On August 4, six researchers published a draft that has detonated quietly across Ethereum's economic establishment. The document has no EIP number, no reference implementation, no audit; it is barely old enough to have been read by more than a few thousand people. Yet it has already produced public opposition from the founder of Aave, an existential warning from the CEO of ether.fi, and a DeFi community described in the analysis as openly hostile.
The proposal is staggeringly simple to describe. Remove issuance incentives for staking above 50% of ETH. Burn a percentage of validators' "idealized rewards" at each epoch boundary. Set the burn to reach 100% when total staked ETH hits 60,250,000. Watch net consensus issuance fall to zero at that point. The staking yield curve, which currently rises monotonically and then flattens, becomes an inverted U: issuance peaks at approximately 19.8% of supply staked, then declines.
At the current staking ratio of roughly 28%, this would cut the net consensus yield from roughly 2.6% annualized to roughly 1.2%. Half. I have been sitting with that number for days, and it has not stopped feeling like a blow.
We assumed a certain trajectory for Ethereum's monetary world. More staking meant more security, more security meant more legitimacy, more legitimacy meant more yield. The curve, we told ourselves, was a law of nature. Nature, it turns out, is up for revision.
Context: The Subsidy We Stopped Questioning
To understand why this proposal matters, you have to understand how the subsidy became sacred.
The Merge of September 2022 refounded Ethereum both technically and morally. Proof-of-work was abandoned not simply because it was inefficient but because its machine-age imagery had become dissonant with the emerging story of Ethereum as a decentralized settlement layer for an open financial system. Proof-of-stake offered something it claimed was superior: security purchased through stake, not burnt energy. The issuance schedule designed for that transition was deliberately generous. A base reward factor of 64 calibrated the flow of new ETH to validators, and the whole construction was understood not as a policy choice but as a protocol axiom โ the way that Bitcoin's 21 million supply cap is treated as a constitutional fact.
The promise was a virtuous circle. Staking grows; security grows; economic weight grows. And for the first years after the transition, the circle turned. The staking ratio has now climbed beyond 28% of supply, a figure that would have been unthinkable at the launch of the beacon chain. Lido, the largest liquid staking protocol, sits atop the validator distribution like an engineered gravitational center. The restaking ecosystem has wrapped itself around staked ETH as the universal collateral of shared security. The Ethereum yield curve has become the floor of DeFi's credit structure โ the risk-free rate of the decentralized world.
Yet a minority of researchers โ with a lineage stretching back through "endgame" discussions and the "minimal issuance" school of thought โ has always viewed the curve with suspicion. Their intuition: a security budget that scales with staking but not with security need is a transfer mechanism, dressed in safe colors. The question is when the pressure stored in that mismatch would be released.
The answer, it turns out, is now.
I can date my own education on this subject precisely. In the aftermath of DeFi Summer, as a university student, I audited the governance mechanics of Curve Finance with a depth that bordered on obsession. Four hundred thousand lines of simulation data taught me something that has stayed with me: incentive structures have gravity. No matter how democratic the surface, the underlying token distribution will bend governance toward its thickest concentration. What I did not anticipate, until this proposal arrived, was that the same principle applies to the protocol layer itself. The distribution of staked ETH, the rewards associated with it, and the issuance curve that feeds it are not neutral plumbing. They are the deepest argument about who Ethereum belongs to.
That argument is now being made public.
Core Part I: The Mechanism, Precisely
Let me be exact about the proposal's mechanics, because the public conversation has already begun to mangle them.
Ethereum's proof-of-stake protocol divides time into slots and epochs. Within each epoch, validators perform attestations โ cryptographic endorsements of the consensus state โ and are rewarded on a schedule determined by the base reward formula. The formula accounts for the total active stake, giving each validator a reward inversely proportional to the square root of network stake, multiplied by the base reward factor. This known construction yields a concave, upward-sloping issuance curve: as staking grows, each marginal validator earns less, but total issuance continues to increase.
