One Ethereum proposal would cut staking rewards 48%

What Is EIP-8363 and Why Would It Cut Staking Rewards 48%?

A draft Ethereum proposal published on August 4, 2026 would gradually burn validator rewards out of existence as more ETH gets staked — and one prominent DeFi founder ran the numbers at today's staking level and got a 48% pay cut. The reaction, from Aave to ether.fi to Bitwise, arrived within days.

EIP-8363, titled "Tapered Issuance Burn," is a draft Ethereum Improvement Proposal that would burn a progressively larger share of every validator's consensus-layer rewards as total staked ETH rises, driving net protocol issuance yield to zero once roughly 50% of supply is staked . It was published to the Ethereum EIPs repository and the Fellowship of Ethereum Magicians forum on August 4, 2026. The headline number driving the backlash comes from Aave founder Stani Kulechov, who calculated that validator income would fall from 2.862% to 1.476% at current staking levels — a 48% reduction .

Quick Answer: EIP-8363 is a draft Ethereum proposal, published August 4, 2026, that would burn a rising fraction of validator rewards as staking grows, reaching zero net issuance yield at 50% of supply staked. Aave's Stani Kulechov calculates a 48% income cut at today's 34.07% staking ratio. It is not approved or scheduled.

The backdrop matters for reading that figure. At publication, roughly 41.3–41.5 million ETH was staked — about 34.07% of supply — across approximately 890,000 active validators, earning an average 2.67% . Staked ETH had risen about 15% since the start of 2026, partly on institutional inflows, with liquid staking exposure at 15.04 million ETH worth roughly $28.2 billion . In other words, the proposal targets a trend that is currently accelerating, which is why the reaction from staking operators was immediate.

One status point belongs up front, because it is the most commonly mis-stated fact in coverage: this is a draft, not an enacted change. EIP-8363 landed two days before the August 6 deadline for proposing non-headliner EIPs for Hegotá, the upgrade following Glamsterdam, and it received 30 minutes on the ACDC #184 call — but it did not reach PFI ("proposed for inclusion"), the weakest formal stage in the process . It is not approved, not scheduled, and not part of any upgrade. A search of the canonical index at eips.ethereum.org/all returned no merged EIP 8363 during research, consistent with draft status. Note also that the authors self-assigned EIP-8361 in the pull request, and much early coverage used that number before editors reassigned it as 8363 — both numbers refer to the same proposal.

How the Tapered Issuance Burn Formula Actually Works

The Tapered Issuance Burn works by destroying a rising share of every validator's consensus-layer reward as the total amount of staked ETH grows. The forum specification defines the burn fraction as b = (D / 60,250,000)^1.5, where D is total staked ETH; at the saturation point of 60.25 million ETH — roughly half of current supply — b reaches 1.0 and the protocol keeps every unit of issuance it just credited . Because the exponent is 1.5 rather than 1.0, the deduction stays modest at low staking ratios and steepens as deposits climb, which is the whole design intent: make the marginal ETH deposited progressively less profitable without hard-capping the validator set.

The burn is applied per duty rather than as a single end-of-epoch adjustment. Validators are first credited the full reward they earned for attestations, block proposals and sync-committee participation, and a participation-scaled burn then removes fraction b of that total, leaving a correctly-attesting validator with (1 − b) of what the current rules would have paid . Two consequences follow from that ordering. First, the burn scales with performance, so a validator that misses duties is burned less in absolute terms but keeps the same proportional haircut. Second, penalties are untouched by the formula — a point that becomes central to the downtime-recovery objection raised in the forum thread.

The table below is arithmetic from the published formula, not a separate forecast, and covers consensus-layer issuance only.

Those raw fractions are larger than the headline numbers circulating in coverage, and the gap is explained by what the burn does not touch. MEV income is explicitly excluded from the mechanism , so a validator's blended yield falls by less than b. Aave founder Stani Kulechov's widely cited estimate of a 48% drop in validator income — from 2.862% to 1.476% at present staking levels — is a figure for total take-home yield including execution-layer revenue, not the consensus-layer burn fraction itself . Roughly 41.3–41.5 million ETH is staked today, about 34.07% of supply across some 890,000 active validators, earning an average 2.67% .

