In February, on a governance call I am not at liberty to name, someone screen-shared a diligence report. Nine sections. Nine green checkmarks. Technical architecture. Token economics. Market positioning. Ecosystem role. Regulatory posture. Team and governance. Risk matrix. Narrative durability. Supply-chain transmission. The project had closed a little over $100 million eleven weeks earlier, and the room โ maybe fourteen people, half of them allocators with mandates I would charitably call 'growth-oriented' โ exhaled together like a single organism.
I asked for the appendix.
Every field beneath every checkmark read the same two words: insufficient information. Not 'low risk.' Not 'pending review.' Nine dimensions, nine green lights, and underneath them, forty-one rows of N/A.
The room read it as a failure of the analyst. I read it as the only honest artifact produced in that entire quarter โ and, if I am being blunt, the only sentence in the deck that had not been priced for a bull market.
I have been the other kind of document. In 2017 I co-founded LibertyDAO, a community treasury that I described in three separate Medium posts as 'self-governing.' It was governed by a multisig with a threshold I had chosen because it felt decentralized rather than because I had modeled what happens when two signers stop answering Slack. When the treasury drained, it did not drain through an exploit. It drained through a governance model that reflected my aesthetics instead of my values. That distinction cost about $400,000 of other people's money and roughly two years of my confidence. I went back to school for a master's in computer science specifically to learn how to formally specify the thing I had failed to specify informally: the relationship between a code structure and the human behavior it silently mandates.
So when I see forty-one N/A fields under nine green checkmarks, I do not see an incomplete report. I see the only part of the process that was doing actual work.
Context: how the nine-box template became the industry's lingua franca
The template is not mysterious. It emerged because institutional capital โ post-ETF, post-MiCA, post-the-2024-real-world-asset wave โ needed diligence that could be compared across deals by people who were not going to read the code. That is a legitimate need. When I designed the governance wrapper for GlobalCommons, a tokenized real-world-asset fund, in 2024, half my job was producing an executive summary that a compliance officer in Frankfurt could sign without understanding what a merkle proof is. The other half was producing a community narrative that an on-chain delegate in Lisbon would actually believe. Those two documents had to describe the same system. That is the entire difficulty of this industry compressed into a staffing problem.
The nine-dimension framework solved the comparability problem elegantly. Technical, token, market, ecosystem, regulatory, team, risk, narrative, transmission โ nine axes, and suddenly a fund analyst could put forty projects into a grid and sort by score. Boards loved it. Listing committees loved it. It converted the unquantifiable into a spreadsheet, which is the oldest magic trick in finance.
And here is what the framework quietly did: it made an empty field look like a filled one. A checkmark is a visual claim of coverage. It does not distinguish between 'we examined this and found it acceptable' and 'we could not examine this because the project would not give us the sequencer logs.' Both render as green. The template's affordance โ nine boxes, begging to be completed โ is a stronger force than any analyst's integrity. A framework with nine slots does not measure diligence; it manufactures the appearance of it.
That is the trap. And in a market where the average raise has tripled and the average time-to-close has halved, the trap is load-bearing.
Core: what the empty fields actually correspond to
Let me walk the nine axes the way I would walk them in an audit, because the interesting thing about the empty report is not that it was empty โ it is which fields emptied first, and why those fields are always the same ones.
Technical. The question is never 'is it audited.' The question is what the audit's scope excludes. In my experience auditing mid-cap DeFi protocols, the excluded surface is nearly always the same three things: the upgrade proxy, the admin key, and the sequencer. An audit that covers the implementation contract but not the proxy through which every state transition passes is a document about a building that does not include the doors. A 3-of-5 multisig with no timelock is not a security assumption; it is a promise. And a rollup with a single sequencer operator has not inherited Ethereum's security โ it has rented Ethereum's branding. Trust isn't verified on-chain. It is verified in the diff between what the docs claim and what the bytecode permits.
