In our current system Rocket Pool as an entity doesn’t profit directly from either node operators or rETH stakers. This means that Node Operators can profit directly from their counterparts, by just waiting for the commission rate to go higher, raising the overall average fee for all rETH stakers.
Problem:
With the current commission system, there is a gaming aspect where many node operators are trying to aim for the highest commission possible and the lowest gas cost to launch their minipools. Allowing for an extremely volatile commission rate structure and curve allows for such a game to exist. This leads to high amounts of ETH deposited in the Deposit Pool not allowing for more rETH participants at those specific times and for a game that is always in favor of Node Operators.
My Solution:
Create a commission rate structure in which the current commission rate is held by an average of the past ‘n’ minipool deposits. This ‘average’ is kept as the starting point. The next minipool must then be a certain distance from the average to be considered a valid commission rate.
There are then 2 curves that’ll help determine the commission rate for the next minipools. The first curve will be for increasing the commission rate and the second will be for decreasing the commission rate.
depending on how the curves work you could have the curve that increases the average at say a rate of 0.25%, but when the deposit pool is lower than a certain amount the average will be affected by a 2nd curve to where it’ll decrease by 0.50% (2x the increasing rate) that way it’ll actually target 10%, and act as a balancing mechanism.
I believe that if the commission rate is supposed to “target” 10% then we should have a system similar to this. Not one that is as volatile one as we currently have.
Many different solutions can solve this problem and I believe this has already been talked about a few times in the discord.
Thanks @Xer0! You beat us to it - the team were going to post the issue this week.
The current commission system is designed to provide a supply/demand mechanism that attracts new node operators when the deposit pool starts to fill. As you said, there are some issues which have developed.
You make a great point that node operators are essentially in control and are increasing the amount paid by rETH holders - this is perfectly rational and understandable. The low ceiling, commission hits max around 1.8k, means that we can get UX issues for large ETH deposits.
Additionally we believe that the supply/demand mechanism is hindering Rocket Pool’s growth (slightly) because node operators are holding back. More insidiously, it hides the true node operator capacity - we only get a hint of it when conditions are perfect (gas+commission).
As you said, the community have discussed this a few times in Discord. The team have discussed it with the community and internally. We believe we should work with the code that we have, rather than reinventing the commission system: lower complexity = lower risk + faster time to market.
Our preferred solution is to lock the commission rate to 15%. By locking the commission rate, it completely removes the gaming and just leaves gas price as a factor. This number is open for debate but we feel it balances: the need to attract new node operators vs rETH APR. Ultimately we will need to lower the commission rate towards 10% but we can do that progressively over time.
100% agree on both the proposed solution and 15% commission rate. This will greatly simplify the minipool deployment process for new NOs, since they no longer need to find the perfect optimization of low gas + high commission.
Only recommendation that I’d have is to do this fairly quickly, since it’s a simple configuration change. Gather feedback for 2 weeks, share the plan, and execute the change. No need to get stuck in analysis paralysis or audit backlogs for this one.
One possible idea to take it a step further might be to remove the 32 ETH deposit option from the Smartnode CLI, or perhaps hide it behind an argument (rocketpool node deposit -a 32). This isn’t exactly related, except that it removes yet another decision that a new NO has to make, and the 32 ETH option makes no sense assuming the DP has >16 ETH.
If the commission rate becomes fixed, it would also be worthwhile revisiting the delay on transferring newly minted rETH to other addresses.
Currently, this is set to ~24 hours to prevent flashloan attacks and manipulation of the deposit pool to affect commissions.
It would be nice to see the delay reduced to as few blocks as possible in order to maximise user friendliness while maintaining the security of the protocol.
I feel a fixed commission would also help the marketing of Rocket Pool as it removes a variable that might not be trivial to explain in an elevator pitch.
I’m a +1 on 15% flat commission (for now), removing the 32 ETH deposit, and shortening the 24H rETH transfer delay (an hour seems reasonable?). I think these are all good ideas.
Count me in as a supporter of the 15% flat rate and freeze shortening. However, I am interested in hearing more about possible downsides. Like if the deposit pool is swamped, will the backlog create a bad UX that reduces demand or drives large depositors to centralized stakers without that issue?
By having a fixed commission, there is no driver for node operators to create minipools when the deposit pool is getting full. This was the intention of the supply/demand dynamics in the first place.
What we expect to happen is that node operators will create minipools when it suits them - when gas is low. This means we should have access to capacity earlier.
That said, all of this is economic theory until we test it out. I believe a fixed commission will resolve some of the issues we are facing but if it doesn’t we can always roll back to a supply/demand system.
I also support the fixed NO rate. I believe the slight incentive over the average commission rate of 13.6% (source: https://www.rp-metrics-dashboard.com/) will be useful at attracting NOs to form minipools without having to “game” the deposit pool and form validators only when the pool is full. This will also improve the staking efficiency of RP by allowing all the ETH in the deposit pool to be staked without a disincentive to node operators to stake only when the cap has been reached.
Further, this affords an opportunity for NO to deposit even when the pool is empty during low gas times but still guarantees that the NO will obtain a 15% commission when the NO queue advances to stake.
I believe it would only affect newly created minipools. While it may sting a little now, you’ll have the option to restake as a 15% pool after withdrawals are enabled, while people with >15% can choose to keep them as long as they want, but ultimately as someone who myself has a <15% minipool, I think this is better for the protocol. I actually don’t know if gas fees would even make it worth it to upgrade. It’s all about those RPL rewards anyway!
I agree that a fixed 15% is a great balance for our current needs and support this idea! The other alternative to help scale and smooth out the commission is to increase the Deposit Pool size.
I do have a question: will we also need a 2-week testnet for future adjustments to a fixed commission?
I think the 15% fixed commission should be implemented as soon as possible.
The potential lack of a dynamic, negative feedback incentive structure is slightly concerning, but it’s obvious the relationship between NO and liquid staker is not ideal and needs to be rectified. As the protocol expands and the deposit pool grows, perhaps the community should consider returning to a dynamic commission calculation.
Additionally, I’d like to tag along and signal my approval for removing the 24 hour block on rEth movement, as well as the removal of the explicit 32 Eth deposit option for NOs.
Node operators want higher commission rate. Isn’t that the problem. The curve should have larger range. Going +20%. Why are we making judgements? the market is smarter. Widening commission rate would be us making less judgements. I am a rETH holder. rETH has so many benefits. Liquidity farming, chilling not worry about node, market making, leverage trading against hyper inflating fiat currency ect. I would not care if I was getting charged +50% commission on my rETH. Hopefully i’m not interpreting the problem wrong
Decent points @sufaccountant. @langers - Is there any consideration to increasing the upper bound on the deposit pool to allow a market based reaction to rETH demand? Seems like that may drive down rETH holder return a bit, but could be a solve to the problem as well. Is there any concern with pegging the commission rate removing the market’s ability to react?
I agree with you, this stings the early adopters/supporters. I have a few 5% minipools. How many people who are just running one or two minipools have 5%? Its a bit of a smack in the face.
Of course node operators would love higher commissions but I don’t think it is a problem. We are growing rapidly and there are many other factors other than commission - RPL rewards are more enticing than commission for instance.
If we increase the node commission above 20% it will impact rETH APR substantially - relative to competing liquid staking tokens. At some point, rETH is not able to compete. Because minipool commissions are locked in, unwinding the average could take a very long time - which is why it needs to be managed.
