Proposal: Introducing $LDO Staking

TL;DR: Introduce $LDO staking module and buyback program. Allow token holders to stake $LDO in exchange for a proportion of Lido DAO revenue (via an $LDO buyback and distribute program). Increase $LDO utility by having $LDO stakers serve as insurance providers of last resort.

Abstract

Over the past few years, Lido has grown from an early stage DeFi protocol to the dominant leader in the liquid staking space with over 12b in TVL.

Despite Lido’s success, $LDO token holders don’t directly benefit from the revenue generated by the protocol and $LDO has no direct utility. These points are a principal concern for current and prospective token holders.

This proposal aims to catalyze a discussion and propose a solution to the utility and value-accrual issues facing $LDO.

Proposal

Goal: Introduce $LDO token utility and align protocol incentives between all Lido associated parties.

Exact Parameters/Mechanism:

  1. Revenue Share
    a. Redirect 20-50% (toggleable parameter based on governance) of future Lido DAO revenue from the protocol treasury to stakers of $LDO
    b. The Lido DAO currently has a treasury exceeding $280m at the time of writing
    i. Since DAO operating costs are an estimated ~$16m annualized, the Lido DAO currently has an estimated runway of 17.5 years, suggesting minimal-to-no operational impact from revenue redirection.
    c. Distribute Lido DAO revenue to $LDO stakers in $LDO tokens weekly through a buyback and distribute mechanism (exact mechanics will be determined separately)
    i. This mechanism allows all $LDO holders (not just stakers) to benefit from the revenue generated by the protocol
    ii. All stakers will ‘receive’ tokens on a weekly basis. All ‘earned’ tokens will be vested for 6-months before distribution.
    iii. I propose the buyback be executed on-chain via a VWAP or TWAMM

  2. Set minimum size for the LidoDAO insurance fund
    a. As previously discussed, set minimum size of the Lido insurance fund at ~6k stETH
    b. If insurance fund reserves dip below the minimum threshold due to a slashing event, the protocol will redirect all revenue from stakers to the insurance fund until the minimum reserve is reestablished.

  3. Staking Terms
    a. 14-day unstaking cooldown for all stakers
    b. Potential loss of up to 30% (industry standard) of staked $LDO in the event of a major slashing event
    i. Similar to stAAVE, staked Lido will be auctioned off to the market
    ii. Loss waterfall will initially be as follows: insurance fund (once exhausted) → $LDO stakers (up to 30% staked tokens slashed) → socialized loss via stETH holder negative rebase
    1. While this may make staking $LDO seem high-risk, I anticipate that the Lido staking program will be improved in the future to include NOs, especially after permissionless validators and DVTs are integrated with the staking router. In this case, I’d anticipate NOs will take most of the slashing risk themselves.

Simulation/Analysis

$LDO Staking Estimations
Note: Assuming staking growth Y + 1 to be the same as past year (58%), and following year (30%) , and Lido mkt share to be 30% of all staked ETH.
Source: TokenTerminal, Dune

FAQ/Concerns

Q: Wouldn’t this cause $LDO to face additional regulatory scrutiny (especially from US regulators)?
A: Perhaps, but the proposed model is an industry standard, with several DeFi protocols adopting a similar standard over the past 5-yrs including Maker, Aave, and Yearn.

Q: Is this proposal likely to increase the potential of SEC enforcement action?
A: The Lido DAO is a fully-decentralized organization with no legal entities. Lido DAO’s only tie to the US is the employment of US-based persons as contractors. Lido has not received any enforcement or warning communication from any US-based regulatory body to-date.

Q: This proposal doesn’t quite address additional utility for the LDO token?
A: Token holders are more aligned with the success of node operators and the general success of the protocol. I also believe that node operators need to be more aligned with the protocol :slight_smile:

Next Steps

After a 7-day period of discussion, I propose moving this proposal to a signaling vote on snapshot.

I am in the process of drafting a second proposal that will impose an obligation on NOs to stake $LDO. Staked $LDO will ensure validators have skin in the game in case they get slashed. IMO this is a necessary strategic move to fortify the alignment of incentives across the protocol. I will share the new proposal in the coming week.

