Hello everyone,
As you all know, Marinade’s goal is to increase the number of validators on the Solana network and subsequently raise its Nakamoto coefficient. I would like to put forth a proposal to the mDAO that, if agreed upon, is designed to address Marinade’s stake pool strategy, create additional benefits to the validator community and strengthen Solana’s censorship resistance. Following your feedback, this proposal could be the first item voted on in the mDAO using our new on-chain governance set to launch very soon.
Problem statement
One of the main struggles for new or small validators is becoming profitable. Getting to a “break-even” situation is really hard and prevents a lot of people from trying to launch their validator on mainnet.
To counter this, validators currently only have a few options:
- Apply for a Solana Foundation delegation that has a backlog of more than a year (and requires running two servers, one on testnet and one on mainnet, which up the costs significantly)
- Launch their validator with at least 5,000 SOL ($410,000 at current price) that they self-stake to their validator
- Manage to get visibility and get a lot of stake from users directly (50,000 SOL at 10% commission to break even)
- Try to get the stake from one or multiple stake pools, which may not be enough by itself to become profitable
They often end up mixing all those options in order to put all the chances on their side.
Some of these scenarios require a huge amount of funds to be injected into the project in order to ensure that their validator will be profitable. For most people, this amount of funding is out of reach and leads them to abandon the idea.
Getting stake from stake pools is also something that is difficult to plan, while also making your validator dependent on an outside parameter, which might be problematic from a validator’s standpoint.
We can add to this that bringing stake from users to their validator is a challenging task that often drives validators into marketing, content creation, etc. in order to get some visibility among users. All of this costs time and money, without even mentioning opportunity cost.
Current Marinade situation
Marinade TVL currently allows a number of validators to be profitable with only the Marinade stake, but for the majority of them, Marinade stake is situated between 15k and 40k SOL.
While this would improve with the TVL getting bigger and bigger, the current situation still leads validators to rely on themselves rather than stake pools.
If we use Cogent’s tool to calculate the profitability of most of these validators having between 15k and 40k SOL staked by Marinade, we can observe that the Marinade stake is not enough in itself to make them profitable.
As we can see, having only 35k SOL delegated on your validator is not enough and makes your node cost you approximately $9,000 USD per year. This forces validators to not only rely on stake pools but also on finding the stake elsewhere.
As a result, they can’t work in full symbiosis with Marinade as they are competing for the same stake.
Finally, those data are calculated with a 10% commission, the maximum commission allowed by the Solana Foundation and by Marinade, which is almost mandatory for small validators to hope to become profitable, whereas installed validators can lower their commission as they get more stake and more “slot-leader” positions on average. This allows them to attract even more stake by being the most competitive.
All of those points create difficulties for new and small validators to blossom and this obviously leads to a more centralized network in the end, as bigger validators survive and smaller disappear or are discouraged to even try. It also pushes validators to compete for the stake with stake pools.
Reflecting on this situation, I would like to offer a solution to the community on behalf of Marinade, that has been actively thinking and working on potential solutions.
Proposed solution: MNDE voting on validators stake to guarantee sticky stake and predictability
Now that on-chain governance is becoming a reality, using gauges for this purpose becomes a possibility. Let’s take a second to redefine a gauge:
A gauge is a governance mechanism that allows holders to allocate the % of the total stake that their holdings represent to a specific target. They can be used to control any type of allocation in a decentralized fashion.
Applied to MNDE voting on validators stake, this is an example of what the gauge could look like:
1M MNDE is locked in governance and participate in the vote on validators stake, controlling 1M SOL staked.
Owning 10,000 MNDE (1% of the total MNDE locked in governance) would give the owner control on how to allocate 1% of the stake.
So, dedicating 10,000 MNDE to Validator A through the gauge would guarantee 10,000 SOL staked to this validator.
If Marinade’s TVL doubles and the number of locked MNDE in gauges stays the same, 10,000 MNDE would then allow to allocate 20,000 SOL to a specific validator, as 10,000 MNDE would still control 1% of the total stake to delegate.
On the other hand, if the number of MNDE locked and used in this gauge doubles, reaching 2M MNDE, 10,000 MNDE would only control 0.5% of the total stake.
