EIP-8361 would burn a rising share of consensus-layer issuance as the staking ratio climbs, reaching 100% at 60.25 million ETH. The stated reason is concentration, not inflation. The strongest objection is that a lower yield may concentrate staking further, not less.
Roughly a third of all ether in existence is currently locked in the staking system. On 4 August 2026, staking participation reached an all-time high of about 33.98% of circulating supply — up from around 29% at the start of the year. On paper that is the network working exactly as designed. Proof of stake pays holders to bond capital in exchange for the right to validate, and more bonded capital is supposed to mean a more expensive network to attack.
Six Ethereum researchers, including the Ethereum Foundation's Justin Drake, have submitted a draft proposal arguing that this is no longer true — that past a certain level, additional staking makes Ethereum less secure rather than more — and proposing to fix it by taxing the exact behaviour the protocol was built to encourage.
Nearly every write-up has filed this as a yield story: staking rewards may fall. That is the consequence, not the argument. The argument is about who is doing the staking.
EIP-8361 was published as a draft on 4 August 2026. It is twenty days old, still a draft, and contested. It has also circulated under the number EIP-8363 because of a numbering conflict — the two refer to the same proposal.
Staked ether is not gone, but it is not liquid either. To exit, a validator joins an exit queue whose throughput is capped by the protocol — a deliberate design choice to prevent a mass, simultaneous withdrawal from destabilising consensus. In practice that means a meaningful share of ether responds to price with a lag measured in days or weeks rather than seconds.
The intuitive read is that this is bullish: a third of the supply is off the market and cannot be sold into a rally. The less comfortable read is that the same mechanism works in reverse. Capital that cannot exit quickly in a rally also cannot exit quickly in a rout, and a full exit queue during stress is a structural feature of the system, not a malfunction. Both readings are correct. Which one matters depends on the day.
The proposal's case is not that staking rewards are too generous in the abstract. It is that the marginal staker has changed.
Early in a proof-of-stake network's life, growth in the staking ratio comes from a wide base: individuals running validators, small pools, dispersed operators. That growth genuinely buys security, because it distributes the power to propose and attest blocks across more independent parties. Later, once the accessible retail base has largely staked, incremental growth arrives overwhelmingly through the paths of least friction — large custodians, exchanges and liquid-staking providers, which offer a token receipt instead of the operational burden of running a node.
At that point, the researchers argue, the security budget keeps rising while the thing the budget is supposed to purchase — decentralised control of consensus — stops rising and may fall. You are paying more for less. The concentration figures are the reason the argument has traction: Lido Finance alone holds roughly 8.89 million ETH and around 61.66% of the liquid-staking market. Whatever one thinks of Lido's governance, a single protocol intermediating that share of the fastest-growing staking channel is exactly the outcome the proposal is aimed at.
This is where the proposal is genuinely elegant, and it is worth understanding as a worked example of monetary policy set by protocol rather than by committee.
EIP-8361 does not cap staking. It does not reject validators, impose a queue limit, or blacklist operators. It changes what a validator is paid. As the staking ratio rises, a growing fraction of newly issued consensus-layer rewards — the issuance paid for attestations, block proposals and sync-committee duty — is burned rather than delivered. The burn fraction scales with the staking ratio raised to the power of 1.5, so it is gentle at low participation and steepens sharply as the ratio climbs.
The endpoint is fixed: a saturation balance of 60.25 million ETH, roughly half of supply at the time of the fork. At that level the burn reaches 100% and a validator performing its duties perfectly earns zero net consensus-layer yield. The change would phase in over 18 months rather than at a single block, specifically so that validators are not all handed a new economic reality on the same day.
At current participation, the proposal projects annual consensus-layer yield falling from roughly 2.6–2.7% to about 1.2% across the transition.
Two things are worth being precise about. First, this is consensus-layer issuance only. Validators also earn priority fees and MEV, which the proposal does not burn — so total validator revenue does not go to zero even at saturation, and the gap between issuance yield and total yield becomes much more important than it is today. Second, burning issuance is not the same as not issuing it. The supply effect is deflationary in the same way EIP-1559's fee burn is, which is part of why the proposal has support from holders who do not stake.
The strongest counter-argument is not that lower yield is unfair. It is that lower yield may make concentration worse rather than better — which would invert the entire justification.
Staking operators do not share a cost base. A solo staker running a node at home faces hardware, electricity, bandwidth, uptime monitoring and the opportunity cost of their own attention — costs that are largely fixed per validator and do not fall with scale. A large custodial operator runs thousands of validators on shared infrastructure with amortised engineering and near-zero marginal cost per additional validator.
Compress the yield and you compress everyone's margin, but you do not compress it evenly. The operator with the highest cost floor becomes unprofitable first. In this case that is plausibly the independent solo staker — precisely the participant the proposal wants to protect — while the custodial operator with a 1.2% yield and near-zero marginal cost simply continues, and picks up the exiting stake. If that dynamic dominates, EIP-8361 accelerates the concentration it was designed to arrest.
Proponents have a response: much of the marginal growth is arriving through custodial channels driven by institutional yield mandates, and those mandates are the most yield-sensitive money in the system — a fund benchmarked against a target return exits a 1.2% yield faster than a committed individual does. Whether the yield-sensitive money is institutional or retail is an empirical question about the composition of current staking flows, and it has not been settled.
The liquid-staking and DeFi side has raised a separate and more direct objection: a large share of on-chain lending and structured products is collateralised by liquid-staking tokens whose economics assume a yield. Halving that yield does not merely reduce returns; it changes the viability of products built on top. Critics argue the inflation reduction achieved is modest relative to that disruption. That is a real cost and the proposal's supporters have not fully priced it.
A draft EIP is a proposal, not a schedule. Most draft EIPs never ship. This one is unusual in the seniority of its authors and in how directly it touches holder economics, which cuts both ways — it will get serious technical review, and it will get serious organised opposition from the staking-services businesses whose revenue is a share of the yield being burned.
The realistic path is not adoption or rejection of this exact curve. It is a debate over the saturation balance, the exponent and the taper length, with something meaningfully different eventually shipping or nothing shipping at all. Watch the parameters, not the verdict.
“A blockchain adjusting its own monetary policy to discourage its own core mechanism is not a common event. The disagreement about whether it will work is more instructive than the proposal itself.”
What makes it worth following even for people who hold no ether is that it is one of the clearest live examples of a system setting monetary policy without a central bank: a published rule, an argued rationale, an open objection, and no authority able to impose the outcome. The debate is the mechanism.
The argument is about composition, not quantity. Early staking growth comes from a dispersed base and genuinely distributes control over consensus. Later growth arrives mostly through large custodians and liquid-staking providers, because those are the lowest-friction routes. So the total value bonded keeps rising while the number of independent parties controlling it does not — you pay more for a security property you are no longer buying. Whether this describes Ethereum today is exactly what is contested.
DrafterDaily covers the mechanisms underneath crypto markets — governance, issuance, custody and the arguments the participants are actually having.
Read more Crypto coverageA race condition between the display logic and the signing buffer breaks the one promise a hardware wallet exists to make — that what the screen shows is what the key signs — without breaking any cryptography at all.
$26 billion. $31 billion. $38.17 billion. Tokenized Treasuries at $6.8bn, $12.88bn, $15bn, $16.21bn. All published in 2026, all defensible, none reconcilable without knowing the date and the definition.
Coverage fixed on the $75 million exemption. The consequential provision is a conditional safe harbor that would let a crypto asset stop being subject to an investment contract — and a preemption clause almost nobody has read.