The draft introduces a single new operation: a burn at the epoch boundary. Before a validator's idealized reward โ the full amount it would earn under the current curve โ is credited, a percentage is destroyed. The burn percentage is a function of total staked ETH. It is nil in the lower staking regime; it rises as the staking ratio approaches the threshold; and it reaches 100% at 60,250,000 ETH, which corresponds to approximately half of current supply.
The consequences are immediate and structural. Beyond the threshold, the protocol still demands the validator's participation. It still imposes slashing conditions. It just does not pay new issuance for that participation. Validator income from the consensus layer, in net terms, becomes zero. The services matter; the subsidy is gone.
The transition design deserves attention: the draft proposes to double the base reward factor from 64 to 128, then allow it to decay back to 64 over roughly eighteen months. At first glance, this is a shock absorber: it creates a period of elevated issuance before the burn curve fully dominates. But look closer. The doubling of the base reward factor inflates the amount at risk of being burned; it does not change the shape of the cliff at the threshold. The behavior shift at the margin โ enter the 50%+ staking zone and exchange all future issuance for nothing โ is not gradual. It is a threshold crossing that validators will experience on the first epoch after activation.
The resulting issuance curve is an inverted U. Below the peak, at about 19.8% of supply staked, the curve remains upward-sloping. Above it, the curve turns downward. At 50% staked, net issuance crosses zero. Ethereum becomes, in a meaningful sense, an asset with a negative supply response to deeper staking.
No major L1 has attempted this. The innovation is real. The question is whether real innovation can survive the gravity of the ecosystem it was designed to change.
Core Part II: The Redistribution
Let me now make the economic structure of the proposal unmistakable.
At current levels, the proposal would compress the consensus yield from roughly 2.6% to roughly 1.2%. This is not a marginal adjustment; it is a halving of the protocol's direct compensation to validators. Execution-layer fees and MEV remain untouched, so the total income of a validator falls by a smaller magnitude. But the composition of income changes more importantly than its absolute level. The stable, protocol-guaranteed component โ the part that made validation a predictable service โ contracts sharply. The volatile, competitive, market-driven component becomes the primary source of revenue.
The redistribution is the mirror image. Non-staking ETH holders receive a quiet but real benefit: reduced dilution. If issuance is curtailed, the supply growth that would otherwise flow to validators is destroyed, and the relative purchasing power of existing held ETH increases. The proposal is, in effect, a transfer of the "issuance tax" that non-stakers currently pay, redirecting it back to them. Validators lose; holders gain; the protocol loses a tool for funding participation.
I write "holders" with a particular emphasis, because the class of non-staking ETH holders is far larger than the class of validators. This may be why the proposal has a genuine populist appeal in some circles: it is an anti-oligarchic gesture, framed in the language of monetary discipline. In a cycle where the market has been sideways and the search for yield has become an increasingly competitive zero-sum game, the proposal offers something the staking economy's critics have long demanded โ an end to inflation that accrues to the already-large.
But the keyword is "redistribution," not "reform." Redistributions have winners and losers, and the losers are not abstract quantities. Every percentage point of consensus yield removed from the base rate is a direct reduction in the income of thousands of individual validators โ people who bought hardware, provisioned servers, and committed to the network's security. Their costs are largely fixed. Their break-even math was written before this proposal. The word that best describes the effect of a 54% yield cut on that population is not "correction." It is "confiscation."
That asymmetry โ a wealthier future for the coin versus a sudden income shock for its security workforce โ runs through everything I have read about this document. It is the contradiction at the heart of the proposal. The authors want Ethereum to be harder and scarcer; the consequence is a protocol that punishes its own operational class. The code is law, but the humans are the bug.
Core Part III: Water Through Roots
I have seen how these transmissions work from the inside. Designing quadratic voting systems for community treasuries taught me a cardinal rule: you cannot change the risk-free rate of a system without repricing every instrument derived from it.