Nothing switches on at once. The specification phases the burn in over roughly 18 months, in about 64 discrete steps of 8.6 days each, so the effective burn fraction ratchets toward its formula value rather than landing in a single epoch . Under the authors' own model, aggregate issuance peaks near 0.5% of supply per year at roughly a 20% staking ratio, then tapers toward zero as the ratio approaches 50% . That is the structural break from today's rules, where issuance scales with the square root of total stake and yield decays asymptotically — leaving a residual incentive of roughly 1.5% at high ratios that never reaches zero . One further mechanical effect matters for later sections: with issuance suppressed, the share of validator revenue coming from issuance falls from about 93% to roughly 70%, mechanically raising MEV's weight in what validators actually earn .

Where EIP-8363 Actually Stands: Draft, Not Law

EIP-8363 is a draft proposal with no approval, no schedule, and no place in any Ethereum upgrade. It was published on August 4, 2026 to the Ethereum EIPs repository and the Fellowship of Ethereum Magicians forum , landing two days before the August 6 deadline for pull requests proposing non-headliner EIPs for Hegotá, the upgrade that follows Glamsterdam . That deadline governed the act of proposing, not any decision on inclusion . Anyone trading on the assumption that validator rewards are about to be cut is trading on a working paper.

Even the identifier is unsettled. The authors self-assigned EIP-8361 in the pull request, and much of the earliest coverage carried that number — including reporting from Crypto Times. EIP editors then found 8361 was already allocated and reassigned the draft as EIP-8363 . Both numbers remain in circulation across press coverage of the identical proposal, so searches on either string return the same document. A useful sanity check for readers: a search of the canonical index at eips.ethereum.org/all returned no merged EIP 8363 during research . Draft pull requests are not published to that index until merged, and the repository's own guidance is to treat proposals as working papers unless they appear at the canonical site .

The proposal did get a serious hearing. On the All Core Devs Consensus call #184 on August 6, EIP-8363 received 30 minutes of agenda time — more than any other Hegotá-block item — with ETHCC co-founder Jérôme de Tychey presenting . Participants pressed on small-validator economics and centralisation, Obol's Oisin Kyne raised the Nakamoto coefficient, and Lido Labs' Greg Koumoutsos argued the draft should be withdrawn on the grounds that there had been insufficient preparation time . Koumoutsos framed the underlying stake in terms of what issuance buys:

"Ethereum is not only paying for slashable ETH; it is paying for decentralization, operator diversity, censorship resistance, and network resilience," — Greg Koumoutsos, Lido Labs (source: Cointelegraph Magazine)

The recorded next steps from that call were narrow and procedural: consider withdrawing the proposal from Hegotá consideration, and answer every comment on the forum thread . Neither is a step toward shipping. Critically, EIP-8363 has not reached PFI — "proposed for inclusion," the weakest formal stage in Ethereum's upgrade process, and the one that merely marks a proposal as worth evaluating for a specific hard fork .

For traders, the practical status ladder is worth holding in mind, because headlines routinely collapse it:

• Draft / working paper — where EIP-8363 sits today: published, discussed, revisable, withdrawable.
• PFI (proposed for inclusion) — client teams agree the idea merits evaluation for a named fork. Not reached.
• CFI / SFI (considered / scheduled for inclusion) — implementation and testing commitments follow. Not reached.
• Devnet and testnet activation — months of client work, typically with public specs frozen first. Not reached.
• Mainnet fork — the only point at which validator income actually changes.

The gap between the first rung and the last is measured in quarters, not days — and the proposal's own transition schedule would then phase the burn in over roughly 18 months across about 64 steps of 8.6 days each . Nothing about the current staking yield of roughly 2.67% across about 890,000 active validators is scheduled to change . What has changed is the disclosure that a serious group of researchers wants the ceiling debate on the table — and that governance signal, rather than any imminent cut, is what the market is presently repricing.

The Case For It: Why Researchers Want a Staking Ceiling

The case for EIP-8363 rests on a structural quirk in Ethereum's existing issuance curve: rewards scale with the square root of total stake, but that pool is divided across every staked ETH, so per-validator yield falls asymptotically without ever reaching zero. Even at high staking ratios, roughly 1.5% remains as a standing incentive to keep adding stake . The authors argue that curve never tells the network when it has enough security — so the protocol keeps paying regardless of whether the marginal validator adds anything.

