Over the past seven days, a single line on a DefiLlama dashboard has been doing something that no token launch, no governance vote, and no developer conference has managed to accomplish in years: it has made the entire long tail of the blockchain industry visible at once. Seventy-one percent of the blockchains tracked by the platform generated exactly zero fees across a twenty-four-hour window. Not low fees. Not falling fees. Zero.
I have spent most of my adult life staring at fee dashboards — first as an economist studying money velocity, then as a fund manager watching on-chain revenue curves at three in the morning during the Terra collapse, and again last year translating ETF regulatory language for pension fund trustees who had never opened a wallet. A number like 71% does not shock me the way it shocks the timeline. What unsettles me is how few people understand what it actually measures. It is not a verdict on technology. It is not a death certificate for sixty-odd networks. It is something far more uncomfortable — a mirror held up to the demand side of an industry that spent a decade obsessing over supply.
So before this number gets weaponized, let me do what I always do with my community: slow down, define the terms, and then tell you what I actually think. Because the story here is not that seventy-one percent of chains are dead. The story is that we finally have a clean, chain-level instrument for measuring whether anyone is willing to pay to use a network — and the answer, for most of them, is no.
Context: What DefiLlama Actually Measures When It Says "Fees"
Let me start with the plumbing, because I have learned the hard way that most market panics are really just vocabulary failures. DefiLlama is an open-source data aggregator that has become, over the last several years, something close to the Bloomberg terminal of decentralized finance. It tracks total value locked, protocol revenue, and — the metric that matters here — fees. When the dashboard reports that a chain produced fees, it is reporting the sum of transaction costs that users actually paid to have their activity included in a block. This is not an estimate. It is not a projection. It is the on-chain record of money changing hands for the right to use scarce block space.
That distinction matters enormously, because fee revenue is the one number in this industry that cannot be faked by marketing. A project can buy followers, manufacture TVL through recursive lending loops, and rent liquidity with inflationary token emissions. What it cannot easily fake is a stranger voluntarily paying a fee to use the network. That is why I have argued for years that fees — not TVL, not developer counts, not Discord members — are the honest scoreboard. TVL is a stock that can be inflated with borrowed confidence. Fees are a flow that must be earned every single block.
The economic logic underneath is simple, and it comes from the same place as the economics of any congested resource. Block space is scarce by design. Every chain, whether it is Ethereum with its twelve-second slots or Solana with its sub-second finality, can only include a finite amount of data per unit of time. When more people want to use that space than the space can hold, they bid against each other, and the winning bids become fees. This is why the fee level of a chain is really a demand signal dressed up as a price. A high-fee chain is a chain where demand persistently exceeds supply. A zero-fee chain is a chain where demand has effectively vanished — or, more precisely, where nobody is willing to pay anything above the floor to get in.
Now, here is where the nuance begins, and where I want to be careful. Zero fees does not automatically mean zero activity. There are at least three technical reasons a chain can register zero fees while still processing transactions. First, the fees may be so close to zero that they round down — this is common on certain layer-2 networks and application-specific chains where the design goal was explicitly to make transactions feel free. Second, the fee may be subsidized by a foundation or an operator that covers user gas costs, which is a legitimate growth strategy and not a sign of failure. Third, the fees may be settled in a non-standard way that DefiLlama's tracking does not capture — a measurement gap rather than an economic void.
I raise these caveats not to soften the headline but to sharpen it. Because once you strip away the measurement noise, what remains is a structural fact that no amount of caveat can erase: the overwhelming majority of blockchains are not producing meaningful, self-sustaining economic activity. And that fact has consequences that ripple from the validator set all the way up to the pension funds I advised last year.
Core Insight: The 71% Is a Demand-Side Mirror, Not a Supply-Side Verdict
Here is the insight I want you to hold onto, and I will put it in bold because it is the spine of everything that follows: the zero-fee statistic is not measuring how many chains were built; it is measuring how few chains anyone actually needs.
We spent the last decade in a supply-side arms race. Every cycle brought a new wave of networks promising higher throughput, cheaper transactions, and more elegant consensus mechanisms. The implicit assumption — rarely stated, always present — was that if you built a better chain, users would come. Build the roads and the traffic will follow. That assumption was always a bet on demand that had not yet materialized, and the 71% figure is the bill for that bet coming due.
