Hook
Last week I was handed a 41-page diligence deck on a mid-cap Layer2. Nine sections. Technical architecture, token supply, governance, regulatory posture, ecosystem dependencies. Every table had a header. Every header had a row. Every row said N/A.
Under team: N/A. Under unlock schedule: N/A. Under audit status: N/A. Under contract revenue: N/A. Under top-ten holder concentration: N/A.
The deck cost the fund that commissioned it roughly eleven thousand dollars, two analysts, and three weeks of calendar time. What came back was not a finding. It was a container built to hold a finding, delivered empty, and formatted so cleanly that nobody in the room noticed the vacuum.
I have done this work for twenty-one years. I know what real diligence looks like, and it looks like fingerprints โ messy, specific, and impossible to fake. An empty framework is a different animal. It is a liability dressed as a deliverable.
Context
The research layer of crypto has quietly become its own industry, and like most industries that sell to allocators, it optimizes for what can be invoiced rather than what can be verified. The standard product is a template: nine dimensions, forty pages, a valuation range, a disclaimer. The templates were originally built by serious people solving a real problem โ how do you compare a rollup to a lending market to a restaking layer in a single framework? Fair question. The answer they shipped was a grid.
A grid has an ugly property. It can be filled by looking anything up, or it can be filled by looking nothing up. The reader cannot tell the difference between a row that says N/A because the data doesn't exist and a row that says N/A because the analyst didn't look. Both render identically. Both bill identically.
I built my first diligence process in 2017 in Ho Chi Minh City, auditing ERC-20 contracts for a $12 million ICO called DragonCoin. I found an integer overflow in the distribution logic that would have let miners mint unlimited supply. I emailed the core team. They patched it before launch. The lesson was not that I was clever. The lesson was that the vulnerability existed in the code and nowhere else โ not in the whitepaper, not in the pitch deck, not in the forty-page grid the promoters were circulating. The only place the truth lived was in a function that had been compiled and deployed. I have never since trusted a document that wasn't a diff.
Core
Here is the mechanism, and it is worth naming precisely, because the incentive structure explains almost all of it.
Research desks are paid on coverage, not on accuracy. Coverage is countable: how many protocols did you write about this quarter. Accuracy is not countable, because nobody forces the desk to score its own calls eighteen months later. If your compensation depends on how many grids you filled, and filling a grid with N/A takes as long as filling it with numbers, you will fill it with N/A and buy back the time. Not out of malice. Out of arithmetic.
The empty report is not a research failure. It is a research product, priced and delivered correctly for the incentive that generated it.
The second-order effect is worse. Empty coverage enters the information market as a legitimacy signal. An allocator sees a 41-page deck with a fund's logo on the cover and reads it as diligence performed. The protocol, meanwhile, waves the same deck in its own marketing. The grid becomes a token. A token with no fundamentals behind it. You have seen this trade before.
I ran into the mechanical version of this in 2020. I wrote a Python script that watched Uniswap and SushiSwap pools for spread, and it executed a few hundred trades into a $45,000 profit during DeFi Summer. The script didn't have opinions. It had two numbers and a difference between them. What it taught me was that sentiment in this market is downstream of incentives โ the narrative moved from "store of value" to "yield farming" because the yield numbers moved first, and the story followed the arithmetic within about 72 hours. Arbitrage is just geometry disguised as finance. Narrative works the same way. It fills whatever vacuum the incentives open, and it fills it fast.
Which brings me back to the N/A grid. When a research product carries no verified facts, the protocol it covers is free to supply its own. The vacuum is not passive. A research gap is not an absence of information; it is an unfilled slot that someone with a marketing budget will fill on your behalf.
I watched this happen in May 2022 with TerraUSD. The on-chain data was screaming hours before the desks published anything coherent. Mint-and-burn ratios, the curve of UST supply against LUNA supply, the widening of the peg on the small venues first. The information was all there, publicly, and free. The coverage that arrived later was elaborate and correct in retrospect and utterly useless in the window that mattered. Narrative control preceded price action. It always does. The people who understood the mechanism were reading chain state, not decks.
This is why the empty report bothers me more than a wrong one. A wrong report generates disagreement, and disagreement generates scrutiny. An N/A report generates silence. It looks like rigor, it smells like rigor, and it carries no claims that can ever be falsified.
Contrarian
Here's the part that will make me unpopular with the desks that commission these things: in this market, the empty analysis is often the more honest artifact.
The populated version is not more informative. It is more confident. Give a junior analyst a grid and a deadline and a protocol with no verifiable revenue, and he will produce a valuation range anyway โ because the grid demands a number, and a number with a footnote reads as work. That number then travels. It gets quoted in a pitch. It gets aggregated into a sector average. It becomes a reference price for an asset that has never once produced a verifiable cash flow.
So which do you prefer? A page that says N/A, or a page that says $0.42 with an asterisk? Only one of them is lying, and it's the one that looks finished.
In a bear market this distinction stops being academic. When capital is cheap, a manufactured valuation costs you an opportunity. When capital is scarce, it costs you principal. The sectors that bled hardest in every drawdown I have watched โ 2018, 2022, the rolling de-pegs since โ were the ones with the most coverage and the least verifiable data. Volume of research and quality of research are inversely correlated in exactly the places you need them to be correlated. I don't confuse coverage with conviction.
Takeaway
The next eighteen months will produce more research reports than the last three years combined, because fear sells diligence. Most of them will be grids. Read the ones that have fingerprints in them โ commit hashes, timestamped wallet movements, a specific function name, a diff someone actually ran.
When you open the next 41-page deck, go straight to the footnotes. If every row resolves to N/A, you have not been handed a cautious report. You have been handed a blank check made out in your own name, and someone is waiting to fill in the amount.
The question worth sitting with is not whether the analysis was thorough. It is who was holding the pen when the grid stayed empty โ and whether they were waiting for you to look away.