We have discussed these options with the Blockdaemon team and they preferred Method 2 because of its simplicity & APR implications
You can think of the APR boost as Amount of MEV rewards received * All-in staking APR (network emissions + MEV) on a continuous basis. A simple way of thinking of comparing the two methods is compounding 100 dollars (and the proceeds from re-staking) daily or weekly at a ±5% APR for 365 days and seeing the difference with Method 1. It works out to be as follows:
Method 1: $36,500 ( = $100 * 365)
Method 2: $37,417.60 with weekly compounding or $37,425.27 with daily compounding
The benefit of Method 2 over Method 1 is $917.60 (weekly compounding) or $925.27 (daily compounding), which is a ±2.5% boost over 1 year, and a ±8% boost over 3 years. So if the all-in APR without compounding is 10% where MEV is half (5%), with compounding the all-in APR is 10.125% where MEV is 5.125%. By the end of year 3, it will have grown to 10.4% all-in APR, where MEV is 5.4%. The boost will be bigger if MEV rewards work out to a higher APR. Note that this calculation does not account for the manual reinvestment of rETH2 into sETH2, which can offer a considerable additional benefit.
In terms of competitors, I believe Lido are planning to reinvest MEV rewards as well, while Rocket Pool are not. Reinvesting would mean that we stay par or beat Lido and remain ahead of Rocket Pool in terms of APR.
Finally, not sure about the tax implications since for various jurisdictions taxes work differently.
If you want to cross-check the calculations for the compounding effect, see this simple calculator.
I think the fact that [1] we don’t pay gas fees and [2] we don’t lose 10% off the cuff makes it pretty compelling. I’d hate as whale to get 1mil back in ETH and then have to pay 10% of that back to StakeWise. I’d rather get 1mil back in rETH2 and then swap / play around as needed.
cheaper imho to go that route vs the loss of the 10% of MEV rewards.
Hmm not sure what you are referring to here StakeWise charges 10% on the rewards, so 10% of MEV is going to be collected by the protocol anyway, whether as ETH or rETH2.
Would the team (@kiriyha / @tsudmi) be willing to share a high level pseudo-code implementation of what would be modified for Option 2? doesn’t have to be detailed, just highlight what would be modified and/or added?
The technical implementation (roughly) is calculating total rETH2 supply as
Total rETH2 supply = Total Validators Balances - 32 * Number of Validators + Total MEV rewards historically accrued in the coinbase address
So whenever updateTotalRewards is called, rewardPerToken will calculate the amount of validator rewards and MEV that should belong to each sETH2 holder and update the balances of rETH2 accordingly. Hope I got this right @tsudmi !
Will sETH2 holders receive the staked ETH and the rETH2, or only the rewards generated by the staked ETH?
Glad to have you back sETH2 holders will receive rETH2, which will represent both the newly staked ETH and the rewards from it.
I guess the payout to sETH2 holders will happen in a proportional way to ones sETH2 holdings?
Since MEV is earned proportionally to sETH2 holdings, then reinvesting & distributing MEV will also happen proportionally to sETH2 holdings. So correct!
But how can this be displayed in the UI and is rETH2 then also getting a yield bearing asset? Is rETH then having its own APY, otherwise I could sell the rETH2 but still get the boosted APY, no?
Not really - rETH2 does not become a yield-bearing asset since the rewards (rETH2) are still being distributed to sETH2 holders, despite a portion of them being minted against ETH principal. And as for selling rETH2 to get boosted APR - that’s true, you can go one step further and compound again. In short, sETH2 is still king.
Hey kiri, this part isn’t clear, please could you clarify?
example: 100 sETH2 staked for 1 year creates 4 rETH2 staking reward + 6 ETH MEV reward (which is restaked daily) so after 1 year we end up with 106 staked Eth in total (made up of 100 sETH2 and 6 staked Eth) and the 4 rETH2. Shouldn’t the 6 staked Eth therefore be given out as sETH2? As it’s literally staked Eth and will be yield bearing - literally sETH2, but you’re saying here it’s better to convert into rETH2 which isn’t yield bearing? In that case where will the yield from the staked Eth underneath go?
If i’m understanding this correctly this will cause an imbalance between actual staked Eth and sETH2 in existence.
If i’m understanding this correctly this will cause an imbalance between actual staked Eth and sETH2 in existence.
It’s correct that sETH2 supply will be less than the amount of principal in the validators, but this will be compensated by the supply of rETH2 amounting to total accrued rewards + total accrued MEV, so sum of supply of sETH2 and rETH2 is always equal to total validators’ balance.
You can think of it this scenario: whenever ETH is staked, it earns rewards. These rewards are distributed as rETH2. Now, whenever you distribute the rewards, they are distributed proportionally to one’s sETH2 holdings. Suppose MEV was distributed on the validator level (simply as ETH rewards that are not accessible ie cannot be immediately reinvested) - we would distribute them as rETH2. No biggie.
