It was near midnight in Polanco, the mezcal going warm on the table, when the PDF finally opened. Forty pages. Nine analytical dimensions. A risk matrix shaded like a weather radar. And nearly every cell held the same three characters: N/A. A junior analyst at a fund I advise had sent it along with a nervous voice note โ "would you read it before Monday's committee?" โ and the timestamps told me he'd spent two weeks on it. The scaffolding was flawless: technical, tokenomic, market, ecosystem, regulatory, governance, risk, narrative, supply-chain transmission. Underneath the scaffolding, nothing. No thesis, no inputs, no object of study. A cathedral built around an empty lot. I laughed. Then I stopped, because I recognized the handwriting of the machine that produced it. I had spent years building frameworks exactly like this one โ clean, neutral, defensible โ and I knew precisely how a sharp person ends up generating forty pages of nothing and calling it diligence.
The demand for crypto research industrialized this cycle. When the spot ETFs opened the institutional door, every fund, DAO, and protocol treasury suddenly needed "due diligence" that resembled the equity reports their compliance teams already trusted. In Mexico City, where I advise family offices and a couple of hedge funds, boutique research shops ballooned from two analysts to twenty in eighteen months, all shipping the same nine-box template. The template has an honest ancestor: equity analysts evaluate a business across earnings quality, balance sheet, management, and moat. Mapped onto a token, the equivalent looks obvious โ audit the code, model the supply, size the market, vet the team. The catch is that a token is not a company: no earnings statement, no board, no audited books, so the mold fills with placeholders. The N/A is not a failure of effort. It is an accurate report of a category mismatch, dressed in the costume of diligence.
I learned that costume the expensive way. After my own book fell from roughly $200,000 through the Terra and FTX unwind, I stopped building better frameworks and started deleting them. What kept me solvent afterward was cruder and less elegant: watching global liquidity and on-chain behavior, and ignoring any column that demanded I "score the team."
Here is the test I now run on any research deliverable, mine included. I cover the conclusion and ask a single question: what decision would this document have changed? If the honest answer is "none," it isn't research. It's liturgy.
The report I read failed that test cleanly. Its tokenomics page had rows for team, early investors, community, and treasury; every row was blank, because the source was blank. To his credit, the analyst refused to guess. The failure sat upstream: someone commissioned a nine-dimension analysis before confirming that even one dimension had inputs. The framework executed automatically, like an oracle-less smart contract returning default values.
That mechanical quality is the real story, and it isn't confined to research. I see the same pattern in the products I cover. Liquidity mining programs advertise an APY, and the APY is the framework; the underlying "users" are the empty cells. Pull the emissions and the TVL column reads N/A within weeks, not because the analysis was wrong but because there was never a business underneath the yield. Subsidized TVL and subsidized attention fail identically: both are frameworks pretending to be fundamentals.
Layer-2 sequencing offers the same tell. For two years, roadmaps promised "decentralized sequencing," and the nine-dimension reviewer dutifully scores it under technical architecture. Then you check who actually runs the sequencer, find one operator holding the admin key, and score the row N/A โ because the decentralization was always a slide, never a system. The framework couldn't see it. A five-minute query of the sequencer's address could.
I saw the mirror image in Bitcoin's hashrate after the fourth halving. The narrative columns lit up with "network security at all-time highs," while the revenue line quietly fell off a cliff and hash power began concentrating into a shrinking set of pools. Analysts kept scoring "decentralization" as a strength because the category existed on the form. Nobody scored the thing that mattered: a form is not an instrument.
Which brings me to the part of the empty report I found genuinely useful. The absence was itself data. A framework that returns N/A across every dimension isn't a neutral void โ it's a confession that the commission was performative. In a bull market, that confession is everywhere, and it's readable if you know the tell. When a project funds a nine-part "research initiative" but publishes no repository, no audit, no unlock schedule, the volume of the framework runs inversely to the substance of the asset.
Let me put numbers behind that. Suppose a freshly funded project announces a $100 million ecosystem fund and a matching nine-dimension "research portal." Strip the marketing: the ecosystem fund is a treasury allocation with vesting, and the research portal is content marketing. The only hard inputs are two addresses and an unlock table. Everything else is N/A with better typography. My rule since 2022: count the inputs, not the dimensions. Three verifiable inputs beat nine elegant blanks.
