Collateral Provision & Restaking

[merklefruit]

Credible commitments are based on some form of credibility. In an untrusted context, this credibility can come from economic collateral, which essentially means depositing any on-chain asset in a smart contract and being subject to protocol-defined conditions for managing rewards or slashing conditions.

In particular, when validators in PoS chains like Ethereum want to provide collateral, restaking comes as a very capital efficient way to do this. For one, the existing large economic security of staked ETH is "sitting idle" when validators don't have to directly fulfil proposing or attesting roles, and secondly the risk appetite/tradeoff for some of these validators might be higher than what the core protocol offers by itself.

For what concerns Bolt in particular, economic security can come from anywhere: simple ETH staking, restaking pools, LP positions, NFTs, or anything else. However, for the sake of simplicity we would like to have all parties agree on what it means to have a "good collateral" that anyone is willing to accept in case of commitment fault attribution rewards or similar.

This is why the current plan is to start with the following setup:

• allow all validators to register by authenticating with their BLS key
• allow them to bind collateral to the operators that will be serving commitments
• this collateral may come from:
  • natively staked ETH in a smart contract (akin to the beacon chain deposit contract)
  • restaked collateral vaults, from existing standard protocols such as Symbiotic and Eigenlayer

Bolt as an AVS

Bolt is not a normal AVS, because there is an essential extra separation between collateral providers and actual Ethereum validators. This is because rewards in Bolt are accrued directly to proposers, just as in MEV-Boost, and not to a separate vault that implements a revenue share model between validators and commitment operators.

On a trust-relationship note, it's also important to remember that Bolt Operators need to have full control over what blocks validators will actually produce. So it is not likely that general-purpose "restaked collateral vaults" will opt-in Bolt without also controlling an amount of validators themselves.

Who should secure Bolt?

We expect anyone that is running validators to provide collateral in any form to Bolt. This mainly includes:

1. Staking pools, including pools with several node operator clusters behind them
2. Solo stakers and small pools that might appreciate a simpler opt-in experience without the complexity of a collateral vault