The EIP That Wants to Cap ETH Staking at 50%

EIP-8363 proposes capping Ethereum's staking ratio at 50% by tapering issuance down to near zero as that threshold approaches, through a mechanism called the tapered burn. The proposal was published two days before the August 6th deadline for consideration in Ethereum's next upgrade, codenamed Hegota, though the text was finalized on July 14th. Hegota is not expected to ship until 2027, and the next real checkpoint is October, when core developers will decide whether EIP-8363 survives for potential inclusion.

Roughly one third of all ETH is currently staked, with nearly 900,000 validators, more than 41 million ETH locked in, and an additional 2.5 million ETH queued with waits of up to 43 days. Proposal author Jerome argues that once more than half of ETH supply is staked, the reserve of unstaked ETH stops functioning as a check on a rogue validator set, making the 50% threshold a no-go zone that is very hard to exit once crossed. He also argues that the longer the ecosystem waits, the larger the deposit base becomes and the more disruptive any eventual correction will be. Jerome estimates staking rewards under EIP-8363 would settle around 1.4 to 1.5 percent, identifies 15 to 20% staking ratio as a healthy minimum, and says 30% is the point where things start to tip.

Jerome contends that under the current issuance curve, solo stakers are already being squeezed out slowly and invisibly as the staking ratio climbs, with no dramatic moment of public debate, and that his proposal protects solo stakers earlier and more deliberately even if nominal yield looks worse on paper in the short run. He also notes that Lido directly asked him to withdraw the proposal, which he declined.

Mike Silagatse argues there is strong consensus among the builder, validator, and node operator communities that EIP-8363 is a bad idea now, and offered a public one million dollar bet that the network becomes more concentrated if the EIP passes. His core argument is that large institutional entities like Coinbase and Binance have no cost of capital and would continue staking even at 50 or 20 basis points yield, while a survey from ETH Staker found around half of independent node operators said they would stop staking entirely if yield dropped to 2%. Mike argues this dynamic means EIP-8363 would drive solo stakers out and leave only large institutional operators, producing the centralization the proposal claims to prevent. He also contends that if yield collapsed as expected, Ether.fi would exit the staking business and market share would likely consolidate to Lido, meaning a proposal designed to fight centralization could produce one dominant liquid staking winner as the only business model that survives at near-zero yield.

Mike further argues that a large portion of DeFi lending activity is built on staked ETH being reused as collateral across protocols through rehypothecation, and that reducing staking would destroy that demand and cause capital to exit DeFi. He contends that capping the reward does not reduce the wrapping of ETH into derivatives and only reduces the reward itself. Jerome counters that the dependence of DeFi on rehypothecated staked ETH is itself a systemic risk, citing a recent CapDAR event that led to a bailout of a major protocol. Mike disputes this framing, arguing that the actual staked ETH was never at risk in the KelpDAO incident and that what was hacked was a cross-chain protocol that then impacted a lending protocol where the failure occurred. Mike also argues that liquid staking tokens are best understood as a thin coordination layer between dozens or potentially hundreds of different entities, carrying less risk than a single entity like Coinbase running one giant server in an AWS availability zone.

Mike agrees that issuance probably needs to change eventually but objects to this specific proposal as the wrong fix at the wrong time, and states that EIP-8363 is not among the top ten problems facing Ethereum. Jerome argues that other networks having 80% or more of their token supply staked is not the right target for Ethereum's specific security model, and that proof of stake is not one universal design. Both sides agree that Ethereum's issuance curve was not designed for a world where a third of all ETH is staked and climbing every month, with the core disagreement being whether capping staking rewards protects network neutrality or engineers the very centralization it claims to prevent.