Think about what a chain actually is in economic terms. It is a marketplace for block space. Like any marketplace, it needs two sides to function: sellers of space (validators and node operators who secure the network) and buyers of space (users and applications that need to transact). The sellers exist in abundance because launching a chain has never been cheaper — rollup-as-a-service frameworks, modular settlement layers, and shared security protocols have driven the marginal cost of spinning up a new network toward zero. But the buyers are finite, and they cluster. They cluster where liquidity is deepest, where the tooling is most mature, where the wallets already work, and where the other buyers already are.
This is the cruel arithmetic of network effects: a marketplace with many sellers and few buyers does not produce cheap transactions — it produces no transactions at all, because the buyers never arrive. The long tail of chains is not competing on price. It is competing on relevance, and most of them are losing.
I have seen this movie before, and it is why one of my signatures has always been that history repeats, but liquidity decides the tempo. In 2017, I watched hundreds of utility tokens launch with elegant economic models and no users. The tokens did not fail because the models were mathematically wrong. They failed because the models assumed a demand that the founders had imagined rather than observed. The blockchains registering zero fees today are the direct descendants of those tokens. The technology improved. The demand did not.
What makes this cycle different is that we can now see the failure in real time, on a public dashboard, denominated in a unit that no one can spin. During the ICO era, you had to read whitepapers and join Telegram groups to sense which projects were hollow. Today, you open a browser tab and the answer is sitting there in a column of zeros. Transparency is a gift, but it is also a mirror, and mirrors are unkind.
Let me quantify the concentration, because the aggregate 71% hides an even more extreme reality. When DefiLlama reports that fee generation is highly concentrated, it is understating the case. A handful of networks — Ethereum and its scaling ecosystem, Tron, Solana, and BNB Chain — account for the overwhelming majority of all fees paid across the entire industry. Tron alone, largely on the back of stablecoin transfer demand, routinely processes fee volume that would embarrass networks with ten times its mindshare. Solana's fee curve has steepened dramatically as consumer applications found product-market fit. Ethereum remains the settlement layer of choice for anything that values credible neutrality. And beneath those four, the drop-off is not a slope. It is a cliff.
The pyramid of public blockchains is not a pyramid of technology. It is a pyramid of demand, and the base is almost entirely empty.
The Death Spiral Is Not a Metaphor — It Is a Mechanism
I want to walk you through the mechanism, because I think the phrase "death spiral" gets thrown around so casually that people stop believing it is real. It is very real, and it operates with the cold logic of a bank run.
Start with a chain that has low but non-zero activity. Its validators and node operators are compensated in two ways: fees from users, and inflationary token rewards. When activity is thin, fees are negligible, so the security budget rests almost entirely on inflation. That inflation is paid in the chain's native token, which means the real compensation of the validators depends on the token's market price. Now suppose the token price falls — perhaps because the broader market is in a drawdown, perhaps because the lack of real usage is finally reflected in the valuation. The validators' real income falls with it. Some validators, facing unprofitable operations, shut down. The network's security and liveness degrade. Applications and users, sensing fragility, migrate to a healthier chain. Activity falls further, fees fall toward zero, and the token's fundamental case weakens again. The loop closes on itself.
This is why I take the zero-fee figure more seriously than the casual observer. A chain at zero fees is not merely a chain with no revenue. It is a chain whose security budget is entirely dependent on the continued willingness of the market to subsidize it through token inflation. That is a subsidy-dependent structure, and subsidy-dependent structures have a well-documented failure mode: they work beautifully until sentiment turns, and then they work not at all.
I watched this dynamic up close during the 2022 bear market. When Terra and Luna collapsed, I made a decision that some of my peers found strange: instead of immediately liquidating and going quiet, I started a weekly transparency series for my subscribers, detailing exactly what our fund was exposed to and how we were hedging. I published the uncomfortable numbers. I told people where we had been wrong. And what happened was the opposite of panic — we retained roughly eighty-five percent of our capital, because the community viewed honesty as a stabilizing anchor. That experience taught me something that applies directly here. Markets do not fail because people lack information. They fail because people lack trust in the information they have. The zero-fee chains have a data problem and a trust problem at the same time, and the second is worse than the first.