This is also where the ZK arithmetic gets uncomfortable. Every team I have reviewed in the last eighteen months has a proving-cost line item that assumes verification demand scales linearly and hardware costs fall on a predictable curve. Neither is true. Proving a mid-sized batch on commodity GPU clusters runs into real dollars per batch once you amortize the hardware, the redundancy, and the operator margin โ and that cost does not care about your token price. Verification on L1 is cheap; proving is not. Operators are currently subsidized by a combination of token emissions and the assumption that gas returns to bull-market levels and stays there. When I model these businesses at 2023 gas prices, the margin goes negative in eleven of the fourteen architectures I have examined. That is not a bearish prediction. It is arithmetic that has been deferred, and deferred arithmetic is the most expensive kind.
Token economics. The supply table is where narrative meets the calendar. Team, early investors, community and liquidity, treasury and ecosystem fund โ four rows, and in the healthy version each row has a vesting curve with a cliff that is longer than the marketing cycle. In the version I usually see, the team allocation has a twelve-month cliff that lands within the same fiscal quarter as the first major narrative catalyst. That is not a coincidence and it is not necessarily malice; it is the natural equilibrium of a market where the raise closes before the product ships.
The number I care about most is not the fully diluted valuation. It is the ratio of emissions-funded yield to fee revenue. If a protocol is advertising 40% APR and 6% of that comes from actual user fees, you are not looking at a yield product. You are looking at a subsidized customer acquisition program with a token as the invoice. And the subsidized acquisition will stop โ it always stops โ and when it stops, the depositors who came for the 40% will leave at a speed that the withdrawal queue was not designed for. I learned this the expensive way in 2020 with EquiSwap, where I built a beautifully balanced pool and discovered that balance is not the same as demand. My pools were perfectly weighted and completely empty of organic flow.
There is a deeper structural point buried here. Even in the blue-chip lending markets โ the ones everyone treats as the mature end of the industry โ the interest rate curves are policy, not price discovery. The kink at 80% utilization, the base rate, the slope on either side of the kink: all of these are governance parameters chosen by vote and revised by committee, not outputs of a market-clearing process. They approximate supply and demand the way a thermostat approximates weather. That is fine as engineering. It is not fine as an argument, and the industry has been making the argument for four years.
| Allocation | Typical share | Cliff | The field that empties | |---|---|---|---| | Team | 15โ22% | 12 months | Whether cliff and catalyst coincide | | Early investors | 18โ30% | 6โ12 months | Whether the round price is disclosed | | Community / liquidity | 30โ45% | none | Whether 'community' includes market makers | | Treasury / ecosystem | 15โ25% | none | Who holds the keys and under what threshold |
Look at the rightmost column. Four rows, and every one of them terminates in a question that the project can simply decline to answer. That is why the token section empties first. Not because the data is hard to get โ because the data is easy to withhold.
Market. 'Priced in' is the most abused phrase in this industry, and the only honest way to test it is positioning. Funding rates, basis, and open interest tell you what the marginal leveraged participant believes. When funding is persistently positive and open interest is at a local high going into a catalyst, the catalyst has already been purchased by people who will need to sell something else to keep it. The empty field here is usually 'who is on the other side of this trade,' and the answer, uncomfortably often, is a market maker with a token loan.
Ecosystem position. Dependency graphs are where I find the most self-deception. A protocol is not its TVL; it is the set of things that break when it breaks. The upstream dependencies โ the oracle, the bridge, the sequencer, the stablecoin โ determine the blast radius. The downstream integrations determine whether a failure is an incident or a cascade. Developer signals are the least gameable metric available: contract deployments, unique deployers, retention of deployers across two quarters. Daily active users are trivially gameable and everyone knows it and everyone still puts them in the deck, because a number that can be inflated will be inflated, and a number that cannot be inflated rarely gets measured.
Regulatory. Howey is a four-prong test, not a vibe. Investment of money, common enterprise, expectation of profit, derived from the efforts of others. The prong that kills most token designs is the fourth, and the reason is structural: the moment a foundation employs a core team whose salaries come from a treasury whose value depends on the token, the 'efforts of others' prong is not a legal argument, it is an org chart. I have watched teams spend six figures on legal opinions that describe a network that does not yet exist and will not exist until after the token unlocks.
