Gm Community stakers!
After a while since the original post, I came to the conclusion that the proposal to increase CSM share should be extended with the CSM permissionless phase transition. Please find the updated post below.
CSM share increase and transition to permissionless
Community Staking Module (CSM) was added to the Lido protocol on October 25, 2024, via on-chain vote #180. This was a significant step towards decentralizing the Lido protocol. Since adding the module, more than 225 Node Operators have joined the protocol!
One of the main factors limiting further decentralization of the Lido protocol is the current target share of the CSM. Due to security and safety considerations, the initial target share was set to 1% of the protocol TVL. Today, CSM has already occupied ~0.51% of the protocol TVL and keeps growing.
On-chain data about CSM validators’ performance collected by the Rated network looks promising. CSM is still in the Early Adoption phase, which ensures that most Node Operators are independent participants, each running up to 12 validators. The average validator count per Node Operator is slightly below 7.
Another important topic to consider is the transition to the module’s permissionless phase. This will make CSM fully permissionless and remove the limit of 12 validators per Node Operator.
According to the on-chain data, the number of new operators joining weekly CSM decreased, indicating that most of the addresses eligible for Early Adoption had already joined. Also, several participants indicated they are willing to spin up more than 12 validators, which is currently impossible.
Given that, it is proposed to end the Early Adoption period and transition CSM to a fully permissionless mode alongside a share increase from 1% to 2% to ensure further expansion of the Lido protocol decentralization.
Permissionless operators not eligible for Early Adoption will have to submit a bond of 2.4 ETH for the first validator (compared to 1.5 ETH for the EA participants). This ensures meaningful sibyl protection and protocol safety should significant MEV stealing or slashing happen.
Given the current distribution of the validators between Node Operators in CSM and the acceptable performance of the module in general, it is reasonable to assume that the risks of increasing the module’s target share and transitioning to the permissionless mode can be assessed using the existing risk assessment. In this assessment, the main risk factors are:
- Module capture by a single Node Operator;
- Poor performance of the validators in the module;
These factors only partially hold for the current validators in the module due to the observed diversity of the Node Operators and demonstrated performance. Hence, one can assume that the risk of an additional 1% of the target share can be assessed as the initial 1% in the original research, which was made using the assumptions but not the actual data about the performance and validator distribution in the module which was not available by the time the research was made.
Without strong objections within the research forum discussion, the proposal will be moved to the snapshot stage around December 12, 2024. Should Lido DAO support the snapshot proposal, the actual on-chain vote might take place in January 2024.