Last autumn a hedge fund client in Mexico City handed me a nine-dimension report on a liquid-staking protocol. Every risk row was graded; every ecosystem arrow was drawn. I asked one question: what does the protocol earn, net of token emissions, after paying validators? The room went quiet. The document had an entire dimension labeled "tokenomics," and none of it answered that. We pulled the on-chain revenue โ the number existed, it simply wasn't on the form โ and found that real fees covered about 8% of the advertised yield. The other 92% was emission. The framework had scored the asset "low risk." The inputs had scored it "subsidy."
The regulatory dimension deserves its own N/A, because it exposes how template thinking fails at the moment of highest stakes. Every form carries a Howey box, and reviewers dutifully tick money-investment, common-enterprise, expectation-of-profit, efforts-of-others. But the answers depend on facts the framework never asks for: who holds the admin keys, whether the foundation can mint at will, what the unlock schedule does to circulating float. I've watched a report score a token "low securities risk" while the founder's address controlled 40% of supply on a two-year cliff. The form asked the legal question. Nobody asked the custody question. That gap โ between the box and the chain โ is where the actual risk lives, and the template is structurally blind to it.
The community dimension deserves its own warning, because it's the one I'm most tempted to over-credit. I've watched Discord servers of forty thousand members dissolve into ghost towns the week emissions stopped, and I've watched a nine-dimensional "community health" score miss it entirely, because heat is easy to fake and retention is hard to model. The behavioral reality โ that enthusiasm is a lagging indicator of liquidity, not a leading indicator of value โ never fits the form.
This is why my own model is so stubbornly boring. Axis one: dollar liquidity โ TIPS yields, the M2 impulse, the direction of the funding rate. Axis two: on-chain cash flow โ fees actually paid, in a stablecoin, to an address you can read. Everything the nine-dimension template asks about, from team to narrative to moat, either feeds one of those two axes or it doesn't matter. Most of it doesn't. I've run this since 2022, and it has one virtue over the elegant template: every cell has an input, because I refuse to include a dimension I cannot measure.
There's a second-order insight buried here. In a bear market nobody pays for a nine-dimension report, because nobody is allocating. The framework economy is itself a bull-market artifact, a product of surplus capital hunting for reasons to move. So the existence of an elaborate research apparatus is a sentiment indicator, not an analytical one. When my inbox fills with beautifully formatted studies of assets that produce no cash flow, I don't read them as research. I read them as a liquidity cycle reporting its own temperature. The volume of the form is the signal. The content is the noise.
Now the uncomfortable part, aimed at my own desk. The reflex in this cycle is to conclude that frameworks are worthless and pure on-chain intuition wins. I don't buy it. The contrarian case is subtler: empty analysis and rigorous analysis can look identical from the outside, which is exactly why good frameworks still matter and bad ones survive. The problem was never the form. A hammer isn't wrong because you swung it at an empty nail.
And there's a decoupling argument worth confronting. Commentators insist crypto has finally decoupled from macro and now trades on its own narrative. My read is stranger: the research decoupled from reality, while prices stayed chained to the same forces they always were โ dollar liquidity, real yields, the M2 impulse. The forms drifted off into slide decks. The assets didn't budge. So the winning posture for 2025 isn't "no framework." It's a two-input framework: liquidity on one axis, cash flow on the other. Everything else โ the team score, the narrative heat, the nine boxes โ is decoration that flatters the allocator without informing him.
The blind spot almost every reviewer shares is that they audit the asset and never audit the audit. Nobody scores the research template itself. Had someone dropped that nine-dimension report onto the same form and asked "does this document have inputs?" the whole product would have scored N/A โ and that would have been the single most valuable line in forty pages.
So I sent the analyst a one-line reply: you built a perfect instrument and aimed it at nothing โ next time, find the object first. The framework isn't the enemy. The enemy is the bull-market reflex that mistakes the shape of diligence for diligence itself. In the next year, the reports that earn their keep won't have nine dimensions. They'll have three inputs, one decision they'd change, and the discipline to write "unknown" where the market has hidden its hand. So which dimensions are you counting โ and how many are just N/A with better formatting?