Their claim is one of diminishing marginal returns. Past some threshold, additional stake buys progressively less security while diluting holders who do not stake, since issuance is paid by everyone and captured by validators. A tapered burn replaces the asymptote with a genuine terminal point: under the authors' model, issuance peaks near 0.5% of supply per year at about a 20% staking ratio, then tapers toward zero at 50% . In that framing, the proposal is less a reward cut than a decision to stop subsidising stake growth once security is adequately funded.

The intellectual lineage matters for judging seriousness. The core mechanism is credited largely to prior work by forum contributor pa7x1, drawing on Anders Elowsson's research into per-duty burns, and the author list includes Ethereum Foundation researcher Justin Drake, ETHCC co-founder Jérôme de Tychey, Lodestar's dapplion, pintail and Ladislaus von Daniels . This is not a fringe submission; it is a compilation of several years of issuance-policy research inside the protocol's own research community, which is part of why the reaction has been so sharp.

Outside the protocol circle, at least one institutional researcher has framed a ceiling as constructive for ETH holders. Grayscale head of research Zach Pandl argued in May 2026 that limiting staking incentives would be "positive for the price of Ether over time" , on the logic that lower issuance means less structural sell pressure and less dilution for holders who never lock up their ETH.

"Limiting staking incentives would be positive for the price of Ether over time," — Zach Pandl, Head of Research at Grayscale, speaking in May 2026 (source: Cointelegraph Magazine).

The supporting argument, then, has three legs: the current curve has no natural stopping point; extra stake beyond a threshold is a cost rather than a benefit to non-staking holders; and a predictable taper is preferable to letting the staking ratio drift toward levels nobody explicitly chose. Whether those legs hold under the objections raised by operators and allocators is the next question — and the one that dominated the ACDC discussion.

For context on how the mechanism is applied per duty and how the schedule phases in, see the detailed thread on the Fellowship of Ethereum Magicians and the summary at DeFiPrime.

The Backlash: Kulechov, Silagadze and Berryman Push Back

Opposition to EIP-8363 arrived within 48 hours of publication and spanned a wider set of constituencies than most Ethereum monetary-policy fights: DeFi founders, liquid-staking operators, solo validators and institutional allocators all objected, largely for different reasons. Aave founder Stani Kulechov produced the number that defined the debate, calculating that validator income would fall 48% under the mechanism — from 2.862% to 1.476% at the current staking ratio . That single figure, drawn from present-day stake levels rather than the 50% saturation point, is why the proposal is discussed as a reward cut rather than a long-horizon taper.

Kulechov's argument was structural rather than arithmetic. "A zero-yield regime accelerates the capture it means to deter," he wrote, adding that "Ethereum's staking ratio is rising because staking works. That is what growth looks like" . The logic: if protocol issuance falls toward zero, the validators who survive are those with the lowest marginal cost and the best MEV access — large professional operators — which is the opposite of the decentralisation outcome the taper is meant to protect.

ether.fi CEO Mike Silagadze attacked both the substance and the process. He called the proposal "bad for decentralization, this is bad for Ethereum adoption, and this is bad for the credibility of the network" , objected that the "EIP [was] released with 48 hours notice for comments," and warned it would "kill a huge chunk of DeFi which is built around the staking ecosystem" . The DeFi point is concrete: liquid staking exposure stood at 15.04 million ETH worth roughly $28.2 billion, with Lido holding 62.7% of the LST market — collateral whose yield underpins lending markets, looping strategies and structured products.

Bitwise's head of Ethereum partnerships, Dr. Steve Berryman, reframed the objection entirely — his concern is not the reward level but the fact that it can be changed at all.

"Institutional adoption requires predictability... even tinkering with the issuance structure can create uncertainty," — Dr. Steve Berryman, head of Ethereum partnerships at Bitwise (source: Cointelegraph Magazine)

Berryman's position is that institutions would begin pricing in "yield governance risk" as a standing discount, and that staking participation would plateau naturally around 2% yields without intervention . Lido Labs' Greg Koumoutsos made the complementary case on what the issuance actually buys: Ethereum "is not only paying for slashable ETH; it is paying for decentralization, operator diversity, censorship resistance, and network resilience" . In the August 6 ACDC #184 call, Koumoutsos argued for withdrawing the proposal from Hegotá consideration on grounds of insufficient preparation time .