A validator on a zero-fee chain is, in effect, holding a claim on future token appreciation to compensate for present service. As long as the token narrative holds, the validator stays. The moment the narrative breaks, the validator leaves, and there is nothing underneath to catch the fall. This is not a hypothetical. It is the structural condition of every long-tail network on that dashboard.
Fee Concentration as a Network Effect Moat
Let me now explain why the concentration is not a temporary imbalance that competition will correct, but a durable feature of the landscape.
Network effects in crypto operate through three reinforcing loops, and each one deepens the moat around the leading chains. The first is the liquidity loop. Deeper liquidity attracts larger traders, larger traders generate more volume, more volume deepens liquidity. This loop lives on the leading chains because that is where the order books, the automated market makers, and the bridges already are. A new chain can offer lower fees, but lower fees mean nothing if there is no counterparty on the other side of the trade.
The second is the tooling loop. Developers build where the libraries, the SDKs, the indexers, and the wallets already work. This is a form of path dependency that economists recognize from every technology adoption curve in history. The tooling on Ethereum and its major layer-2s is mature enough that a competent developer can ship a product in weeks. On a long-tail chain, the same developer spends those weeks fighting infrastructure gaps. The time cost of building on the periphery is a tax that no fee discount can offset.
The third is the trust loop. Users, especially the institutional users I advised during the ETF process, gravitate toward networks with a track record, a credible security budget, and a recognizable governance structure. Trust compounds. It also concentrates. And this is where I want to bring in the second of my guiding ideas: culture is the code that compels human adoption. The leading chains are not winning purely on technical merit. They are winning because they accumulated a culture — a shared set of expectations, habits, and norms — that makes people feel safe transacting there. That culture is not in the whitepaper. It is in the muscle memory of millions of users who reach for the same wallet, the same bridge, the same explorer every time.
A long-tail chain cannot buy that culture with a grant program. It cannot mint it with a token incentive. Culture is the slowest-moving asset in this industry, and it is the one that decides which chains still have users when the incentives run dry.
The competitive implication is stark, and it connects to something I noticed during the 2020 DeFi Summer. Back then, I directed a portion of our fund into Aave and Compound liquidity pools, and the thing that determined whether a pool retained capital was not the advertised yield — it was the user journey. When the interface was frictionless, capital stayed. When users hit a confusing approval flow or an unclear risk parameter, they pulled out and went somewhere friendlier. I coordinated with product teams to smooth those friction points precisely because I understood that UX is not a cosmetic concern. It is a capital-retention mechanism. The same principle applies at the chain level. A chain that makes it hard to transact will lose its users to a chain that makes it easy, regardless of how elegant its consensus algorithm is.
The blockchains at zero fees are, in most cases, chains that never solved the user journey. They offered a theoretical alternative to the incumbents without offering a practical reason to switch.
The Ghost Chain Phenomenon and the Subsidy Economy
Now I want to name the thing that the 71% figure is really pointing at, because it has a name in our industry and we should use it honestly: the ghost chain.
A ghost chain is a network that raised capital at a high valuation, launched with fanfare, distributed a token, and then produced almost no economic activity. It has validators, because the token incentives pay them. It has a website, because someone built one. It may even have a foundation and a grants program. What it does not have is users who pay. The ghost chain is the physical manifestation of the gap between the narrative that funded it and the demand that never showed up.
I first started thinking seriously about ghost chains during my NFT work in 2021. I managed a portfolio that invested in generative art projects, and I deliberately sought out female digital artists and curated collections around community ownership rather than speculation. The lesson I took from that period was that cultural utility — the sense that owning something connects you to a group of people who care — is a genuine driver of value, not a soft one. But cultural utility only works when there is an actual culture. A ghost chain has no culture. It has a Discord that went quiet six months after launch. Value in this industry is either cultural or it is temporary, and ghost chains are the proof.
