What is going on in Ethereum regarding the validator set?
There are nearly 900,000 active validators on the Ethereum blockchain who are helping to process transactions and keep the network aligned on the state of the world. This is a ~160% growth since the Shapella upgrade in April this year, when validators were able to exit as well as withdraw their rewards.
These new validators have been drawn in due to the ~4% potential returns for putting up their 32ETH collateral and taking part in the critical infrastructure task of proposing and attesting to blocks.
However, whilst growth in the validator set could contribute towards decentralisation (although likely not if it’s via liquid staking providers like Lido who already hold a large % of staked ETH), there are alarms being raised about the drawbacks of a very large validator set.
The alarm was raised back in September by Ethereum core developer dapplion and protocol meeting coordinator Tim Beiko that at the current rate of deposit queue growth, then
“…the share of ETH supply staked will reach 50% by May 2024, 75% by September 2024, and 100% by December 2024.”
This would be notable growth from the ~24% of ETH that’s currently being staked and could see a reduction in network activity if too many people are staking and not enough people are using their ETH.
Another potential challenge raised by the growth in validators is network latency and sprawl. This is where the number of validators to share information with is too large to effectively achieve. The result is block reorganisations and missed blocks which reduces transaction throughput and adds chain instability around the tip.
An additional drawback of the increasing validator set size is a reduction in the reward rate for each validator since rewards are distributed to validators through their selection to propose and vote on new blocks. However the more validators in the network, the less likely your validator is to be picked for these tasks and be rewarded for them.
In response to the growing validator set there have been a number of proposed solutions. One which garnered discussion back in June was around increasing the max stake amount per validator from 32ETH to 2048ETH to allow for consolidation from multiple validators into just one. This EIP, 7251, is still in draft with no current plans for inclusion in an Ethereum upgrade. https://eips.ethereum.org/EIPS/eip-7251
However the solution which has been approved for inclusion on the next Ethereum upgrade, Dencun, is EIP7514 — introducing a hard cap on the number of validators that can enter per epoch. https://eips.ethereum.org/EIPS/eip-7514
Currently the number of validators that can enter or exit per epoch (every 32 blocks, ~6mins) is dynamically adjusted depending on how many validators are in the active set. At the current size, this is 13 validators per epoch. However this new EIP will introduce a cap of 8 validators regardless of validator set size — reducing the number of new validators that can enter the staking set from 2,9250 per day, to 1,800 per day.
The aim is therefore to slow down the growth of the validator set whilst additional solutions to reduce sprawl and centralisation can be explored.
My 2sats on the situation are that the projections regarding the % of ETH which will be staked by the end of this year and next is overblown — it hinges upon a 100% full activation queue at all times when in fact that’s now on a downward trajectory on the activation queue and at times recently has been all but empty. However with EIP-7514 slated for inclusion in Dencun then it looks like the churn limit is destined to be activated. Time will tell whether this impacts activation queue entry time if there’s a burst or sustained increase in new validators looking to start staking or any reshuffling of existing stake between providers or staking types.