The opposition was not unanimous. Grayscale head of research Zach Pandl said in May 2026 that limiting staking incentives would be "positive for the price of Ether over time" , and commentator Leo Lanza pushed back on the "stealth tax" framing by noting Ethereum's sub-1% annual inflation compares favourably with gold's 1–2% supply expansion . Still, observers describe the exchange as the most significant debate over Ethereum's monetary policy since EIP-1559 — and the forum thread raised four further objections that go beyond rhetoric.

Four Structural Risks the Forum Flagged

The four risks raised on the Ethereum Magicians thread are distributional, not theoretical: a flat percentage burn hits solo validators hardest, downtime recovery lengthens by 3.6–14×, liquid-staking loops break below a 1–1.5% yield spread, and issuance falling from roughly 93% to 70% of validator income mechanically raises MEV's share of the pie . Each one points the same direction — toward operators with scale.

1. The burn is flat; costs are not. EIP-8363 deducts the same fraction b from every correctly-attesting validator regardless of who runs it. A home validator running a node on residential power and hardware carries a far higher proportional cost base than an operator amortising infrastructure across thousands of keys, so an identical percentage cut compresses the solo operator's margin first. That asymmetry sat behind the small-validator and centralisation concerns raised during the 30 minutes EIP-8363 received on the ACDC #184 agenda on August 6, 2026, where Oisin Kyne also invoked the Nakamoto coefficient .

2. Downtime becomes far more expensive to work off. Inactivity penalties in Ethereum's consensus layer are denominated independently of the reward burn, so shrinking rewards while holding penalties constant stretches the time needed to recover a balance. Forum participants put the extension at 3.6× to 14× depending on the staking ratio . Operators with redundant hardware and on-call staff absorb that; hobbyists on a single machine may not.

3. The liquid-staking spread is thin enough to close. Looping strategies — borrowing against an LST to re-stake — only clear when the staking yield exceeds the borrow rate by roughly 1–1.5%. Liquid staking exposure stood at 15.04 million ETH, about $28.2 billion, with Lido at 62.7% of the LST market . Cutting protocol yield from 2.862% to 1.476% at current levels puts that spread inside the noise band — the mechanical basis for ether.fi CEO Mike Silagadze's warning that the change would remove a large share of the DeFi built around staking .

4. ETF products are underwritten on distributable yield. US spot ETH ETFs structured around passing staking income to holders have roughly $11.2 billion of inflows exposed to a reward cut of this size . That is the same objection Bitwise's Dr. Steve Berryman framed as predictability risk rather than a complaint about yield levels.

The composition problem underneath all four. EIP-8363 leaves MEV income untouched . As protocol issuance drops from about 93% to 70% of validator revenue, MEV's relative weight rises by construction — and MEV accrues disproportionately to operators with better relay access, block-building relationships and latency. The forum's read is that a proposal aimed at limiting concentration would, through this channel, shift the remaining revenue toward exactly the sophisticated operators it worries about . None of these objections is fatal on its own; taken together they explain why Lido Labs' Greg Koumoutsos argued for withdrawing the draft from Hegotá consideration rather than iterating on it under deadline.

Monetary Policy Predictability: EIP-8363 vs Bitcoin's Fixed Supply

The "Ethereum is losing its monetary edge to Bitcoin" line attached to EIP-8363 is press and commentary shorthand, not language the proposal's authors used. The Ethereum Magicians thread for the Tapered Issuance Burn contains no direct comparison to Bitcoin's fixed supply , and The Defiant's coverage of the Kulechov and Silagadze objections carries no head-to-head Bitcoin argument either . What does exist in the record is a structurally identical critique — one about who can change the rule, not about which chain issues fewer coins.

That critique runs as follows. A monetary rule that a governance process can revise on 48 hours' notice for comment is a weaker commitment than one enforced by the software every full node runs. Bitcoin's own framing makes the distinction explicit: bitcoin.org states that developers cannot force protocol-rule changes because users choose which compatible software to run, and Bitcoin Core validation materials present full validation as the mechanism that protects users against miners violating rules such as the 21 million coin limit . Ethereum's issuance curve is a parameter set by rough consensus among client teams and researchers, revisable through the same all-core-devs pipeline that reviewed EIP-8363 on August 6, 2026 . Bitwise's Dr. Steve Berryman named the resulting cost directly: institutions would begin pricing in "yield governance risk," because "even tinkering with the issuance structure can create uncertainty" .