I’m excited about the future of $LDO and believe that this advancement will help underscore the protocol’s commitment to $LDO token holders as well continued growth and resilience.

10 Likes

There’s definitely a good case to be made for something like this, but just wanted to flag early comments to avoid any confusion down the line, and focus on a productive discussion:

$280m is counting LDO at current market values. This is bad practice, highly misleading and does not reflect the actual value of the treasury. The treasury currently has 20k ETH, 10.7k stETH, 10.3m DAI and 2.3m USDT. Please stop counting the value of LDO in treasury. We rely on @Hasu mental model for treasuries as an illustration of the reasoning.

Regarding the budget, the post you link to is almost a year old and was a draft. Since then there have been several actual budget requests that have come and gone. The currently approved budget requested $24m through to the end of the year. Much of this is for audits for further Lido v2 developments including the Staking Router and Dual Governance.

Other than that, look forward to a fruitful discussion for this interesting proposal. We will give it some thought too and provide our views.

22 Likes

A large % of the LDO in Lido’s treasury could be put to work securing the protocol and earn a return.

But the amount of stETH earned by Lido atm is 10,700 stETH($20 mil atm)

If people are curious what Lido makes every rebase I would check this out. It the latest rebase showing what node operators and the treasury earned 24 hours ago

2 Likes

This is a great proposal. Highly support it

I think this proposal is alright. stkAAVE style mechanism is generally sound, and considering buyback to acquire the LDO to be distributed to stakers along with staking lockup period, it provides value accrual not just for staked LDO but all LDO holders as a whole.

Do have some general comments though:

  • The value of staked tokens as insurance is probably not very high, considering LDO can be expected to fall sharply in the event of a significant slashing event requiring recapitalization beyond insurance fund
  • 6k ETH insurance fund doesn’t seem sufficient to have confidence in being able to self insure against small to moderate losses (even excluding large systemic slashing events which are effectively un-hedgeable)
  • Lido continues to have significant token denominated operating expenses from liquidity mining for stETH, and I don’t think it makes sense to be making distributions to token holders in a case where the protocol is not actually profitable yet
  • Lido has limited development resources which could be better spent on other priorities that grow the protocol rather focusing on distributing existing revenue

So while I think the broad strokes of the proposal are fine, I’d be more comfortable approaching this as a medium term (12-24+ month) goal. Additionally, I think targeting a much higher capitalization/insurance buffer (0.2-1%, ~12,000-60,000 ETH based on current circulation) before distributing earnings makes sense. And this insurance backstop could theoretically be put to use providing liquidity in Curve or Balancer stable LP, which can reduce Lido’s operating costs linked with liquidity mining.

tl;dr: grow the pie before focusing on how to slice it up.

Gonna leave this video here to end on a light hearted note :laughing:

“It’s not about how much you earn, it’s about what you’re worth.”

22 Likes

Looking at the AAVE price action after introducing the staking/safety mode, it’s very depressing.

Also MKR proved that buyback is a very very bad idea. It simply doesn’t work in crypto. I bet LDO token holders would rather be distributed native revenues i.e. stETH than LDO.

Please change the proposal accordingly!

1 Like

Awesome proposal! Excited to see it put forward to a vote. However, I’m not sure about the idea of requiring node operators (NOs) to stake $LDO tokens. I think it makes sense being optional with the added benefits to them (e.g. $LDO yield and maybe lesser fees). Since they already have to stake $ETH, adding the $LDO staking requirement may potentially reduce the number of people interested or able to provide NO services.

Why over complicate ?

Stake LDO, share stETH to LDO stakers, because it is a eth backing platform, as rewards from protocol over ETH staking rewards fee, not the other way around.
Use a fair % to distribuite.
Keep part in the treasury for development and maintenance.

So far that is what I think…
Need to think more.

1 Like

tldr

  • probably gud eventually in some form but we prefer to wait for dual governance and an ecosystem to form around the staking router
  • highest impact is to focus on long-term product things, let the pie grow a bit more first
  • dont design for ponzinomics, design for a better protocol
  • eg flap/flop auctions, staggered redemption curves, mint/burn on price thresholds, etc ideas welcome

wall of text

Generally we agree with @monet-supply for similar reasons.

