Ethereum

Here is a quick overview of EIPs that promote validator consolidation, before ACDC! 

The fast finality roadmap requires consolidation, to allow more stake to weigh in on consensus decisions with less overhead. We therefore wish to nudge stakers in this direction.

Developing transparent finance for @ethereum

EIP-8061 increases the exit and consolidation churn, so more stake can exit and consolidate each epoch. Reduced queues increase liquidity and help the network reconfigure stake after adverse events. We route exits through the consolidation queue when the exit queue is longer.

A higher deposit churn might also be beneficial. A concern is faster growth in active stake, but the equilibrium staking level is arguably best addressed by reducing the yield. Since a higher deposit churn is more controversial, the EIP is currently focused on exits/consolidations.

EIP-8062 charges a small fee for sweep withdrawals on 0x01 validators. It’s calibrated to roughly match the EL gas costs of the operations, at the average base fee over the last year. The fee is taken out as a percentage: 0.05%, for consistency, if the validator set size changes.

The fee is rather marginal, all things considered. We might suspect that $100k/year for an entity managing 5% of the stake is an insufficient incentive to consolidate. An alternative specification therefore has a higher fee. But before we get to that, we must review EIP-8068.

EIP-8068 addresses the lower capital efficiency of compounding 0x02 validators (blue line). There is on average 0.75 ETH not contributing to the effective balance (EB) as the balance compounds. They (0x02) therefore earn a lower yield than skimming validators (green circles).

EIP-8068 also addresses that 0x02 validators can repeatedly withdraw down below their EB, thus earning outsized yields (red lines). The solution is to set the EB hysteresis upward threshold to the neutral +0.5, and to let the EB shift after partial withdrawals at a neutral point.

What if we want to address the lower capital efficiency of 0x02 validators without reconfiguring the EB bands? As shown in this plot, the EIP-8062 fee is insufficient on its own, because that 0.75 ETH drag on the yield has a stronger effect. We could then just impose higher fees.

The EIP-8062 fee could be increased to 0.5% as shown by the full red line. At that point, an entity managing 5% of the stake would lose $1M/year when not consolidating, which is a fairly strong incentive. Under such a fee, we could reconfigure the EB bands as in EIP-8068 later.

This concludes the overview. Linking the EIPs:

https://eips.ethereum.org/EIPS/eip-8061
https://eips.ethereum.org/EIPS/eip-8062
https://eips.ethereum.org/EIPS/eip-8068

For questions on EIP-8061, you may also reach out to the EIP champion 
@fradamt. Other EIP contributors: 
@toniw.eth, Ben Adams, Maria Ines Silva, @caspar.

Explore the EIPs / ERCs mentioned in this cast:

However, a higher fee wouldn’t solve the other problem EIP-8068 addresses. It is still possible to game the EB by withdrawing down to 32.75 ETH (at a 33 EB), which can harm consolidation. Therefore, a simplified alternative specification in EIP-8068 addresses only this concern.