35% - DAO Distribution . Used as token holders see fit via proposals, initially used for Liquidity Miningprograms.
35% - DAO Treasury . This is a treasury reserve to be used for operations, grant programs, and strategic partnerships. This treasury is controlled by Marinade’s multisig and will be unlocked and governed by MNDE holders.
Besides this reserve, the treasury also receives mSOL protocol fees:
2% of staking reward fees generated by Marinade protocol
25% of liquid unstake fees generated by the Unstake liquidity pool
30% - Team . This is an allocation for current and future contributors. The distribution has not started yet and those funds are currently controlled by Marinade’s multisig.
Thanks for raising the point, @meyerbro, it’s one I hadn’t considered (I hadn’t realized team members were validators).
As you can see in my posts, I’m aligned with you and @laine about Marinade keeping most (or potentially all) the stake distributed algorithmically. I do think there are safeguards in place which make @Cerba’s proposal rather low risk:
I get the concern of a potential conflict of interest. Given votes are transparent and on-chain, however, wouldn’t this kind of corruption be obvious? (eg. a set of wallets always voting for a specific validator)
If such petty corruption would be obvious, then it seems to follow that engaging in it would destroy Marinade’s value proposition, and thus devalue the MNDE the team holds - rational self-interest would compel team members to pressure each other to act honestly [1];
Finally, even if we do think team members might favor others who are validators, wouldn’t keeping the percentage low (say, the proposed 10%) make this kind of behavior even less appealing vs. the potential blowback?
Would love to read your thoughts on those.
[1] eg. what happened in OpenSea, where Nate Chastain abused his position to engage in petty corruption, but his actions were against the team’s interests so he was expelled.
Thanks FWIW I’m in favour of maintaining algorithmic distribution for a majority of Marinade stake, but I don’t oppose some percentage being distributed per this proposal’s suggestion via MNDE voting.
Something mentioned on yesterday’s Solana validator round table was also to possibly give validators higher voting power, I’m not entirely sure yet whether I think that’s a good idea or not but wanted to mention it here so it’s on the record, and in doing so I believe the most effective way would be to attach such increasing weighting to votes placed by accounts that hold MNDE AND are withdraw authorities of vote accounts. That being said anyone can create a vote account so the weighting could either be stake weighted or there could be a minimum stake required to obtain the increased voting power (or have it linked to the algorithmic score!! that would be a nice crossover between the two methods)
The team allocation most contributors get (including me) is not nearly that high as some of the MNDE wallets I have seen from people/companies that are gathering mnde tokens for the last year.
I also think every contributor should have the freedom to run a validator themselves.
But I do get your point and concerns. I would vote yes to “MNDE given to team members shouldn’t have voting power to select validators”.
To speak for myself. And I will always be completely open and transparent about this; I would never use any team token allocation to vote for my validator. I agree it would be unethical. like @PlayerOfBits said, such things would be pretty obvious.
But, I did buy a bunch of MNDE on the market like everyone else, that I would use to lock in and vote for the validator of my choice.
I’d be OK with this as well. It would require team members to maintain some wallet hygiene, and I agree with @stakeconomy that team members should be able to vote with MNDE they have acquired themselves, but I’m not against restricting voting with the MNDE assigned as payment.
Then again, that’s easy for me to say, since I don’t particularly value assigning stake weight to any specific validator (although that may change once I see the feature in practice). Curious what other team members think.
My view is that it’s not problematic for someone on the core team to run a validator or even for someone on the core team to vote for their own validator so long as there are clear guard rails in place and there is transparency around the process.
If the goal here is to distribute Marinade’s stake fairly, maximize impact for smaller validators, and increase Solana’s Nakamoto Coefficient, I would argue that the core contributors probably have the best sense for how to accomplish that. No need to handcuff the most informed voting power out there.
…
I don’t think the right solution is as simple as this.
Perhaps the best route would be to let the core team vote how they see fit but with caps on the total amount of their personal voting power they can assign to one gauge/validator.
What would that number look like? Maybe it’s 25% of personal voting power, maybe it’s even higher? Idk what the right number is yet, but I would love to spend some time discussing it here.
But again, transparency is the key - and recent history has taught us that having the right tools in place to monitor these things is highly important.
BTW, I would be more than happy to help build these tools too!
Thank you for the great work and outlining the proposal. I think it is a great idea to integrate some support mechanism for new validators into the Marinade Pool to raise the Nakamoto coefficient. While I like the current proposal, I think there are some limitations / things to be cautious with.
1. Helping smaller validators to get profitable
Validators can secure a “base-stake” to operate and increase their planning security. This is great, but can become problematic without a limit for big players
Big players can arbitrage the leverage MNDE offers away. The scenario quoted below will incentivize big players to set up a validator and gain yields of 28.8%-57.6% on their invested capital. For big funds it can be worth to act like this with millions of dollars until the MNDE price reaches a point that makes it unattractive for big players (and much earlier for new validators) to use MNDE for delegating stake. So we should make sure to only make gauges capital efficient for the right people.