The counterargument is that Ethereum already wins on the number that the comparison is nominally about. Commentator DCinvestor pointed out that "ETH issuance is already lower than Bitcoin's and than annual newly mined gold supply," arguing that consistency of the rule matters more than its absolute level . Leo Lanza made the same point against the "stealth tax" framing, noting that Ethereum's sub-1% annual inflation compares favourably with gold's 1–2% supply expansion . Under the authors' own model, issuance peaks near 0.5% of supply per year at roughly a 20% staking ratio before tapering toward zero at 50% — so the dispute is about the stability of the schedule, not its generosity.

"Institutional adoption requires predictability… even tinkering with the issuance structure can create uncertainty." — Dr. Steve Berryman, head of Ethereum partnerships at Bitwise (source: CoinOtag)

One market caveat belongs alongside this framing. Coinbase Institutional wrote on May 15, 2026 that Bitcoin's resilience appeared supported by ETF and treasury demand while ETH/BTC remained below both its 50-day and 200-day moving averages, despite improving staking and treasury narratives . That is nearly three months before EIP-8363 was published. The relative-demand gap predates the proposal, so any chart that pairs a falling ETH/BTC ratio with this draft is showing correlation the timeline does not support. The predictability argument against EIP-8363 stands on governance grounds; it does not need — and is not supported by — a causal price story.

What Happens Next for EIP-8363

EIP-8363's near-term path narrows to two options: withdrawal from Hegotá consideration, or a revised draft that answers the small-validator and centralization objections raised on All Core Devs Consensus call #184 on August 6, 2026 . Both were recorded as next steps on that call, alongside a commitment from the authors to answer every forum comment. The proposal never reached "proposed for inclusion," the weakest formal stage, so nothing is scheduled and nothing is decided .

The single most useful thing to track is the Nakamoto coefficient argument Oisin Kyne put to the call. It reframes the debate away from yield levels and toward the measurable question: does a per-duty burn make the validator set more or less concentrated? The proposal's stated purpose is to stop subsidising stake that buys diminishing security, but forum critics argued the same flat burn taxes home validators and institutional operators identically despite very different proportional cost bases, and that shrinking issuance from roughly 93% to 70% of validator income raises MEV's relative weight — a revenue stream that concentrates rather than distributes . Any revised draft that does not model that trade-off directly will face the same objection again.

A second proposal deserves separate tracking, not conflation. "Validator Redirected Revenue," posted to Ethereum Research on June 21, 2026 by Kleros founder Clément Lesaege, would let validators signal via hard fork to redirect up to 10% of staking rewards to ecosystem funding when 51% signal a rate above zero . It is explicitly pre-EIP and drew its own structural criticism — EIP editor emeritus Micah Zoltu argued unresolved cartel and allocation-game risk made the rest of the design moot, and Lesaege conceded a 51% majority could take 10% of everyone else's yield while calling the risk theoretical and capped . Different mechanism, different stage, same underlying question about who controls validator income.

Observers have called this the most significant debate over Ethereum's monetary policy since EIP-1559 . The lasting effect may be less about the formula than about the precedent. Bitwise's Dr. Steve Berryman framed institutional exposure as pricing in "yield governance risk" — the possibility that a rule can change, independent of whether it does . With US spot ETH ETFs holding roughly $11.2 billion of inflows built around distributable yield, and liquid staking exposure at 15.04 million ETH worth about $28.2 billion, that risk premium has a large surface to attach to .

Practical takeaway for anyone holding staked ETH or LSTs: do not reprice positions on a draft. The concrete trigger to watch is whether EIP-8363 is formally withdrawn from Hegotá consideration or resubmitted with decentralization modelling attached — check the Ethereum Magicians thread and ACDC agendas, not headlines . Until a proposal clears PFI and appears merged at the canonical index , the operative number for staked ETH remains today's average 2.67% yield across roughly 890,000 active validators, not the modelled 1.476% .

Frequently asked questions

What is EIP-8363?

