In anticipation of the Merge update in September 2022, skeptics warned it could lead to centralization within the Ethereum network. Now, a year later, Danny Ryan from the Ethereum Foundation admitted a “corporatization, a centralization, and a systemic threat” within the network. Even Buterin acknowledged these concerns to some extent in a recent blog post outlining potential approaches to mitigate such risks.
The three ‘doors’ in question are the three validators who now control over 85% of Ethereum’s liquid staking validation.
Notably, in May 2022, Vitalik believed no single liquid staking service would exceed 15%. Now, with three services collectively holding 85%, it’s evident that those expectations have been far surpassed.
(Glassnode’s report: stETH’s Effect on Ethereum)
Worse, other areas of Ethereum’s proof-of-stake system also show signs of centralization, becoming a serious problem.
The risk of liquid staking providers like Lido being attacked by the government or malicious actors is rather low (in Lido, operator nodes are spread between ~30 companies) but should not be taken lightly. Besides, there’s also a very material risk of the “cartelization” of Lido, as CoinShare’s recent report points out.
A long-term solution, meanwhile, remains unclear. Just as the dynamics of liquid staking were not fully grasped when staking was initially introduced, any fix incorporated into a future hard fork must carefully evaluate its potential long-term implications and unintended consequences.
Some shorter-term solution, however, is in the works. @Dapplion, for example, proposed EIP-7514, which would hard cap (of either 16, 12, 8, 6, or 4) the number of new validators that can join per epoch, extending the time it takes to reach staking milestones like 50%, 75%, or 100% of staked ETH. However, this would assume a scenario where the activation queue stays full, and staking demand remains unaffected by any self-balancing decay mechanisms.
This stopgap measure doesn’t address any underlying issues but gives the Ethereum developer community enough time to devise a lasting solution for staking dynamics.
From the ‘GitHub governance’ side, Vitalik Buterin has shared a blog post outlining his view on integrating certain protocols into Ethereum’s core code as opposed to keeping them “on top” of the blockchain.
The protocols Buterin discussed include ERC-4337 for account abstraction, ZK-EVMs, private mempools, code precompiles, and liquid staking, each of them, as he concludes, involves “a complicated trade-off” that will persistently evolve.
Buterin expressed concerns about the concentration of Ethereum’s liquid staking providers, too. He pointed out that Lido and Rocket Pool, two of the top providers, have some safety mechanisms in place, though these may not be enough.
Rather than relying on “moralistic pressure” to encourage using a more diverse set of staking providers, Buterin suggests making some changes to Ethereum’s protocol, making liquid staking more spread out and less controlled by a few big players.