The 1.5% cap counteracts this behaviour somehow, which is nice. See more under (3).
Proposal: Only validators with <XXXk SOL in total stake should be eligible to benefit from gauges. This would go further than the 1.5% cap of the pool i.e. we could offer usage only for validators with up to ~200k SOL staked. This is a range where validators can operate with a solid profit. For sure validators with >500k SOL should not profit from the MNDE token if this program really aims at helping new validators
2. How much of the Marinade Pool should be controlled by MNDE?
I think 10% is a good starting point
Going above 15-20% is potentially dangerous imho because Marinade does not keep its promise of algorithmically distributed staking. This could lead to a smaller pool size and eventually also to a decline of the MNDE value. Just from my feeling, 30% would definitively be too much. But I also like the comment mentioned to separate the discussion about the actual size of the MNDE controlled part from the discussion about the initial mechanics
3. Should the stake controlled by MNNDE considered as the “same stake” as the stake delegated by Marinade’s delegation strategy?
When answering this, think about the initial rationale to cap the Marinade stake to 1.5%
As far as I know, validators are capped to 1.5% of Marinades pool to support decentralization and ensure a diversified base of validators
Adding a second cap of 1.5% and allowing validators to get more stake through the gauges is against the rationale of capping validators stake in the first place. Consequently, validators should not be able to get more than 1.5% of the pool through MNDE.
Adding a second cap of 1.5% does not help new validators. It makes it only easier for validators who are already successful.
First off I think this proposal is an excellent idea and I’m in full support. Here are a few reason’s why
Reason 1: A lot of foundation validators are stuck at 10% commission and are slowly losing stake from the foundation and have not been able to attract organic stake due to the very high commission. If they halve their commission to 5%, a more reasonable number for attracting organic stake, they could possibly put themselves on a path of sustainability, but their marinade stake would not go up significantly and the short term losses may not be sustainable.
You can see this from the mariande stake data where the average 10% commission validator is at an eyeballed estimate of ~15-19k stake from marinade and the average 5% validator is at 19k to 25k stake from marinade. (If there is contentoin on those numbers I can take futher time to find exact numbers) at 7% apy In these two instances the 10% validator receives 126 sol a year from Marinade stake and the 5% validator receives 68 sol which could be the difference between profitability or not.
If there was a sustainable path forward via providing value to the community, attracting additional stake organically, and through Marinade, perhaps they could gather enough stake to strike out on their own and be competitive with out the foundation stake.i.e. have more competitive returns.
This proposal would allow some validators to possibly make the move from 10% down to 5% or lower and have a long term sustainable path without going out of business in the mean time.
Reason 2 By putting a direct value allocation on MNDE the value of MNDE its self will go up. This will in turn allow for higher liquidity mining values and further increase the defi value proposition of mSOL attracting more stake making for a higher value of MNDE.
Reason 3 As more validators are able to have a sustainable path, with competitive rates, the overall returns of mSOL will increase. This would make marinade a more attractive option for some of the stake from whales who currently stake with validators with better returns.
Conclusion I see this as an important step to increasing the value of both MNDE and the value of holding mSOL which will put us well into the fly wheel of value. (even if you are familiar with the flywheel of value concept, I highly recommend this wonderful article, https://www.strategy-business.com/article/The-Flywheel-Philosophy)
Unless the Chefs have received an exorbitant amount of MNDE at very low prices I do not have an issue with them using their earnings as a Chef to do what they please with their MNDE. To me it would be the equivalent of them receiving USDC and trying to say what they can or cant do with their earnings. In fact If anything I would trust the chefs to act in the good of Marinade as holders of MNDE more then if they did not hold it.
Now I would have a different view point if the chefs controlled a very large percentage of MNDE and got it at un unfair valuation.
What if say a validator starts providing incentives for individuals to vote one way or another. e.g. allocate your votes to me and recieve X amount in profits, or X many Y tokens… It could be natural for economic competition to arise in such a system. Would we look to prevent this or allow it to happen?
I’ve read the proposal and while I believe it comes with good intentions, it is not a very good idea. Marinade started great by using an algorithmic delegation that is much more objective and based on facts rather than on human factors. If you check the majority of the PoS networks, they are super centralized from a stake perspective - people tend to follow the herd and stake with the same validators ranked in top by stake, stake can be bought, many deals are made behind the doors, it is terrible.
I can’t really follow what is the true benefit of adding a human voting that can definitely be easily bought, subjective and through deals behind the doors. It looks more like adding an use case to the MNDE token but probably destroying a positive influence for Solana as a whole. What will happen as a consequence is that small validators will be chosen by humans which will selectively define how much stake one should receive but not based on true on-chain metrics.
The proposal has close to no impact on the huge delegations made by institutional investors or to the whitelabel validators operated in the Solana ecosystem. The real problem lies actually exactly here - since the idea would be to distribute the stake better and not to define a whole new small - medium validator set based on subjective factors.
