Allow for time-delayed rETH burning at protocol rate using validator exits
Suggestion:
Currently rETH can only be redeemed at the protocol rate when sufficient liquidity on the EL exists. If it doesn’t, rETH holders that need to exit are forced to sell at a discount on secondary markets. There is no direct mechanism for rETH holders to exit validators.
While node operators can take advantage of a discount of rETH on secondary markets by exiting validators, this only becomes attractive when the discount becomes large enough and therefore only serves as a worst case insurance from the perspective of rETH holders.
I’m suggesting to add a mechanism that lets rETH holders “request unstaking”. This could then lock the rETH for a period of time. In regular intervals, validators would be exited (would require EIP 7002 and Saturn 2) to fullfill these unstaking requests. At that point the rETH is burned and ETH can be claimed.
Advantages:
The value proposition of rETH is improved by removing any risk of not being able to exit the position or having the risk of a loss on exit
This may attract new holders that need a stronger liquidity guarantee but might not end up using this exit mechanism regularly
Disadvantages:
The value proposition of being a node operator is hurt, since there now would be a risk of validators being exited when the operator would prefer to keep them running
Open Questions:
How does the unstaking mechanism work in detail and how do we ensure it can’t be exploited?
How are exited validators chosen? FIFO? Random?
Note: I understand this submission is ineligible for compensation in the event I am chosen as a review volunteer.
rETH burn queue with smoothed Node Operator opt-in random exits
Suggestion:
Credit @knoshua for describing the problem statement
Currently rETH can only be redeemed at the protocol rate when sufficient liquidity on the EL exists. If it doesn’t, rETH holders that need to exit are forced to sell at a discount on secondary markets. There is no direct mechanism for rETH holders to exit validators.
While node operators can take advantage of a discount of rETH on secondary markets by exiting validators, this only becomes attractive when the discount becomes large enough and therefore only serves as a worst case insurance from the perspective of rETH holders.
Thus, I suggest we create an incentivized opt-in system to encourage people to allow their validator to be randomly exited periodically for matching in the burn queue. The system would have the regular minipool/megapool, a new rETH burn queue, and a new minipool/megapool that includes the chance to be exited.
To make this system stable, we must split the arb value amongst three parties: the rETH holders taking the time exposure risk of buying discounted rETH and waiting in the queue, the node operators in the opt-in validator exit minipool/megapool, and the node operators exited at each interval. The flow to each of these parties should be adjustable so that it can remain fluid with market dynamics.
Advantages:
rETH is a better product as it has a low risk exit route regardless of market value, building trust and reducing the likelihood of depegs happening in the first place.
More holders - some may require this form of exit
Reduced strain on liquidity - time insensitive exits can wait for exits instead of relying on DEXs.
Disadvantages:
added protocol complexity with more decisions to be made by the node operator
reliance on market incentives may create situations where not enough node operators opt-in
Open Questions:
How do we design this?
How much value do we direct to each faction?
Do we need to incentivize the exiting node? If so, should we model that value flow after gas costs?
Is this system stable after long periods of premium in which the bonus APR for the node exit group is ~0?
Can we model the impact of variables like gas, time, and lost APR from exit queues?
Do we even need any value flow to rETH outside the ability to exit on peg while the market suggests a discount?
Implementing a smoothing pool lotary Background: Many node operators don’t participate in the smoothing pool for fear of missing out on a large EL reward block. The lottery will act as a hedge against the risk of missing out on a large EL reward block. Recommendation: Implement a smoothing pool lotary system that works by minting a lotary ticket against the smoothing pool. If the operator happens to propose a block with a reward greater than the value of the ticket, the operator can burn his ticket and collect the entire block reward (proportional to his bond). Proceeds from the mining of tickets are allocated to rEth holders. Tickets may have an expiry date and must be renewed. The renewal fee may be less than the original ticket price. Example: An operator deposits 0.1 ETH in the smoothing pool and mints a NFT lottery ticket. If he later proposes a block with an EL reward of 1 ETH, he can burn his ticket and collect the full reward proportional to his bond of 0.25 ETH. If the operator proposes a block with an EL reward of 0.2 ETH, he does not need to burn his ticket because he will only receive 0.05 ETH reward. Advantages: Additional income for the smoothing pool from minting tickets; Incentive for NOs running a traditional LEB8 to join the smoothing pool and don’t fear missing out on a big block reward; More node operators joining the smoothing pool and the additional income, means higher rETH APR and incentive to mint rETH. Disadvantages: Slightly less income for all operators in the smoothing pool; Requires smoothing pool rework.
Suggestion:
The rETH APR displayed on the website is a static value. Users may appreciate a dynamically updating APR site that also displays historical data or clear comparisons to cbETH (and maybe stETH once we get our APR up there). Offer a mint button.
Advantages:
Seeing a live updating value may attract users. Seeing it outperform the competition may draw them too.
Disadvantages:
We may sometimes end up looking bad.
Open Questions:
Should APR be the only thing? Should it mention something like live node operators registered count?