Liquid staking tokens are the most direct derivative of the staking yield. When the consensus yield falls by more than half, the base yield of stETH and weETH falls proportionally โ all else being equal. The liquid staking protocol's fee remains, and its margin remains, but the underlying asset's income stream is thinner. A fall in LST yield is not a niche event. It is repricing across the whole credit stack.
Consider the lending protocols. Aave uses stETH as collateral. The demand for borrowing against ETH is structured, in part, around the yield that collateral earns. When stETH yields drop sharply, collateral becomes less attractive, and the borrowing market's equilibrium shifts: lower utilization, lower rates, less leverage. Stani Kulechov's objection is exactly this in microcosm โ it is the fear of a liquidity vacuum at the center of the DeFi economy. He is not defending a particular business model. He is defending the stability of a pricing architecture.
Restaking compounds the problem. EigenLayer and its successors have built a market in which staked ETH simultaneously secures the base chain and external applications. The revenue to restaked validators is the same base yield, plus additional AVS rewards. Cut the base yield, and the economics of restaking change materially. The same ETH that secured one network now offers a lower baseline reward for securing several. The consequence is a repricing of AVS risk, a reduction in the size of the market for shared security, and a potential flight of capital away from restaking toward simpler forms of yield.
The ether.fi CEO raised a different but equally crucial concern: solo stakers. The economics of solo staking are already marginal. A home operator covering hardware, electricity, and the opportunity cost of locked ETH leans heavily on the deterministic income from consensus issuance. Cut that income by half, and the hobbyist ceases to be a viable participant. The survivors are those who can access institutional capital โ or who can capture outsized MEV.
I want to underline that this is not an accidental side effect. In a protocol whose stated values include decentralization and permissionless participation, the design of a low-issuance burn curve has a direct, predictable effect on participant diversity. The proposal, if enacted as written, will not simply reduce staking; it will change who is able to stake.
Core Part IV: The Governance Abyss
The politics of the proposal are as significant as its mechanics.
The draft was published two days before the Hegota upgrade's EIP submission deadline of August 6. Deadlines in Ethereum governance are more than formalities; they are the pulse of protocol-level decision-making. Publishing a controversial, unvalidated economic restructuring into that window is either a demonstration of questionable strategic judgment, or a calculated attempt to force attention in a moment where it could not be ignored.
Either interpretation should trouble anyone who cares about Ethereum's governance culture.
What shapes Ethereum's future is not an elected parliament. It is a decentralized assembly of researchers, core developers, application builders, and stakeholders who coordinate through rough consensus in call logs, GitHub issues, and corridor conversations. In such a system, legitimacy depends not merely on correctness but on process. A proposal that arrives unvalidated, unaudited, and without a reference implementation โ at a deadline โ has placed enormous weight on its authors' authority. If that weight is sufficient to carry it into a formal EIP process, governance has been exercised through reputation rather than through evidence.
The community has responded accordingly. The DeFi sector is reported to be hostile, not merely skeptical. Aave's founder opposed it publicly, calling the direction harmful to Ethereum, and the CEO of ether.fi has warned of dire consequences for individual stakers. The speed and intensity of the outrage suggest the proposal struck a nerve far deeper than the standard policy disagreement. It has surfaced a chronic tension that has been accumulating since the Merge: the research elite's vision of an optimal protocol versus the economic interests of the ecosystem's operational layer.
I think of this as a structural fault line. Ethereum has historically been able to accommodate discord by deferring major changes and allowing technical merits to be debated over long horizons. The proposal's compressed timeline makes that accommodation impossible. It forces the protocol's governance into its most adversarial posture, and then asks the community to deliberate under deadline. Silence is the only consensus that never forks โ but the draft was crafted to maximize the volume of speech, not silence.