The subsidy economy that sustains these chains deserves a hard look, because it explains why the 71% figure persists rather than self-correcting. Many zero-fee chains are not merely idle; they are actively subsidized. Their foundations or treasuries pay for user gas, or their token emissions fund liquidity mining programs designed to manufacture the appearance of activity. This is not always cynical. Some of it is a legitimate customer-acquisition cost, the crypto equivalent of a startup eating the cost of shipping to win market share. But a subsidy is only rational if it eventually converts into organic demand. When the subsidy ends and the fees remain at zero, the experiment has failed, and the failure was disguised by the subsidy for as long as the treasury held out.
Here is the uncomfortable part for anyone holding long-tail tokens. If a chain's only economic activity is subsidized, then the token's value is not backed by cash flow. It is backed by the hope that the subsidy will one day be replaced by demand. That is a narrative, not a business. And narratives have a way of evaporating precisely when you need them most.
Real Yield and the Repricing of the Entire Sector
The 71% figure did not appear in a vacuum. It arrived at a moment when the market's dominant aesthetic had already shifted toward what we call real yield — the demand that a protocol's returns come from genuine, verifiable revenue rather than from token emissions. This is the single most important narrative shift of the current cycle, and the zero-fee data is its sharpest supporting evidence.
For most of crypto's history, the market rewarded growth of any kind. Users, volume, TVL — all were treated as signs of health, regardless of whether they generated cash flow. The bear market that followed 2022 broke that habit. When the easy money left, capital demanded proof that a network could earn its keep. And what the market discovered was that very few networks could.
The shift is not merely sentiment. It is the maturation of how the market prices risk. A chain with real fee revenue has a defensible valuation anchor: the present value of its fee stream. A chain with zero fees has no anchor at all, which means its valuation is a pure function of narrative momentum. When you strip the narrative away, you are left with the token's cost of production — the electricity, the hardware, the opportunity cost — and nothing else. That is a brutal repricing, and it is happening quietly across the long tail right now.
I saw a version of this during the institutional ETF process last year. When I was drafting policy briefs for traditional finance executives, the question that came up again and again was not about technology. It was about cash flow. Pension trustees who had spent decades evaluating equities and bonds wanted to know where the revenue came from, who paid it, and whether it was durable. The Bitcoin ETF structure we helped shape addressed those concerns for a single asset. But for the broader market, the implication is unavoidable: institutional capital does not chase narratives. It chases demonstrable economic activity. And by that standard, the 71% of chains producing no fees are not merely unattractive — they are invisible.
The real-yield aesthetic is also why I think the zero-fee statistic will have a longer tail of consequences than a typical news cycle. It is not a one-day story. It is a confirmation of the framework that sophisticated capital has already adopted, and confirmations of that kind get incorporated into valuation models permanently.
The Denominator Problem: Reading the Data Honestly
Now let me do what I always insist on doing with my community — challenge the number itself, because intellectual honesty is the only real edge in a market this noisy.
The 71% figure has a denominator problem, and it is significant. DefiLlama's list of tracked blockchains is not a curated roster of serious networks. It includes chains that are abandoned, chains that are testnets or remnants of testnets, chains that were announced and never really launched, and chains that exist only as a ticker in some aggregator's database. If a meaningful fraction of the denominator consists of what I would call zombie chains — networks that no reasonable person would describe as live competitors — then the 71% is systematically inflated. The headline would still be striking at, say, 55% or 60%, but it would be a different story.
There is also a volatility problem. A twenty-four-hour snapshot is a thin slice of time. A single airdrop, a single game launch, a single fee spike on one chain can move the aggregate. The 71% is not a stable constant; it is a reading from one moment in a market that breathes. Any conclusion that treats it as a permanent law of nature is overfitting to a single frame.
And there is the definitional problem I raised earlier. "Zero fees" conflates two very different conditions: genuinely dead networks, and networks that have deliberately engineered fees to near-zero as a growth strategy. A young layer-2 that subsidizes gas to win users is not the same as a ghost chain that never had users to begin with. Lumping them together flatters the pessimists and obscures the real distinction.
So let me be precise about what I actually believe. I do not believe that seventy-one percent of blockchains are dead. I believe that a large and growing share of blockchains have no sustainable economic model, that the true figure is probably lower than seventy-one but still alarming, and that the trend line — whatever the exact number — points in one direction. The direction is what matters. The exact reading is a detail that the media will get wrong and that serious analysts should treat with caution.