EIP-8363, titled "Tapered Issuance Burn," is a draft Ethereum Improvement Proposal published on August 4, 2026 to the Ethereum EIPs repository and the Fellowship of Ethereum Magicians forum . It would burn a progressively larger share of every validator's consensus-layer rewards as total staked ETH rises, using the burn fraction b = (D / 60,250,000)^1.5, where D is total staked ETH. At the 60.25 million ETH saturation point — roughly 50% of supply — the deduction reaches 100% and net protocol issuance yield falls to zero . The burn applies per duty across attestations, block proposals and sync-committee participation, phasing in over roughly 18 months in about 64 steps of 8.6 days each. MEV income is untouched. Authors include Ethereum Foundation researcher Justin Drake, ETHCC co-founder Jérôme de Tychey, Lodestar's dapplion, pintail, pa7x1 and Ladislaus von Daniels . The authors originally self-assigned EIP-8361, which EIP editors reassigned to 8363, so both numbers appear in coverage .

Has EIP-8363 been approved?

No. EIP-8363 is a draft that has not reached PFI ("proposed for inclusion"), the weakest formal stage in Ethereum's upgrade process — it is not approved, not scheduled, and not part of the Hegotá upgrade . It landed two days before the August 6, 2026 deadline for proposing non-headliner EIPs for Hegotá, the upgrade following Glamsterdam, and that deadline governed proposing EIPs rather than deciding inclusion . The proposal received 30 minutes on the ACDC #184 agenda on August 6 — more than any other Hegotá-block item — with de Tychey presenting and participants raising small-validator and centralisation concerns; recorded next steps included considering withdrawal from Hegotá consideration . Consistent with draft status, a search of the canonical index at eips.ethereum.org/all returned no merged EIP 8363 during research, since draft pull requests are not published there until merged .

How much would EIP-8363 cut ETH staking yields?

Aave founder Stani Kulechov calculated that validator income would fall 48% under the mechanism — from 2.862% to 1.476% at the staking level prevailing when the draft was published . That figure reflects roughly 41.3–41.5 million ETH staked, about 34.07% of supply across roughly 890,000 active validators earning an average 2.67% . The cut deepens as the staking ratio climbs: because the burn fraction scales as (D / 60.25M)^1.5, net issuance yield reaches zero at 50% of supply staked . Under the authors' own model, issuance peaks near 0.5% of supply per year at about a 20% staking ratio before tapering toward zero. MEV income is excluded from the burn, so total validator revenue would not fall by the full 48% — but issuance dropping from roughly 93% to 70% of validator income raises MEV's relative weight .

Why do people compare EIP-8363 to Bitcoin's monetary policy?

The Bitcoin comparison is press and commentary framing, not a claim made by the EIP authors — the Ethereum Magicians thread contains no direct Bitcoin fixed-supply comparison . What critics actually raise is a predictability argument: Bitcoin's 21 million coin limit is enforced by node validation, and bitcoin.org states that developers cannot force protocol-rule changes because users choose which compatible software to run . Ethereum's issuance schedule, by contrast, can be revised through governance. Bitwise's head of Ethereum partnerships, Dr. Steve Berryman, framed the objection in exactly those terms — "Institutional adoption requires predictability," and "even tinkering with the issuance structure can create uncertainty" — arguing institutions would price in "yield governance risk" . On the absolute numbers, commentator DCinvestor pushed back on the framing, noting that "ETH issuance is already lower than Bitcoin's and than annual newly mined gold supply" and that consistency matters more than the level .

What's next for EIP-8363?

The recorded next steps from ACDC #184 were for the authors to answer every forum comment and to consider withdrawing the proposal from Hegotá consideration . Four objection categories need answers: flat burns tax home validators identically to institutional operators despite higher proportional costs; recovery from downtime lengthens 3.6–14× because penalties stay constant while rewards shrink; LST looping becomes uneconomical below a 1–1.5% yield spread; and US spot ETH ETFs built around distributable yield have roughly $11.2 billion of inflows at stake . Lido Labs' Greg Koumoutsos argued for withdrawal at the call, citing insufficient preparation time, while ether.fi CEO Mike Silagadze objected that the EIP was "released with 48 hours notice for comments" . Watch the Ethereum Magicians thread and subsequent ACDC agendas for the decision, and treat the canonical index at eips.ethereum.org/all as the test for whether anything has actually merged .

Enjoyed this article? Subscribe to get new stories by email whenever they're published.

Subscribe