The presence of core developers among the authors adds a further complication. In my years of observing this governance space, "internal credibility" has always cut both ways. It buys attention, but it also converts a technical proposal into a factional marker. The proposal is already being read as the research wing's campaign against the economic establishment. That framing will outlive whatever technical fate the draft experiences.
Core Part V: The MEV Spiral
Here is the darkest prediction I can make about the proposal's long-run effects, and I want to be explicit about my uncertainty: the burn curve's interaction with MEV could produce a concentration spiral the authors did not model.
The reasoning is straightforward. With consensus issuance approaching zero, the dominant income source for validators becomes execution-layer rewards: transaction priority fees and MEV. MEV is not distributed uniformly; it is a winner-take-most market. Sophisticated block builders with exclusive order flow and high-frequency infrastructure capture a disproportionate share of extractable value. Small validators, who currently meet their break-even partly through deterministic issuance, would depend increasingly on a lottery-structured income source that systematically favors the largest operators.
This is not a hypothetical. The post-Merge data show that MEV capture is highly concentrated, and that solo stakers, when they earn MEV at all, capture a tiny fraction of the available surplus. A shift in the income mix toward MEV therefore implies a shift in the validator population toward entities that can successfully compete in the MEV market. The consequence is a slow, steady erosion of validator diversity โ the exact property that the staking subsidy was originally intended to purchase.
The feedback loop is vicious. Lower issuance โ greater dependence on MEV โ concentration of MEV capture โ concentration of validator set โ reduced decentralization โ increased vulnerability to coordinated attacks and censorship. The proposal, by eliminating a broad-based subsidy while leaving MEV untouched, is functionally a regressive transfer from the tail of the validator distribution to its head.
The advocates of minimal issuance would respond that MEV was always the point โ that security should be paid for by the users who extract value from the chain, not by inflation. It is a coherent position. But it fails to account for the market structure of MEV. A tax on users of the chain is acceptable; a subsidy to the largest block builders is not. The proposal does not distinguish between the two.
We built a kingdom of ghosts in the machine: validators who reveal themselves each epoch, attestation by attestation. The ghosts are losing their income; the machine keeps spinning. Something will be quietly redistributed.
Core Part VI: Regulators in the Mirror
One of the least explored implications of the proposal is its regulatory resonance. In the Howey-test analysis of staking, the "expectation of profits" element has always carried heavy weight. The SEC's arguments in enforcement actions against staking products have leaned on the income that stakers expect. The proposal complicates that analysis.
If the protocol is configured to burn the issuance component of validator rewards, and if validator compensation shifts toward transaction fees and MEV โ payments for operational services โ then staking begins to resemble a service contract rather than a profit-seeking investment. A validator operator is paid for performing work, not for participating in a joint profit-sharing enterprise. That characterization does not conclusively change the legal outcome, but it changes the vocabulary of the debate.
I want to be honest about the limits of this observation. The Howey analysis is multifaceted, and the presence of staking rewards is only one factor. It would be naive to think this proposal, if passed, would immunize Ethereum from securities claims. But in a regulatory environment where every aspect of digital asset mechanics is scrutinized, a shift in the reward structure away from issuance and toward fees could reduce the most semantically damaging feature of staking.
This could be an asset for the network. It could also be a trap: if the proposal reduces the legal exposure of staking while simultaneously concentrating participation among institutional validators, then the regulatory spotlight may simply move from "profit expectations" to "control concentration." The worst constitutional outcome would be an Ethereum that is at once less decentralized and more legally exposed.
Contrarian Part I: Give the Devil His Due
Having laid out my objections at length, let me now steelman the proposal more aggressively than its advocates have managed.
The subsidy critique is defensible. The Ethereum staking issuance curve was designed in an era when the network needed to attract participants. That world ended. Staking participation has exceeded 28% and continues to climb, and Lido's dominance is a structural problem no one has solved. The marginal security value of each additional staked unit of ETH declines as the attack cost becomes asymptotic. A network that continues to print billions of dollars of new ETH each year to reward staking that contributes little incremental security is running a welfare program, not a security apparatus.