Now since MEV is distributed on the execution layer, these ETH rewards are accessible, and all we do compared to the scenario above is also stake them on the validator level to accrue even more rewards. It’s akin to compounding, and while it’s not as straightforward as before with sETH2 and rETH2, the main logic remains the same: sETH2 is the interest-bearing token, rETH2 is not. I hope this makes sense G!
Hi everyone, I’m late to the party but thankfully not too late. I have a couple of questions.
In your post you mention three methods but only two are described. Just a typo or did you miss a third method?
With regards to the rETH2 distribution and your explanation here: This means that you collect MEV and general transation fees on the EL until you reach 32ETH and then just distribute a new validator to one of the node operators which ultimately leads to an accrual of more rETH2 because more validators are running. Did I get it correctly?
My third question goes into a totally different topic and might be relevant on a different forum post. Therefore my point can be split in two topics:
3.1. When should we include mev-boost or a different tool on the technical level? My question refers to the fact that mev-boost is a very new tool and might come with issues as it is a) another software component which can fail in general and b) since it is pretty new (is still in alpha I think) it could come with general issues that might lead to bad consequences. This was also mentioned in the ACD call from yesterday (19th August) so the recommendation from Danny Ryan was to maybe do a smooth transition from non mev-boost setup to mev-boost setup.
3.2. The second part is the current discussion about censorship. I think as DAO we need to decide whether it matters if validator operators will access censorship resistant or censored relays with mev-boost. I have a very strong opinion about that topic but maybe this needs to be discussed in a separate post.
Total staked Eth = Seth + reth…but over time not in the proportion you might think, which is confusing but results in apy growing (vs overall beacon chain) for Seth holders.
Playing this forwards to when withdrawals are open, seth redemptions will lead to further proportionate imbalance of underlying Eth staked vs Seth supply. As redeeming 1 Seth will obviously get 1 Eth, but that Seth has a claim on >1 underlying staked Eth. This will drive the price of Seth ahead of Eth over time, incentivising people to get Seth, either by staking fresh Eth or redeeming reth to buy Seth. To use some trendy crypto terminology, we could in effect say we have generated a ‘flywheel’ for Seth adoption… this is great news for the protocol.
You are pointing at an interesting issue actually.
But first: once withdrawals are open, you will be able to redeem both sETH2 and rETH2. The amount of sETH2+rETH2 = the amount of staked eth + consensus layer reward eth + execution layer MEV eth (restaked).
So 1 sETH2 would have a claim on exactly 1 eth, same for each rETH2 also 1 ETH. At least the way it would work now.
It looks likely that the consensus layer reward eth will be claimable in an earlier fork. If so they will be claimable before sETH2 and restaked MEV eth.
In effect at some point the need for rETH2 disappears, at that point consensus layer rewards could be restaked immediately.
Might it then not make more sense to not issue sETH2 for the restaked MEV rewards? Instead as your idea would suggest let them accrue to sETH2, resulting in 1 sETH2 becoming worth more than 1 ETH. OR instead of issuing these rewards as rETH2, issue them as sETH2? In which case sETH2 remains at 1 ETH, but you just get more of it.
At some point the need for rETH2 disappears and we can follow the same logic with those rewards as we did for the restaked MEV rewards.
I actually like this approach and it allows for the compounding effect to happen automatically. Take the MEV rewards and mint sETH2 and not rETH2 - so while initially you don’t get additional rewards - long term, it does increase your APR because you’ve already compounded the rETH2 (as sETH2).
Does this tokenomics model make sense? (not a financial whiz by any means). Whatever gas we’d have to spend minting the rETH2 rewards to distribute, would be the same for sETH2?
After some time to consider both options and leaning on the advise of others. I would prefer option 2 of restaking & distributing MEV rewards as reth2. It utilises and shows the benefit of having 2 token model. Allowing compounding and creating more opportunities to increase TLV of the pools.
Hey Tobias! Great stuff on all points - 1) I fixed the typo, 2) that’s correct, 3) the discussion about the technical side of MEV extraction will follow.
Thanks for your thoughts here @ottodv - as always a strong contribution
The main reason why we’d rather not make sETH2 a repricing token is that it becomes difficult to manage liquidity in the exiting sETH2/ETH pool (since sETH2 value will start trending up) and because it makes the tokenomics a tad more complicated (having both the repricing mechanism and the rETH2 accrual mechanism is tough to explain to new users). Given that it doesn’t result in a higher APR than the 2nd method, we decided not to mention it at all.
highest possible APY for Seth2 holders due to frequent restaking of MEV rewards
I actually like this approach and it allows for the compounding effect to happen automatically.
It doesn’t unfortunately - still requires us to deposit MEV rewards into the PoolDeposit contract & distribute the rewards. Also note that it doesn’t increase anything in the end - the “compounding” on the sETH2 level and the rETH2 that your “extra” sETH2 generates are both captured by the Method 2 anyway
Thank you all for your responses & thoughts! We will now be proceeding to a Snapshot vote on the matter to solidify the DAO position on this important topic
Thank you for bumping it to my attention.