Pros
As I explained above, securing an initial stake in order to break even is one of the main struggles of new validators.
Current validators that rely on stake pools or whales also are in a situation where planning long-term may be hard, as a whale or a stake pool can lower its stake or even fully unstake at any time. Validators gauges would:
-
Offer a possibility to validators to guarantee a part of their stake is an efficient way to allow them to plan and secure their position as performing validators in the long run. This should result in more validators and a more decentralized network.
- Having this guaranteed stake may also lead them to attract more stake from users or even from Marinade or other stake pools, as they perform and validate transactions efficiently.
-
Allow some validators to lower their commission and become even more attractive because a part of their stake (and of their profitability) is guaranteed by their MNDE holdings.
-
Invites validators to buy and lock MNDE, making them active members of the DAO. They would gain voting power to take decisions on the delegation strategy, on the future of Marinade, and they would join and constitute the security layer for Solana that Marinade is.
-
Finally, this would lead to more MNDE locked in governance, as MNDE has a utility and an additional financial incentive for validators. This would also bring buying pressure on MNDE which is currently lacking.
Cons
-
One of the main things that might worry people with gauges is “What happens if a big validator with deep pockets buys all the MNDE and delegates everything to himself?”
-
While being a possibility, MNDE has been distributed in the fairest way possible and it would be very costly to accumulate MNDE given the available liquidities.
I believe that by being transparent on the rules of the game, everyone will have the same incentives to accumulate (or not) MNDE and will have time to do so, as the emissions are still happening for a foreseeable future. -
Marinade would also make this information clear and communicated to all validators, in order to level the playing field. I believe that the financial incentives and the current distribution will help create a free market that regulates itself around offer and demand, preventing one or few entities from accumulating too many MNDE.
-
-
MNDE is still being emitted (but in a predictable and reasonable way). This would mean that validators might have to sometimes “reinvest” in MNDE in order to keep the same power over the stake.
- To answer this, I would hope that as validators get profitable more quickly and easily, a part of their benefits may be oriented towards investing in their position in MNDE governance. Investing in Marinade governance may also reap more benefits than just allocating a stake in the long run, as a validator would directly be involved and implicated in the decisions taken regarding Marinade and its treasury.
-
Finally, some people might be worried that this goes against the ethos of Marinade itself.
- Only validators eligible to Marinade stake in the first place (commission max 10%, not part of the super-minority, not blacklisted for cheating the system, etc.) could receive stake via those gauges in order to keep on raising the Nakamoto coefficient. This prevents this system from being used to lower the Nakamoto coefficient or to significantly lower mSOL APY.
Initial settings suggestion
The initial idea would be to allocate 10% of Marinade total stake to the Validators gauges. Those validators would have to be eligible to Marinade stake in the first place.
Example
Say Marinade has 10M SOL under management and allocates 10% of it to MNDE governance voting, 1M SOL.
Say half of all circulating MNDE tokens gets locked in governance to vote = 30M MNDE.
Say half of locked MNDE tokens will participate in validators voting.
= 15M MNDE will vote over 1M SOL delegation.
1M MNDE would control 66,666 SOL to delegate, which would be more than sufficient to become profitable for a validator at 10% commission, returning around $900/month. (See Cogent’s tool)
Final thoughts
Another thing that can be reflected on is the value that validators would get out of their dollars when choosing between buying SOL and staking it to themselves or buying MNDE to control some Marinade stake.
$1 million would currently buy approximately 15M MNDE, enough to control 1M SOL, which would represent $24,000 - $48,000 each month for validators (with 5 to 10% commission). This represents $288k - $576k per year.
On the other hand, $1 million would buy roughly 12,000 SOL, which would bring $5,687 every month if self-staked ($68k/year).
This situation would lead to a validator earning 29-58% per year on its investment through the staking fees secured by MNDE.
On the other hand, self-staking those 12,000 SOL would net a return of 6.8% on the investment.
This makes MNDE extremely capital-efficient for validators.
Finally, the last advantage of this system is that it also gains in power as Marinade TVL grows.
The mDAO will be able to utilize the upcoming on-chain governance using MNDE tokens to vote on this proposal. But before we do, mDAO members, validators, the floor is now yours to discuss this idea!