Apologies for the criticism but I trully don’t believe this will have a positive benefit for the current small validators, except for those that might have some contacts with MNDE token holders or owning themselves MNDE tokens, which doesn’t sound like a fair system.
Algorithmic staking is something which all networks should implement in their original staking mechanism in order to better distribute the stake.
Looking into this idea a bit more… from what I can see on Solana Beach there’s actually a decent amount of validators with > 500k SOL delegated that sit outside of the so called “security group”. I wonder if the chefs have any idea how often this subset of validators actually receive stake from the existing algorithmic delegation strategy (if at all).
Thanks for all the feedback, folks. @Cerba and myself have discussed this and we hear your concerns. We need to think about how this could be designed to address the issues that you have raised so far - I would not expect this to be reaching a governance vote before we do that.
I’m inclined to think that this might be a transparency issue, so it bears thinking about what sort of information could put the community at ease. For instance, it’s perfectly possible that the concern isn’t as trivial as “team members” (which could be someone holding merely a handful of MNDE), but in general large, concentrated stake swaying the protocol’s actions. In that way, it might be connected to the topic of vote multipliers as well.
Let us take some time to get data, create a few models, and come back with suggestions.
Maybe this has already been discussed elsewhere, but I’m also wondering how Marinade will determine which validators are assigned a “gauge” to vote on. As the number of newer validators (who probably need delegation the most) increases, what will the process look like? There are already over 1600 of them per Solana Beach. Will they have to apply? Will this also be decided by a DAO vote? Is there a maximum number of gauges that can be assigned using Tribeca?
Great, I managed to fat-finger delete my own reply.
The topic hadn’t been brought up here, but thanks for raising it - it just came up on a call that @cerba and I were having. Marinade currently only delegates to the top 300 performing validators. However, the DAO could decide that validators between (say) 301-1000 are able to apply to the vote gauges, which could get them more stake and could also help them improve - pulling themselves up by their bootstraps.
Will this also be decided by a DAO vote?
I think the general mechanics need to be decided by a DAO vote - who gets to apply, what the parameters and requirements are, and so on. Running every validator application through a Tribeca vote would be overkill, and once we have voted on them, then they apply to any future potential validators (until we vote to change them).
Folks, I wanted to follow up on the concerns regarding the team using their allocation to vote on validator selection.
Is a concern that the on-chain gauges could allow someone with a large amount of MNDE to bend or break the Marinade delegation rules?
I personally think that the delegation rules should stand above all, even gauges. This means that the maximum 1.5% Maximum Stake Cap would hold.
If it does, and 10% of Marinade’s total stake is controlled by gauges, then that means that if the entirety of the 30% team’s stake were to collude and vote together, it would at most be able to max out the stake limit for two validators.
Given a scenario where the global limits hold - is the team voting with their stake still a concern?
I am in favor of enabling validator voting with a 10% allocation to start.
MNDE Value = Governance Control:
As Marinade is decentralizing governance, we must recognize the DAO will act via token voting for any major decisions. That governance power is unsustainably cheap given the current MNDE price. With 24mm of MNDE locked in governance, a malicious actor could pass any proposal for only $2mm (51% of locked).
Marinade’s primary goal should always be ensuring decentralization and security of Solana but that doesn’t mean the protocol can ignore economics. Outstanding mSOL and MNDE price have been steadily declining while competitors are growing. Liquid staking is a winner take all market - it is imperative mSOL wins over our mercenary competitors.
The MNDE token needs utility and a strong price to help ensure Marinade thrives in the long term. This proposal demonstrates MNDE’s utility and will help provide a floor price as validators look to acquire stake. This is a positive for all of Solana.
Eligible Validator Guardrails:
I agree Marinade should require strong performance from any validators asking for gauge votes. This ensures we do not compromise our primary goal of Solana decentralization & performance. Allocation caps can further prevent any prospect of abuse.
Team Self-Delegation:
I struggle to understand why this is a problem. If delegates have met strict performance requirements, why would it matter whether the validator is operated by a team member or third party? Rational self-interest would prevent any MNDE holders from voting in a way contrary to the long term value of the DAO.
If validators are upset they are losing votes, they should buy more MNDE. It is quite cheap right now. If delegated validators are not benefitting the network, we should tighten the requirements further regardless of the operator.
Some of the proposed solutions to this team delegation question introduce a massive amount of complexity that could reward questionable workarounds. With only a 10% voted allocation, we should start simple and see how things actually play out. We can always introduce restrictions later if actual problems emerge but those will be challenging to undo once in place.
You make an excellent point - in a scenario where a global 1.5% cap holds, I would not be concerned at all.
However, I don’t think that applying the 1.5% cap universally across both algo and governance-controlled delegation would be a good solution. @brian_smith_0 is right - the MNDE token needs more utility, not less. We ultimately want validators to recognize the true value of the token and compete to acquire as much as possible. If the governance-directed portion is capped at all, that cap should be much higher.
And as long as there is transparency and the team continues to act in the best interest of the protocol and its goals, we will be just fine.
