Should Lido on Ethereum be limited to some fixed % of stake?

[Chuck]

I’m quite shocked by some of the ‘Ethereum safety most prioritize’ tones. Their logic is: Ethereum 2.0 must be decentralized. Lido holds too much power, which is a threat. We must diminish the threat by limiting its growth.

It sounds reasonable, right? But, have you even thought about the Lido DAO’s contribution to Ethereum 2.0 during the past year and a half? Without Lido, Karken or Binance may already dominate the staking market.

If you guys really care about the health and safety of Ethereum 2.0, WHY NOT TRY TO HELP LIDO DAO TO HAVE MORE CLOSER AND RELAYABLE PARTERSHIP WITH ETHEREUM, RATH THAN TRYING TO KILL IT I think most of the community members will welcome more Ethereans to join our governance process and build a more healthy future. Don’t forget that the Lido DAO is mostly built upon the Ethereum ecosystem and we share the same future.

[FelixLts]

First of all, awesome to see this extensive write-up and the lively discussion here.

I’m personally opposed to Lido hard limiting itself as I don’t believe that this would address the wider issue that lead to this outcome, which imo is based on the delegation-unfriendly native Ethereum PoS design that heavily favors centralized participants with custody over customer funds, especially because of the lack of liquidity until the merge.

Lido realized the need for a non-custodial delegation protocol and liquidity of ETH stake early, build a great product, and is ultimately succeeding because of that. In the meantime, there are many more products coming to market following this demonstration of PMF, and I think it would be unfair to intervene in those competitive dynamics. Imo a native delegation/liquidity protocol design, which would have allowed staking providers to compete directly without having to build a protocol or develop custom APIs would likely have improved decentralization and alleviated the situation Ethereum finds itself in now.

Similarly to Flashbots, I personally see Lido as a strong force and leader upholding crypto/Ethereum ethos and values; building in the open, continuously improving, open for feedback, actively contributing in many ways (funds, education, development, adding additional NOs,…), and in this way benefiting the decentralization of Ethereum at large.

I do believe efforts such as the node operator diversification strategy, evolving the governance model to include stETH, lowering referral rewards, and watching over leveraged staking are viable steps to address some of the concerns and risks voiced in this discussion.

[Chuck]

Technicially, you can’t put an end to an LDS protocol emerging. Even if Lido is self-limited now, there could be a Nido or Mido emerging and passing the threshold we now settled.

My opinion is let’s focus on how to improve Lido’s governance model to make Lido act as a protector of Ethereum, other than a threat.

Optimism’s model inspired me. We can set up a Council House to govern the Lido protocol. $LDO holders can delegate their voting power to one of the council members. $LDO’s power also can be separated into two parts: voting power and dividend right. While a $LDO holder delegated his/her token to a council member(with a fixed locked up time), the dividend right will be triggered and has the right to share part of the protocol income. Actually, this is a democratic election system that works. The public selected representatives and delegated their political power to them to protect their interest.

[Leo_Glisic]

Hey Danny,

Could you help me understand this part better?

This profitability comparison metric works well when all operators are using “honest” techniques, but if any amount of the NOs defect to utilizing destructive techniques such as multi-block MEV or adjusting block release times to capture more MEV, then they skew the profitability target such that honest NOs will eventually be automatically ejected if they do not join in on the destructive techniques.‘’

Are these techniques only accessible to NOs if the operator set is concentrated enough? Or is this something solo stakers will be able to do as well?

If it’s the latter, isn’t this then a general problem anyway?

[knoshua]

Multi-block MEV requires proposing blocks back to back. The chances of that happening scale with the number of validators you control. So primarily this becomes a concern if one operator controls a relatively large share of the total. Secondarily one could also imagine a group of operators building a system that lets them work together to extract multi-block mev and share the profits.

If such a system emerges, it will be more profitable and operators will have incentive to participate, completely independent of any profitability comparison metric that some protocol employs. I view this as a base layer issue, not really a LSD protocol issue.

