you have not performed due diligence. You have performed formatting.
