“Full censorship will be extremely difficult because solo stakers exist on Ethereum.” Vitalik Buterin, Home Staking Summit
Solo stakers are the heart of Ethereum: thousands of independent validators, aligned with no company or jurisdiction, are the network’s first and last line of defense for neutrality and censorship resistance. Solo stakers need real income to sustain their operations.
The goal is simple: preserve real yield for Ethereum participants, and with it, solo staking. The gray line is real yield under the current issuance curve; the blue line is the proposed change (EIP-8363). Today staking always beats holding ETH, so ever more ETH gets staked even as returns collapse; past the ~64% mark nobody earns a real profit. EIP-8363 settles staking before the 50% mark, so stakers keep earning real yield while, as a side effect, native ETH is also diluted far less. It need not be the final EIP on this matter; it is one promising approach, but an idea like this probably has to be implemented in the near future to protect the security of Ethereum itself.
Go deeper. This page is only an attempt to communicate the amazing work done by brilliant minds on this matter. Start with issuance.wtf, the curated hub of the issuance debate, then the expert writeups themselves: Elowsson’s Minimum Viable Issuance, the EIP-8363 specification, pintail’s staking-ratio analysis, and realbigsean’s issuance explorer. I highly encourage you to dig through the research yourself, or use AI tools to help unpack the high-level concepts.
Model. Current consensus spec: annual protocol issuance
I = 166.32 · √S ETH (S = ETH staked), from BASE_REWARD_FACTOR = 64
and 82,181 epochs/yr, so consensus APR = 166.32 / √S. A solo staker keeps 100% of
consensus rewards plus a pro-rata share of execution-layer tips + MEV (a transfer of existing ETH).
Net supply growth π = (I − burn) / supply. Real yields:
staker (1 + APR) / (1 + π) − 1; unstaked holder 1/(1+π) − 1.
EIP-8363 overlay (proposed). A fraction b = (S / 60.25M)^1.5 (capped at 1) of each
validator’s consensus rewards is deducted and burned, so net consensus issuance is I·(1−b), so staking
yield from issuance tapers to zero at 50% of supply staked (60.25M ETH). Tips/MEV are unaffected, and the reduced
net issuance also lowers supply growth. The transition lever models the spec’s ~18-month phase-in: the effective
base reward factor starts doubled (128) at activation and decays linearly to 64, scaling rewards and the burn’s
reference uniformly (the spec’s 65-step staircase is approximated continuously). The default shows day one of
the transition; drag the months lever to 18 for the permanent regime.
Where does staking settle under each setup? The shaded zones on the chart are estimates, not protocol constants. A staking market settles where net yield falls to the risk premium stakers demand, called the reservation yield in Elowsson’s minimum-viable-issuance framework, with the equilibrium yield ultimately set by the lowest-cost staker type (liquid-staking tokens, not solo stakers). Current curve: there is no healthy resting point: Hasu & Konstantopoulos projected 80–100% staked once liquid staking removes the frictions, and past ~64% (the gray zone) after-tax real yield turns negative for every staker, yet holding unstaked stays worse still, so the race continues. EIP-8363: its rationale leaves the landing spot to the market but guarantees it sits below 50%, where the taper zeroes issuance yield. Two calibrations of a standard supply curve (yield-elasticity 2) pin it down: anchored at Elowsson’s illustrative point (2% yield at 25M ETH staked) it intersects the permanent curve near 25% of supply; anchored at today’s revealed behavior (~40M ETH at ≈2.8% nominal), near 24%. Sticky existing stake and risk premiums that fall as staking matures argue for headroom above those points, hence the blue zone at roughly 25–35%. Update these ranges as better estimates emerge.
Taxes. A rough estimate only: the tax rate applies to the full nominal reward at receipt (the common treatment for staking income); the dilution loss is not deductible, which is why tax bites harder than it looks and real yield can go negative at high staking rates. The unstaked holder’s dilution is unrealized, hence untaxed here. Capital gains on later sale, deductions, and wealth taxes are ignored; set the slider to 0% for the protocol-only view.
Why isn’t real yield zero at 100% staked? Issuance alone would give exactly zero: everyone dilutes everyone equally. What remains comes from fee transfers paid by transaction senders: tips/MEV go to stakers, and the burn keeps supply growth below gross issuance. Set tips/MEV and burn to 0 (and tax to 0%) and the line lands at 0.00%.
Assumptions. Perfect uptime, no slashing, average proposer luck; yields are protocol-denominated (share of total ETH supply), independent of ETH’s fiat price. Axis starts at 5% staked; below that, per-validator yields grow without bound. Defaults ≈ mid-Aug 2026: ~40M ETH staked (33.2%) of ~120.6M supply, tips + MEV ≈ 78k ETH/yr, base-fee burn ≈ 12k ETH/yr (30-day annualized). At 0% tax, current-protocol numbers cross-check with realbigsean.com/ethereum-issuance. Background on the issuance-policy debate (proposed curve reductions, EIP-8363): issuance.wtf. The break-even analysis mirrors pintail’s staking-ratio post; its ~64% figure matches this page’s default 40% tax rate.