  • some sort of balancing incentive for LDO holders is probably needed in the long-run
  • the current proposal is designed for LDO value extraction maximization rather than balancing the Lido on Ethereum protocol as a whole–we are better off in any case focusing on long-term impact
  • It is also likely far too early as it would be nice if we could let the Staking Router ecosystem mature and launch Dual Governance before refocusing

The more governance levers we introduce to the protocol, the further we stray from developing a thin, neutral protocol. The Staking Router and Dual Governance are important steps in the right direction, but a governance-controlled payout is a step in the opposite direction.

Btw we’d go further than that even and suggest that stETH with the Staking Router and Dual Governance could become the pivotal decentralized umbrella liquid staking token to face off against centralized entities, as any decentralized pools of validators could theoretically join stETH through a SR Module. These are developments that are well worth focusing development interest and focus on over the next few months.

In our view it would be better to have an automatic system with no governance input rather than one where governance can control the payout ratio.

Example such systems

Maker flap/flop auctions: In reality, the flop system would likely not work or not work well – in a scenario where stETH is undercapitalized from a giant slashing event that wipes out the surplus, it would likely be very difficult to raise enough ETH through LDO sales at that point and any efforts to sell would make it even harder.

Staggered redemption curves: The clever Gyroscope stablecoin team have other ideas to help bolster the defense of their protocol, such as decreasing redemption curves to slow down a ‘run’ on the assets in the event of a market shock. For stETH it could look something like removing the commitment to 1:1 withdrawals if the protocol surplus goes below a critical threshold level or 0. This could buy time and avoid issuing LDO in an extreme market scenario. This has further benefits by increasing the cost of governance attacks.

Mint/burn on valuation thresholds to bolster and burn the surplus: The other balancing factor that is missing from this proposal is a trigger to issue new LDO when the valuation is appropriately high. The thresholds could be set without oracles, as a function of the Surplus to Token ratio and AMM LDO/ETH prices.

In any case, what we should be solving for is a more perfect protocol, rather than try to introduce narrow tokenomics that we believe will game the price. The constraints that systems like this follow probably look something like:

  • Possibility to make threshold levels immutable
    • for eg % of totalSupply of stETH or a ratio of surplus to AMM price etc
  • Automatic and uncomplicated mechanism with explicit rules understood upfront
  • No privilege for token holders ‘in the know’, even-handed equal treatment of all LDO token holders alike

Agree with @monet-supply on needing more capitalization. 6k is arbitrary, quite low (out of date?).

LDO as bonding token

Regarding LDO as a bonding token for Node Operators, we get the gut feeling that this is ‘early bagholder rewarding’ tokenomics that doesn’t actually secure the protocol and ‘endogenous collateral’ shouldn’t be the bonding token in any case.

Other thoughts

A stETH-powered L2 would be great!

21 Likes

Nice proposal!

I just want to highlight that implementing a cooldown period limits the potential integrations that can be built on top of any such system. For example, at Inverse Finance, we’d love to implement staked LDO as a collateral option in the future when it’s ready; some really cool concepts can be built, such as self-repaying loans using the rewards. However, any kind of lock or cooldown period makes this a lot more challenging to do in a safe way due to the need for instant liquidity for liquidations.

2 Likes

I agree that it’s unquestionably important to tie the success of the protocol to the LDO token. Revenue sharing is the way to do that, as so many other protocols have shown. If this is not done in the next few months, we could see LDO start to go down from farmers dumping.

It is important, however, to understand the dynamics of people locking their LDO. People eventually need to make a profit. If you’re buying back LDO and distributing that, people just accumulate LDO → no realized profit. An alternative would be to buy ETH and distribute that. People would then be realizing their profit on buying LDO without having to ever sell.

3 Likes

Regarding LDO as a bonding token for Node Operators, we get the gut feeling that this is ‘early bagholder rewarding’ tokenomics that doesn’t actually secure the protocol and ‘endogenous collateral’ shouldn’t be the bonding token in any case.

Strongly disagree with this. I think a second layer of LDO/ETH (a combination of those) bonding (which gets slashed locally based on individual NO performance) among NOs is def required as right now NOs have no skin in the game. If they get slashed the Insurance Fund has to cover for their losses at the expense of LDO holders.

