Lido is smart contract layer that connects stakers and node operators. Lido as an organisation does not have or runs a single validator.
If your concern is forking you want more validators and more diverse set, not the other way around that this proposal brings. Having majority under 50% cap is easier than having majority under more ETH staked.
Letās assume a total supply of 120 million ETH.
If nearly 100% of ETH are staking then 70% of validators e.g. running software that reverses a tornado cash transaction represent 84 million ETH having the incentive not to get punished for that.
Compare that to a staking rate of 45% where 70% of the validators would represent 37,8 million ETH. So letās assume we have a scenario where the 30% minority of validators are punishing these 70% because of the reversed transaction and there is a fork. What do you think is more disruptive..
a) staking rate of 45%; 55% of a all ETH is neutral with respect to the fork and will engage with the side of the fork that they deem represents the most useful ethereum universe
b) staking rate of 100%; 0% of all ETH is neutral with respect to the fork and all 70% of the eth representing the side of the fork that censored will also favor that side of the fork
Vanilla ETH doesnāt get a vote in a fork, so ā70 % of all ETHā is not a relevant metric.
Staked ETH is what backs the validators and hence what decides forks
Having a higher participation in staking thus broadens the āvoter baseā and increases stability
Both are political crisis that is larger than the dao hack and thinking that a) is notably more survivable IMO is wishful thinking.
Otoh, in the world with larger share of staking the situation is less likely to happen because the staking layer is more representative of holders.
In a world where most ETH is staked indirectly via staking providers like Lido the ethereum research community should just be honest to themselves, accept that proof of stake was overengineering and delegated proof of stake is sufficient.
Because of the 2 suboptimal alternatives
a) most stake is centraly managed by a staking provider like lido and
b) all users appoint central staking providers via delegated proof of stake
the b) one is favorable IMO. Here at least users have a much lower barrier in switching and appointing a competing provider when the current one misbehaves.
They for sure hold certain powersā¦ e.g. on the messaging layer. So they can influence things, censor transactions (e.g. by threatening exclusion from the set of validators). Saying theyāre just a set of smart contracts is dishonest. Theyāre not permissionlessā¦ you canāt just join the set.
Vanilla ETH has an influence in which fork wins. Eth represents real people and mind share. A dead fork is a fork without value
You can post bond and start validating as Lido node operator today.
There are over 400+ node operators that started to participate into the CSM.
I mean we should focus on real issues that this proposal causes not the anti-lido hate. Thinking stETH that did not grow share for 2 years is going to win over Ethereum is short sighted. We want to onboard trillions of assets onchain.
Loudest voices like Bankless and Jasper from RP actually admitted they were wrong about Lido so we should give it a rest and focus on actual multi layer impact of the proposal to whole Ethereum ecosystem.
What a nothing burger of an argumentā¦ you say āLetās not talk about lidoā and call concerns about lido āanti-lido hateāā¦? But where do you give any reason to why not be concerned regarding Lido? It is obviously something to be concerned about.
āYou can post bond and start validating as Lido node operator today.ā
Alright, coolā¦ and what if Lido governance doesnāt want you to be part of their validator set? Can they kick you outā¦? I donāt even need to read their code to know that they can. It is obvious they can. Or how else would they be able to act when operators with bad performance join their validator set, reducing the average validator return.. When rewards are socialized like with lido then someone needs to be in control to make sure performance is held up.
Your entire argument is based on your own assumptions and imagination. The actual financial market does not necessarily operate according to the scenario you envision.
I can provide an example:
Suppose there is a large service provider whose business includes enabling users to stake ETH on its platform, but staking is only a small part of its overall business. In order to attract more users to its platform, it decides to issue a platform token and airdrop a portion of the token supply to platform users.
The airdrop criteria require users to stake their ETH on the platform.
In this scenario, even if the staking ratio approaches 50%, there may still be users who participate in ETH staking purely to qualify for the airdrop. The primary motivation for these users is likely the potential airdrop rather than the staking rewards. They may even accept a staking reward of zero.
Therefore, a situation could arise where the ETH staking ratio reaches 50%, while the staking reward for ETH is effectively 0.
In this scenario, the platform itself does not really care about the ETH staking yield, because its main goal is to attract new users and guide them toward using its other services.
Similarly, users who stake ETH on this platform may not care much about the staking yield either, because their primary goal is to obtain the potential airdrop.
Another group consists of independent stakers. Due to risk considerations, they choose not to participate in this platform and continue using the current staking mechanism. However, once the staking ratio exceeds 50% and staking rewards fall to zero, these independent stakers may be forced to exit because they are bearing costs without receiving any return.
To summarize:
How does your proposed solution address this scenario?
My understanding is that your proposal not only fails to solve this problem, but may actually make ETH more centralized.
As an independent staker, as long as staking provides me with a positive return, I may continue participating. However, if the return becomes zero while I still have to bear additional costs and risks, how many independent stakers will continue to exist?
Excessive focus on the staking ratio can easily cause people to overlook other important issues. Regardless of how high the staking ratio becomes, Ethereum should always strive to remain sufficiently secure.
From this perspective, the key factor is whether the power among the different groups participating in staking is sufficiently decentralized.
There are several critical groups:
In particular, individual stakers and the existing on-chain LST ecosystem should be given enough room to survive and develop. Otherwise, it may actually accelerate the concentration of power.
For example, if the survival space of the on-chain LST ecosystem is squeezed to the point where these participants can no longer sustain themselves, but the market demand for liquid staking still exists, this demand will be more likely to be absorbed and controlled by centralized service providers.