From this perspective, the proposal is a market correction. It eliminates the excess subsidy, reduces dilution, and preserves the security that the network actually needs. The objection that validators will "exit" is not a contradiction of the proposal; it is the stipulated mechanism of adjustment. If the market value of validation services cannot sustain the current validator population, then the population should contract. The network emerges cheaper, scarcer, and more valuable per coin.
This is the "minimal issuance" school, and it has deep roots. Justin Drake has argued publicly for an endgame in which the emission curve asymptotes toward zero, with security paid for by the active users of the chain rather than by passive holders. This proposal is the clearest expression of that worldview yet placed on the table. It deserves to be taken seriously on those terms.
The proposal is also not obviously hostile to decentralization in its long-run design. If the subsidy shrinks, the prize that drives Lido's dominance may also shrink. A smaller issuance subsidy could make the market share battle less profitable and less consequential. The protocol's security would rest on the actual value of the chain's activity, not on the artificial value of its inflation.
I am genuinely torn. The diagnosis of over-subsidy is correct; the prescription needs careful scrutiny. It is rare in this industry to see a policy proposal this honest about its trade-offs.
Contrarian Part II: Equilibrium Is Not Justice
But the steelman collapses, in my view, on three counts.
First, equilibration through exit is not a process that rewards the most deserving participants. It rewards the largest, the most capitalized, and the most strategically well-positioned. When the base yield is halved, the validator population does not shrink uniformly. The solo stakers, the community nodes, the small operators in regions with high costs โ they leave first. The institutional operators with access to cheap capital and concentrated MEV capture remain. The resulting equilibrium may be efficient by some narrow measure, but its distributional character is the opposite of what Ethereum's stated values require. A mechanism that achieves efficiency by destroying the small participant is not a Pareto improvement; it is a social reordering that betrays the network's ethos.
Second, the proposal's treatment of MEV as an acceptable residual income source ignores the market structure of MEV entirely. Shifting validator income from issuance to MEV does not merely leave concentration levels unchanged; it actively increases the competitive importance of the most concentrated income stream in the ecosystem. The proposal shorts the very thing it purports to optimize for: the diversity and independence of the validator set.
Third, the proposal's timing is intrinsically hostile to the proposition that it is a sober technical correction. No audit, no reference implementation, no long-form economic modeling released to the community โ the authors made a compressed, high-stakes move in a governance window that made debate difficult. A technical proposal about the protocol's economic constitution should welcome scrutiny, not evade it.
There is a middle path, and I have argued for it in multiple contexts: a progressive issuance curve that flattens but does not zero-out, with a modest bonus for small validators, and a transition that is long enough for the ecosystem to adjust without trauma. The authors have chosen a cliff instead. The cliff is intellectually elegant; it is operationally cruel.
Equilibrium is a description, not a justification. A market can equilibrate toward a concentrated, centralized, and brittle validator set. It has done so in other proof-of-stake networks. The question is not what the equilibrium will be; it is whether the equilibrium is one we can live with.
Contrarian Part III: The Precedent That Follows
The final argument against the proposal in its current form is one of governance precedent.
Ethereum's monetary rulebook is not a collection of parameters awaiting optimization; it is the constitutional layer of a multi-trillion-dollar economic network. Changing issuance, burning rules, or the validator compensation schedule is not a routine upgrade. It is a constitutional amendment. The fact that this amendment is arriving as a fast-track draft, with high-stakes timing, changes the cost of the precedent more than the draft's particular parameters do.
If Ethereum accepts that the consensus emission curve is revisable through a rapid, researcher-driven process, every future emission parameter becomes a possible vector for governance capture. The stake of those who believe in a near-invariant supply schedule for ETH โ and there are many, including major institutional holders โ is precisely that the monetary rulebook is not up for grabs every cycle.