This is why, when I audit anything in this space, I refuse to work from a single number. During the 2017 ICO cycle, I built my reputation not by reading code but by reading communities — joining the Telegram groups, sensing the anxiety around vesting schedules, and organizing town halls to explain the economic model to retail investors before panic could set in. That work taught me that the raw number is almost never the truth. The truth is in the context around the number, and context is precisely what a headline strips away.
The L2 and Appchain Paradox: When Zero Fees Is a Feature
There is a dimension of this story that the pessimistic reading misses entirely, and it is the one I find most intellectually interesting. A large portion of the zero-fee chains are not failed experiments at all. They are layer-2 networks and application-specific chains that were designed to have near-zero fees, and their fee level tells us almost nothing about their health.
The economics of the layer-2 model are fundamentally different from those of a layer-1. A layer-2 does not need to generate fees to survive; it needs to settle to a layer-1 that does. Its users are paying fees to the base layer in aggregate, even if the individual transaction costs on the layer-2 round to zero. When I look at the rollup landscape, what I see is not a set of dying chains but a set of chains whose fee economics are deliberately deferred.
And this is where I want to place a longer-term marker, because I have a specific view on how this plays out. The current era of ultra-cheap layer-2 transactions rests on a subsidy of its own — the cheap blob space introduced by the Dencun upgrade. That blob space is a finite resource, and it is being consumed at a rate that I expect to saturate the available capacity within roughly two years. When it does, the cost of posting data to the base layer will rise, and rollups will have to pass that cost through to users. The fees on rollups will not stay near zero forever; they will double again once the blob subsidy exhausts itself. The zero-fee reading we see today is, for many layer-2s, a temporary artifact of an underpriced resource, not a permanent property.
I should be clear that this is a structural view, not a prediction of a specific date. But the logic is sound: you cannot permanently consume a scarce resource below its cost. Something has to give, and what gives is the fee. When that happens, the layer-2s that have built genuine demand will survive the repricing, and the ones that have built only subsidy-driven activity will not. The blob saturation will, in effect, be a stress test — a moment when the layer-2 field separates the economically real from the economically hollow.
There is a parallel here to the application-chain boom. Rollup-as-a-service frameworks have made it trivially easy to launch a chain, which is wonderful for experimentation and terrible for signal. When anyone can deploy a network in an afternoon, the mere existence of a network stops meaning anything. The fee data is valuable precisely because it cuts through the noise — it does not care how easy the chain was to launch. It only cares whether someone paid to use it. The appchain that subsidizes its way to zero fees is, from the data's perspective, indistinguishable from the ghost chain. Only time and the exhaustion of subsidies will tell them apart.
Transmission: How Zero Fees Ripple Through the Stack
A statistic this broad does not stay contained. It transmits through every layer of the industry, and I want to trace those pathways because they are where the real portfolio implications live.
Start at the bottom, with the node operators and validators. On a zero-fee chain, the security budget is entirely inflationary, which means the operators are, in effect, betting on the token. When the token falls, they exit. When enough of them exit, the chain's liveness is at risk. This is the mechanism by which a fee drought becomes a security problem. For the infrastructure operators, the long-tail chains are a diminishing business. The rational move is to concentrate on the chains with real fee revenue, which accelerates the consolidation.
Move up to the exchanges. Listing decisions are shifting. Exchanges have always cared about volume and liquidity, but the real-yield era is teaching them to care about fee generation and genuine user activity as listing criteria. A long-tail token with no on-chain fees and no organic users becomes harder to justify listing, harder to market, and harder to retain. The delisting risk is not theoretical; it is the natural end state of the repricing.
Then there is the infrastructure layer — the RPC providers, the indexers, the block explorers, the wallet teams. These are businesses, and businesses allocate resources where the demand is. A zero-fee chain does not generate the query volume that justifies running infrastructure for it. The result is a quiet withdrawal of support, which makes the chain even less usable, which drives the few remaining users away. The infrastructure layer amplifies the network effect rather than counteracting it.
For DeFi protocols, the impact is more muted, simply because DeFi was already concentrated on the leading chains. The protocols that matter were never deployed on the long tail to begin with. But there is a subtle effect at the margins: the liquidity that once trickled to the periphery will retreat entirely, further deepening the centralization.