The same perspective can be applied to the scenario I described above. The core issue is not simply the staking ratio itself, but how to maintain a diverse and decentralized staking ecosystem while ensuring Ethereumās long-term security.
Tobi, two points:
The if in āif all ether is stakedā assumption is doing a lot of heavy lifting here and slowly feels like fear-mongering. I still donāt see any strong empirical evidence that this is the equilibrium Ethereum is heading towards. Even in systems where staking is basically the default behaviour (e.g. Cardano, where most users delegate because there is basically nothing else you can do with ADA), participation has not reached 100%.
The second point jumps a step. The observation that āLido governance controls which operators participate in Lidoās validator setā is correct. But the conclusion that ātherefore increasing Lidoās share creates Ethereum-level centralization riskā requires a counterfactual.
It does not magically become solo staked. It goes to Coinbase, Binance, ETF issuer/custodians, DATs, or other delegated staking providers.
Those alternatives have their own (often stronger) centralization and political risks.
Also, control over Lidoās validator set is not the same as control over Ethereum consensus. Lido can select/remove its operators and influence stETH governance, but it cannot force non-Lido validators to follow it or unilaterally change Ethereum.
The strongest version of your concern is probably: āEthereum should be careful about any single coordination layer becoming too large.ā That is a valid discussion. But the relevant comparison is not Lido vs a world where everyone solo stakes. That world does not exist.
The real question is: what coordination layer do we want the marginal ETH to flow through?
Of course 100% is a theoretical thought excersice. It will never literally reach 100%, but can we appreciate - following my example - that thereās a reason why >50% of staked eth in total can approach a less optimal outcome than we have now? My main point is total eth staked being significantly over 50% is bad - especially when concentrated among a few staking providers (where there are at least soft powers to influence behavior). And I donāt think the assumption that given the current staking curve weāll get there is unlikely and dismissable.
I donāt have a problem with the people behind Lido, Lido is just the most obvious example. So yes, whether the staking is done by coinbase, lido, doesnāt matter. Thatās precisely why Iām arguing for a change in the staking incentive structure (this proposal) that overall limits the amount of staked ETH. Not specifically the amount of ETH staked by Lidoā¦
I agree with the problem EIP-8363 is trying to solve. My concern is that we still donāt really know who exits first when staking rewards fall. Solo stakers have fixed costs, while large operators benefit from scale.
Thereās also the demand side. ETH staking APR is already around 2.5%, well below long-term US Treasury yields. If we push it materially lower, ETH may become less attractive as a yield-bearing asset. That trade-off deserves more consideration too.
So Iād just like stronger evidence before making a major change to the issuance curve.
I think itās not ETHs job to provide people with yield. Itās Ethereumās job to deliver a useful network to the world. Some staking opportunities for people are just a by-product of that. Ultimately Ethereumās success will be based on how useful it is to the world and the special thing ethereum has to offer is an uncensorable, neutral, decentralized network where everyone can run code. Not being a yield bearing asset for de/trad fiā¦
But this is exactly the part Iām not convinced about: what do you think the validator set looks like after this change, and why is it better?
Why should we expect that the marginal ETH leaving the ā>50% stakeā comes from the large providers we are concerned about and not from the smaller operators with higher cost structures?
Reducing the quantity of stake and improving the composition of stake are two distinct goals.
Iām not making the argument that the validator set will look much better in terms of share of homestakers. I anticipate the distribution will mostly stay the same, maybe relatively there will be slightly fewer home stakers. Those who home stake are doing it out of conviction to a certain degree anyways.
So, Iām not expecting that and it is not my argument.
Thatās exactly why I am for this proposal because it reduces (or sets bounds for) the overall quantity of staked ETH.
See my argument regarding forks in that regard.
OK; I think weāve reached the core disagreement.
I disagree consensus power is simply a function of validator stake āweightā. Social consensus, users, applications, exchanges and economic activity all matter in a fork. Staked ETH matters, but it is not the whole picture.
Where I am completely unconvinced is the confidence around the assumed outcome. Statements like āhome stakers are mostly conviction drivenā or āthe distribution will mostly stay the sameā feel highly theoretical and dismissive of how markets actually behave.
We should look at where stake is actually coming from, who is marginally entering/exiting, how different cohorts respond to lower yields etc. Otherwise we optimize for a magic theoretical equilibrium rather than the Ethereum we have in reality.
We canāt know. Home stakers are by design anonymous. And thereās no certain way to proof that youāre a home staker. Big stakers can split their stakes, use VPNs, multiple IPs, etc.
So given we canāt know, calling for measured evidence for things that are inherently unmeasurable (at least reliably) - feels like concern trolling and stalling. You like the status quo and block any change with calls for demonstration of the undemonstrable..
We need to trust our logic and intuition here a bit. And I already conceded the point that this wonāt increase the relative amount of home stakers.
It is pretty straight forward to say:
An opposing validator minority must have the chance to win the economic fork battle. It can only do so if a significant amount of ETH can join that fork (to win >50%).
fully agree with this, how can you discount āCTā? X is part of ācore governanceā
Well I argue that āwe canāt knowā is not a reason to make a major monetary policy change.
The burden is not on critics to disprove a hypothetical equilibrium. The burden is on the proposal to show that this equilibrium is likely and that this intervention improves the outcome. Itās optimizing for a theoretical failure mode with unproven assumptions about how the real system will evolve.
I consider this proposal dead on arrival.