The proposal's supporters would respond that the protocol's monetary policy should be as responsive to community preference as any other dimension. But that responsiveness is a double-edged sword. It is what transforms a neutral settlement layer into a contested political arena where consensus can be gamed by whoever controls the narrative window.
I have watched DAOs fracture on exactly this terrain. When a governance mechanism becomes a tool for sudden redistribution, the community splits into factions that no longer share enough common ground to govern. Ethereum's strength has always been the opposite form of stability: a deliberate pace that gives legitimacy time to form.
The price of the proposal, if it passes in anything like its current form, may be that it becomes a precedent for future constitution-toppling exercises. That is a cost the draft does not price.
Signals to Watch
Given where we are, I want to offer a practical observation map for those of us who think in market and governance signals rather than in abstractions.
First, watch the staking ratio. If the proposal gains traction and the community begins to price a lower net consensus yield, expect a measurable reduction in new deposits into the beacon chain, and eventually an increase in exit requests. A weekly decline of more than a couple percent in net staked ETH would be a strong market signal that participants are voting with their keys.
Second, watch the LST discounts. The ratio of stETH to ETH, and of weETH to ETH, is the most immediate pricing barometer of staking economics. A sustained discount beyond the historical mean indicates that the market is repricing the underlying yield, not just speculating about the proposal.
Third, watch the core developer calls. The proposal's fate rests, at least procedurally, on whether it is picked up for discussion in the all-core-devs meetings and assigned a formal EIP number. If it is postponed repeatedly, it will likely die in committee. If it is prioritized, expect an extended and unstable governance season.
Fourth, watch the public positioning of Lido, Aave, and ether.fi. A coordinated response โ a counter-proposal, a public statement, a governance campaign โ would mark the beginning of a genuine partisan conflict. A quiet absorption would indicate that the DeFi establishment has other channels for influencing outcomes, channels that may be less visible but more effective.
Fifth, watch the exchange staking products. Coinbase, Binance, and others will adjust their staking APRs if the base rate drops. When the terminal user begins to see lower staking rewards, the retail narrative will shift from "Ethereum yields" to "Ethereum costs," and the prolonged dominance of staking-as-income will be tested.
Takeaway: The Conversation Is the Signal
I do not think this proposal will pass in its current form. It is too sharp, too abrupt, too hostile to the operational class it affects. But it will be amended, shaped, and partially absorbed โ because the problem it identifies is real.
The subsidy is over-sized. The concentration it enables is dangerous. The security budget has become a rent transfer to intermediaries. Those are not inventions of the proposal; they are the facts of the post-Merge era.
What is at stake is not whether the curve should be flattened. It is how the curve should be flattened: whether through a cliff that destroys the smallest validators, or through a slope that gives the ecosystem time to reorganize its compensation culture. The difference between a cliff and a slope is the difference between governance that honors process and governance that imposes will.
We built a kingdom of ghosts in the machine. Those ghosts have families, capital expenses, and commitments. They are the humans behind the validator keys, and their presence is what gives the machine its texture of trust. The proposal's authors have reminded us that the machine can be redesigned; they have also reminded us that redesigning it without the ghosts is a recipe for haunting.
The conversation, though, is the signal. Ethereum's governance is entering a season in which the protocol's economic constitution will be up for consideration in ways it never was in the first decade of its existence. The staking subsidy question is not a technical footnote; it is the question of who participates in the network's defense and on what terms.
I found my own clarity in this debate in an unexpected place โ the writings on algorithmic altruism that I have been developing for my current work. The conclusion is simple: in the design of collective systems, efficiency is not an end. It is a means to the preservation of the humans who live inside the system. A security subsidy that protects diversity is worth paying; a security subsidy that protects rent is not.
The code is law, but the humans are the bug. And the only way to debug the present is to govern the future without pretending that distributing the losses is a matter of dispassionate mathematics.
To govern the future, we must debug the present.