And then there is Europe. MiCA gave the continent apparent clarity, and clarity is genuinely valuable โ but the cost structure it imposed is a fixed cost, and fixed costs are regressive by design. Own-funds requirements under the CASP regime scale from โฌ50,000 to โฌ150,000 depending on service class, which sounds manageable until you add custody arrangements, governance requirements, complaint handling, and the ongoing reporting surface. For a consortium, that is a rounding error. For a twelve-person protocol team, it is the difference between a European market and a European footnote. Regulation rarely kills small projects by prohibiting them. It kills them by charging them the same admission fee it charges the incumbents. The empty regulatory field in that February report was not 'we don't know the rules.' It was 'we have not priced the compliance function, because pricing it would have changed the raise.'
Team and governance. Governance health is measurable and almost never measured. Voter participation as a percentage of eligible supply, top-10 delegate concentration, the ratio of proposals that reached execution versus proposals that were withdrawn after private negotiation. That last one is the tell. A DAO where nothing is ever withdrawn is not a DAO with consensus; it is a DAO with a back channel. LibertyDAO taught me that the governance model is the moral backbone of the protocol, and that a threshold chosen for aesthetics is a threshold chosen for failure. Code is law, but people are the soul โ and the soul is visible in the withdrawal rate of proposals, not the pass rate.
Risk and narrative. The risk matrix is the section most likely to be filled with adjectives, and the narrative section is the section most likely to be filled with adverbs. What I want instead is a gap table: market expectation versus actual delivery, on users, on revenue, on technical milestones, with a dated forecast and a dated outcome. The gap is the tradable object. Everything else is commentary.
Transmission. Finally, the propagation path. A single protocol event does not stay local. It moves to the bridge that custodied it, to the lending market that accepted it as collateral, to the exchange that listed the token, to the market maker that hedged it, and eventually to a TradFi counterparty that believed it had bought a diversified exposure. In 2022 that path was a curiosity. In 2026, with tokenized treasuries sitting inside institutional portfolios, it is the whole ballgame โ and it is the field most likely to be left blank, because the analyst who fills it in has to name firms.
Contrarian: the framework is the vulnerability
Here is the part that makes people uncomfortable, and it is not the part about empty fields.
Nine dimensions is a centralization vector. Whoever defines the axes controls the conclusion, because the axes determine what counts as evidence. A framework that weights 'narrative durability' as a scored dimension will systematically reward projects with good communications teams and penalize projects with good cryptography and bad decks. I have watched this happen at the listing-committee level: two projects, one with a formally verified core and a monosyllabic founder, one with a Notion page and a former-consultant head of growth. The second one scores higher. Not because the committee is stupid โ because the instrument is measuring the wrong thing and doing it consistently.
There is a second failure mode, and it is the one I am most guilty of. 'N/A' can become a performance of rigor. It is possible to produce a document so scrupulously hedged that it says nothing, offends no one, and protects the analyst completely. That is not honesty; that is risk transfer dressed as epistemology. The discipline is not to refuse to fill the box. The discipline is to distinguish three states that the template collapses into one: known, unknown but knowable, and unknowable. The first goes in the report. The second goes in the report with an owner and a deadline. The third goes in the report with a price โ because an unknowable risk is not an absence of information, it is a probability distribution, and refusing to assign one is itself an unhedged position.
A blank field is a signal, not a verdict. The February report was right to leave forty-one rows empty. It was wrong to leave them unpriced.
Takeaway
The next cycle's winners will not be the projects with the best nine-box scores. They will be the projects whose diligence documents have footnotes โ the ones where the sequencer is named, the unlock cliff is dated against the catalyst calendar, the compliance cost is a line item rather than an assumption, and the gap between narrative and delivery is disclosed before someone else discloses it for them. Decentralization is a verb, not a noun; it is the ongoing, expensive, unglamorous act of making the invisible visible on a schedule. The question I would put to every allocator reading this: when the next green-checkmarked report lands on your desk, will you ask for the appendix โ and if you find forty-one empty rows, will you read them as a reason to walk away, or as the only honest page in the room?