For NFT and gaming projects, the transmission is more painful. Both depend on vibrant, low-cost environments with active communities, and both were disproportionately built on chains that now show zero fees. The projects are not necessarily bad. Their host chains simply never had the users to sustain them. This is a reminder that where you build is a strategic decision, not a technical one.
And for traditional finance, the effect is indirect but real. Real-world asset tokenization is consolidating on the compliant, high-fee, institutionally trusted chains, because that is where the counterparties and the regulatory comfort are. The zero-fee long tail is invisible to institutional capital, and it will remain so.
The Contrarian Angle: Decoupling Zero Fees From Zero Value
Now let me offer the counter-intuitive reading, because I have been building toward it and I do not want to leave you with a one-sided picture.
The consensus interpretation of the 71% figure is that it proves the public-chain sector is overbuilt and that most networks are worthless. I think that interpretation is half right and half dangerously wrong, and the distinction matters for anyone allocating capital.
The half that is right: the sector is overbuilt on the supply side, and most networks do lack a sustainable economic model. That is a real and durable signal, and I would not fight it.
The half that is wrong: it treats zero fees as a permanent verdict rather than a point in time, and it ignores the deliberate strategies that produce zero fees on purpose. Some of the most interesting networks on that dashboard are zero-fee because they are early, not because they are dead. A chain that has just launched, that is subsidizing adoption, and that is building toward a demand base is not a ghost chain. It is a startup. And conflating startups with corpses is exactly the kind of error that leads investors to miss the next generation of infrastructure.
Here is the deeper contrarian point. The decoupling of fees from value is real, but it cuts in a direction most people have not considered. It is not that zero-fee chains are secretly valuable. It is that the fee metric, while the most honest available, is still an incomplete measure of a network's strategic position. A chain can have low fees and high strategic value if it is the indispensable settlement layer for something larger. A chain can have high fees and low strategic value if its fees come from a single, fragile demand source. The fee data tells you where the money is today. It does not tell you where the money will be tomorrow, and conflating the two is the classic mistake of extrapolating a single metric into a worldview.
This is why I keep coming back to culture as the deeper signal. Fees are a lagging indicator of demand. Culture is a leading indicator. The chains that will still matter in five years are not necessarily the ones with the highest fees today. They are the ones whose users would not leave even if a cheaper alternative appeared. That is a much harder thing to measure on a dashboard, and it is the thing I spend most of my analytical energy trying to read.
I learned this the hard way. In 2021, when I curated those generative art collections, the projects that held their value through the hype cycle were not the ones with the highest short-term trading volume. They were the ones whose collectors felt a genuine connection to the work and to each other. The volume faded. The culture held. The same logic applies to chains, and it is why I refuse to reduce the entire industry to a single column of fee numbers.
Takeaway: Positioning for a Market That Has Stopped Rewarding Supply
So where does this leave us, in a market that is chopping sideways, that has stopped rewarding the launch of new things, and that has started asking the oldest question in finance — where is the money coming from?
I think the answer is that we are in the middle of a slow, structural repricing that will not finish this cycle. The public-chain sector is separating into two tiers: a small group of networks with durable demand and real fee revenue, and a long tail that will either find a niche or fade. The 71% figure is not the end of that process. It is a snapshot of it, taken from a dashboard that has accidentally become the most honest scoreboard in the industry.
For my community, I would frame it this way. The chop we are living through is not a waiting room before the next bull run. It is a sorting mechanism. The market is using the quiet to decide which chains deserve to survive the next cycle, and the instrument it is using is the fee. If you are allocating capital, the question is no longer which chain has the best technology. It is which chain has users who will keep paying when the subsidies stop and the incentives end.
And here is the thing I want to leave you with, because it is the question I ask myself every time I look at a dashboard full of zeros. If the demand is not there yet, is it because the technology failed — or because the demand was never going to exist, and the last decade of building was a bet on a future that the market has quietly declined to fund? That is the question the 71% is really asking. And unlike the number itself, it is not one that a twenty-four-hour snapshot can answer.
History repeats, but liquidity decides the tempo. And right now, the tempo of most of this industry is silence.