[Edouard_Stakin]

While I understand the concerns around decentralization of ETH staking, I do not think that Lido should self-limit. Here are the reasons why:

• Self-limiting Lido would just create an opportunity for another liquid staking pool to take over, and this pool might not self-limit itself later on.
• It would be unfair to have a decentralized player like Lido self-limiting itself, while centralized exchanges will likely not. Having a self-limited Lido will mostly benefit these centralized competing solutions.
• Lido really improves the Eth staking experience which can be quite complex to many individuals and institutions, it is a deserved success. Without Lido, the market would still be dominated by a few centralized entities. We would probably not have the current ecosystem of Eth staking DeFi products that we enjoy with stEth too. I want Lido and others to keep focusing on innovation and ecosystem development.
• Lido is not perfect, but it is improving day by day. Its commitment to decentralization (selected diversity of entities, clients, and set-ups) is transparent and exemplary.
• Lido has given a chance to many non-custodial Eth staking providers that would not have been able to gain much voting power due to the current Eth staking flow. Lido on Ethereum is composed of tens of selected professional validators that are geographically distributed, and each validator shares a small percentage of the voting power. Each Lido onboarding wave brings in more decentralization. This is a much better solution than having a few centralized entities splitting the Eth voting power between themselves.
• Self-limiting Lido creates an unhealthy precedent for the Eth staking market and hinders innovation. It might send a conflicting message to some entrepreneurs building Eth staking solutions.
• New technologies are being developed that will allow Lido to become ever more decentralized in the future.

[Anna_Vareniuk]

It’s a super informative and highly intense discussion. Thanks to every participant for their thoughts on a subject.

Accumulating a significant percentage of ETH in one protocol may sound like not the best idea in the world and can be (and is) associated with some risks.

But a one-way street approach, in this case, is unjustified (why does one have to limit oneself while others do not and even could benefit from it and possibly become a threat in the future). Either we reach an agreement with all the stakeholders and publicly announce our commitment or stick to the winner-takes-most market strategy.

The Lido team is fully open and is always ready for discussion. Fully agree with

@FelixLts “Similarly to Flashbots, I personally see Lido as a strong force and leader upholding crypto/Ethereum ethos and values; building in the open, continuously improving, open for feedback, actively contributing in many ways (funds, education, development, adding additional NOs,…), and in this way benefiting the decentralization of Ethereum at large”.

and @Edouard_Stakin statements: “Its commitment to decentralization (selected diversity of entities, clients, and set-ups) is transparent and exemplary.”

Highly possible that Everstake would vote against Lido’s self-limiting

[katernoir]

I am against any self-imposed limit on Lidos stake. If ETH2 relies on artificial limits to address these concerns, the problem should be looked for somewhere else. I don’t think a decentralised network should rely on the social layer to fix the centralization concerns.

As a fact, Lido’s core idea was to fix known design issues in ETH2. The lack of a native delegation feature heavily favours centralised actors with custody over user funds. Lido was built to offer a competitive alternative to CEXs whose share would probably be much higher if Lido wasn’t around.

At the core, Lido always promised to push decentralisation on the beacon chain. Looking at the operator limit, the largest operators run ~1.85% each (7391/398716). Compare that to centralised entities running their stake on a single node operator.

Lido shows that it cares about the beacon chain by being open for discussions like this, self-analyzing through things like the Scorecard, funding various altruistic endeavours via LEGO and aligning with core dev teams of various ETH clients through grants and onboarding them as node operators, which gives them a great income stream.

Of the two options discussed (which I am both against), I feel that steering deposits via fees is preferable to setting a hard limit. However, increasing fees on stETH token holders in total is not a right thing to do and might even have legal issues.
Fees were communicated to existing stETH holders, they may not follow governance on a day to day basis and will expect fees to stay where they are. Even if they did, before withdrawals, it’s not possible for them to exit stETH to a competing product without a loss due to slippage. Only imposing the additional fee on new deposits also opposes the thought of one liquid staking token as these two different fees would not work with a single, fungible token.

The same argument about rising fees for Lido could also be made the other way around: If for altruistic reasons and the hope to better distribute stake on the beacon node, other DAOs could also lower their fees until a certain market share is reached. Obviously this cannot be pushed to 0%, since it threatens node operators capability to maintain a high quality service.

