EIP-8361

Both in percent per year, by total ETH staked — from nothing staked to the entire supply.

At 41.7M staked today: 2.58% yield, 0.89% dilution.

Consensus-layer issuance only, at full participation: the protocol pays out 64 × √(total staked) per epoch, so issuance rises with the square root of the stake while a non-staker’s balance stays flat. Yield is that issuance divided by the stake, so it excludes tips and MEV, which are transfers from users rather than new supply. Fee burn, which offsets supply growth for stakers and non-stakers alike, is excluded too. Supply assumed 120.5M ETH. Dilution is issuance ÷ supply, the same base as the yield, so the gap between the curves is exactly the transfer from non-stakers to stakers: yield ÷ dilution = supply ÷ staked. At 100% staked that is 1 — the yield is entirely dilution and a staker’s real return is zero.

A burn of b = (D ÷ D_sat)^1.5 cancels issuance entirely at 60.25M staked. Drag the phase-in to see the elevated base reward factor decay away.

At 41.7M staked today, the burn is 58%, leaving 2.19% yield and 0.76% dilution, against 2.58% and 0.89% today.

Same assumptions as above — consensus-layer issuance at full participation, no tips or MEV, no fee burn — with the draft’s deduction applied on top: every assigned duty is charged b = (D ÷ D_sat)^1.5 of its idealised reward and that ETH is destroyed, so net issuance is the current curve times (1 − b). With D_sat = 60.25M, issuance peaks near 23.91M staked (20% of the supply) and falls to zero at 50%, past which the clamp holds it there. Both curves reach zero together, so beyond saturation staking earns no issuance and holders are not diluted by it. The phase-in only scales the whole schedule; it does not change that shape.