Not including any bonding was a great way for Lido to scale quickly but as the stakes get higher we can’t just rely on human layer of coordination among NOs, and the DAO to handle that. The incentives/alignment needs to be more economic on that front.

At 5% take rate the profit margins are huge for NOs and they don’t contribute to the insurance fund at all. If a NO is responsible for managing $100m I believe that they should be able to post atleast 1-2% of that amount as LDO/ETH to show alignment and help cover some part of insurance fund. Otherwise we are left with Insurance fund being solely DAO treasury growing at 5% of staking yield per year.

And If you check the chain most of the NOs (atleast with public addresses) have already sold their LDO. Let’s see if any NOs want to come out and publicly disclose their holdings and talk about their alignment

3 Likes

Additionally, I think targeting a much higher capitalization/insurance buffer (0.2-1%, ~12,000-60,000 ETH based on current circulation) before distributing earnings makes sense. And this insurance backstop could theoretically be put to use providing liquidity in Curve or Balancer stable LP, which can reduce Lido’s operating costs linked with liquidity mining.

I agree with this. Think Lido needs more insurance as TVL grows at a fast rate but I don’t understand why Lido DAO (LDO holders) are the only ones who needs to pay for it from their 5% take. Imo to grow the insurance pool it’s important to introduce another parameter in terms of NOs stake rather than solely squeezing out DAO.

Its insane how social slashing still isn’t implemented here and why no one has brought it up. Just few weeks ago the RockLogic GmbH slashing incident occurred and I think its a great time for the DAO to realize that they are sharing 5% of the revenue with NOs and these NOs should contribute to more aspects of the protocol like insurance fund and posting some collateral to have skin in the game and show confidence in their own capabilities.

Running nodes isn’t a big favor when you’re making insane margin of profits at the expense of holders. There’s huge misalignment there imo

4 Likes

I agree with @monet-supply that the LDO stakers return should probably be ETH denominated and be used for LP like maybe ETH-wstETH LP tokens like how curve returns 3pool tokens to CRV stakers

4 Likes

Agree with bonding just not using LDO as the collateral, but open to changing views obviously.

Some measure of bonding or insurance is definitely desirable, especially as permissionless-style modules are added to the Staking Router, but anything other than stETH and ETH is basically second tier (or worse) for any kind of serious slashing or outtage event, and therefore isn’t a serious option for anything at scale. This is because if there is a serious event, any other kind of collateral will just get market-sold by other holders and then its utility for the purported use case of making stakers whole greatly diminishes.

My opinion is that collateral should always be primarily in ETH or stETH, and that secondary collateral can be in LDO or other tokens (e.g. specific modules can propose to have different tokens as utility or bonding tokens in addition to ETH).

7 Likes

Ask them to bond ETH and LDO both, if they do an “oopsie” then the regular procedure of shutting down the validators etc happen but on top of that portion of their ETH gets slashed as well to cover the insurance, on top of that some portion of their LDO gets taken and gets re-distributed to stakers.

I think LDO should be used for reputation based scoring and every NO should be required to stake atleast some and staking more ETH and LDO gives them more reputation and more incoming stake. This should be extremely useful when staking router is live and protocol wants to allocate incoming ETH into different routers and various NOs within a router

The jobs of NOs is to run validators well and protect the network, not to pump LDO bags. If they don’t perform well the DAO’s job is to not give them stake.

Buying reputation is how protocols get rekt by malicious actors. Bonds can be a way to increase a score, but the quality of the bond matters, and there needs to be a cap for it to not be a large attack vector.

3 Likes

I don’t disagree. I’m not saying anyone can buy ETH and LDO and get lot of stake, the current validator set is curated anyway so its more like if you do that you get slightly more than the one who doesn’t cause you are more aligned and are staking some reputation behind it.

Safety of the protocol and building up a healthy insurance should be what the design should aim for but I think alignment can be improved more if you bring some sort of reputational stake which gives slightly more advantage to an already chosen NO

Also not a fan of the narrative that NOs are running the validators and “protecting the network” so LDO holders should bow down to them or something. If NOs fuck up LDO holders, stETH holders will have to bear that massive cost while they won’t face any consequences.
I think a fair and clear economic alignment between different actors combined with good protocol design makes the system more robust

1 Like