In my opinion, the only argument / concern that needs further addressing here is the concentration of LDO voting power. We should rather discuss how to distribute voting power more widespread to make LDO governance more resilient from hostile takeover.

[itamarl]

I’d be inclined to vote against this proposal. Potential risks have been identified and the focus should be on mitigating those risks rather than patching this through self limiting Lido. It’s a patch that could too easily be reversed and without any assurance that other players would self limit themselves.

Most risks can be mitigated at Lido level (clear suggestions have been made and there is also a path for further decentralisation) but mitigations should also be researched at the Ethereum base layer as nothing prevents another less well intentioned player from taking market shares or validators from colluding without anyone being aware of it.

[butta]

Lido could just buy rETH to compensate for their centralization, and thus keep their network share <30%.

[Chuck]

Maybe we can solve this issue in another way.

The main issue is that the growth speed of Lido is far more advanced than the Ethereum staking industry growth. Ok, how about having a rest and letting the speed down as well as make ourselves much stronger?

We can build up a proxy protocol upon the $stETH mint protocol. Whenever there is a $stETH mint demand, the proxy protocol will be triggered and find the better price of the demand. Yes, just like what has been down on the Ethereum staking front page. The difference is that it is not just a notification, which might be ignored, but on a protocol layer and deeply connected with the core stETH mint protocol. As the discount will most probably exist before the withdraw from beacon chain is available, this function will make us grow more stable.

It might solve two issues in one step. The depeg issue might be better solved because of natural demand from new players. As the depegged gap goes closer, we may not need to put such a huge DAO treasury budget on LP incentive. Without so much direct fund incentive, the growth will be much healthier.

[skozin]

The issue with this approach resides on the legal side of things: minting stETH and buying stETH might be different events tax-wise depending on the jurisdiction (and will definitely be in some). As much as I’d want to ignore the legal side, I don’t think we can since it might have unforeseen consequences for our users.

[vsh]

Heads up - will be crafting a vote on this in the coming days.

[vsh]

I tried to summarise all the actionable options proposed in the thread, and the impression I’m getting is:

• there is a number of voters who would consider “no changes to Lido’s approach” as an option of choice
• there are multiple actionable direction Lido could take otherwise (e.g. stop accepting deposits; introduce deposit fee; increase staking fee; put deposits to rocketpool; all of it can be conditional or time-bound on top of it, leading to countlesspotentially actionable options), but most of them are not concrete yet and it’s not clear what is the strongest option the team should put effort to explore in more details (re: technical feasibility and implementation timelines).

So I propose to have a potential multi-step vote that will start with two-options question:

Should Lido continue as is or it should consider one of the forms of self-limiting?

If you think Lido should consider decreasing inbound stake flow in any shape, form or severity - vote for the second option, and we’ll find out what exactly we should be exploring on the second leg of the vote.

If you think there’s no problem to respond to, or if you think it should be responded in a way that does not involve self-limiting (e.g. governance minimization, dual governance, etc), vote for the first option.

I’ll wait a couple of days to collect feedback on that approach, and if there’s no objections, will continue with the snapshot vote.

[zuzu_eeka]

After an extended discussion, we have moved to vote on the topic of whether or not Lido should self-limit.

https://snapshot.org/#/lido-snapshot.eth/proposal/0x10abedcc563b66b1adee60825e78c387105110fa4a1e7354ab57bc9cc1e675c2

The snapshot starts 24.06 at 9 AM UTC and ends 01.07 at 9 AM UTC
Please, send your votes!

[zuzu_eeka]

The Snapshot ‘Should Lido consider self-limiting’? got the quorum and closed!
Results:
no, don’t self-limit - 80M LDO - 99.81%
yes, let’s self-limit - 156K LDO - 0.19%

Thank you for your participation!

[marcbcs]

A bit late but this is a great idea @Chuck! I just finished reading BTB’s Ethereum chain security post and it made reflect on the risks we are creating for Ethereum Ethereum - Chain by Chain Security (Part 2 of 3)

I hear @skozin’s concerns about the tax-wise events but the user would not know (unless he does some deep research) if the stETH he is receiving is newly minted or bought from a pool, therefore for tax reporting purposes he would still truthfully report a mint not a purchase. Probably best to check with Lido’s Legal Counsel though