mDAO proposal - Unstake issue compensation

Governance
#1

Greetings, governoooors!

What do you propose?

Marinade will redirect the protocol fees it would normally get directly into the SOL staked pool, for a duration of 3 months OR until 2646 SOL have been sent back to the pool (which could be faster if the TVL grows in those months).

The start time depends on the community approving the proposal, but we expect it to be on the next epoch after the proposal is executed.

What is the rationale behind the proposal?

Background

As you may know, a cascade of events during epoch 284 led to the unstaking of approximately 6.33M SOL delegated to Marinade which resulted in mSOL price not going up as it should have. A full post-mortem is available in the Changelog.

As communicated right after this incident, Marinade wants to bear the responsibility of this incident and help mSOL holders. Nonetheless, as the missed mSOL appreciation would have compounded and as APY and TVL are changing over time, it is impossible to sort out a change in Marinade settings that would make sure that the compensation is perfectly accurate and precisely reaches everyone concerned. We have settled on what we consider a fair middle ground that should help balance out mSOL’s missed appreciation.

How did we arrive at this approach?

First of all, it’s important to state that mSOL price cannot be manipulated and Chefs had to look for a way to compensate that would actually be possible to implement with how Marinade smart contracts work.

As it’s not possible to directly send SOL to the Marinade smart contract without minting mSOL, so we can only envision a progressive approach directed at the fees.

Therefore, the option that Marinade contributors agreed upon would be to lower the management fees taken each epoch by Marinade (currently at 2% on rewards) to 0%, effectively reducing Marinade’s revenues and redirecting them to mSOL holders. Using this system would allow the mSOL price to grow at a faster rate than what it would if it was only getting fees, slowly catching up on the missed epoch, only by changing this parameter.

If we consider the amount of SOL that mSOL did not appreciate on that epoch, we can consider the TVL at time, multiply it by its yearly APY and then estimate the APY for one epoch. This yields:

6339496 * 0.061 / (365.25 * 24 / 60) = 2646.88060233 SOL

In order to compensate for this event, Marinade Chefs suggest that the fees of the Marinade protocol are redirected to the SOL staked pool for a duration of 3 months OR until 2646 SOL have been sent back to the pool (which could take less than 3 months if the TVL was to grow during that time).

What is the expected positive impact of this change?

mSOL will have extra appreciation during the time specified by the amount that would otherwise go into the project’s fees, up until 2646 SOL. This will financially benefit current and future mSOL holders.

Any other considerations?

Main considerations for this decision:

  • This will benefit mSOL holders but negatively impact Marinade’s treasury, which is something MNDE holders need to vote on.
  • Chefs cannot directly manipulate the price of mSOL (which would be a systemic security risk to the project), so we are left with an approximation to help mSOL appreciate using Marinade’s operating capital.
  • Marinade’s treasury lost out on fees during this period along with the mSOL appreciation. Liquid staking provides more flexibility and direct access to our stake, as well as extra APY from use on DeFi, but one of the trade-offs is that we may need to weather unstaking events such as this one. While we can’t guarantee that Marinade will do a similar cash outlay in the future, the team is willing to take a hit and help mSOL get closer to where it should be.

This incident did have some positive side effects, given it also:

  • Allowed a full rebalance of the stake, meaning the system re-distributed the stake in a more optimal manner.
  • Led to the team implementing more checks during this process, making it better. Specifically, we added a sanity check on the total score returned by the processing, removed the dependency to StakeView, and corrected the Emergency brake that had malfunctioned.
12 Likes
DAO proposal - Delegation strategy update & fee structure changes
#2

While I wouldn’t consider my position a hard no, I do feel that the value of mSOL was not harmed greatly and materially I don’t think any one person noticed a difference. However a growing organization, MDAO, would greatly notice 2647 sol. I think having a strong treasury and using the DAO to build and fund community tools would be far more valuable then ever so slightly increasing the value of mSOL over time. This is coming from someone who holds far more mSOL then MNDE.

My counter proposal would be to set aside those funds to pay for builders to make tools that help all of Solana. e.g. The recently built Decentralizor.

That’s my 2 cents :slight_smile:

11 Likes
#3

It is hard to make a call on this. I believe both proposals (Cerba’s and Congent’s) are good. Maybe some more information could be useful. For example:

How much of impact over MDAO treasury would have feeding 2647 Sol into the pool? This can somehow compromise any of the MDAO capacity to make new developments or to keep up with its schedule/roadmap? Are there any impacts over workers wages, etc?

Although I agree with Congent that the impact of this incident is minimal to holders, if none of the DAO functions is compromised with the official proposal, brought by Cerba, there is a reputational issue, given the fact that Marinade already promised to cover the costs of the incident in its announcements. Therefore, I lean towards agreeing with the official proposal (and I hold more MNDE than mSol right now :))

Lastly, in another proposal, perhaps we should also vote for constantly and gradually building a security fund with monthly revenues for the case of any unfortunate incidents that may happen in the future.

4 Likes
#4

Do you mean, as a counter-proposal, that this amount be earmarked for something like the grant program instead?

#5

Yes if it doesn’t hurt the DAO too much. I’m assuming its doable from Alex’s proposal. I’d love to see that returned to the community through grants to builders for anything that helps the Solana Community. Better yet if it helps decentralize stake but I’m not sure it needs to be limited to just that. Just getting people to build for the greater Solana network helps all. mSOL holders included.

3 Likes
#6

Full disclosure: I worked with Cerba on this and, as a MNDE holder and team member, I have a vested interest in Marinade having a solid treasury.

This is a big part of it for me - Marinade has stated publicly that it would help mSOL get back to where it would have been, so we need to make a best-effort attempt at fulfilling that commitment.

Having said that, as @Cogent_Crypto pointed out, I do feel we all need to weigh the trade-offs, including Marinade not having that amount in the treasury.

The team may feel the budget impact for planning in the short run (I’ll leave other team members to comment on that) but in the long run, it’s money that did not go towards protocol development (albeit invisibly).

Interesting suggestion! In most protocols I’ve been involved in, token holders end up acting as the backstop, since the staked tokens can be slashed and sold at market. We don’t have slashing in Marinade, given that NFTs are always under your control. What you suggest would be similar, though, in that token holders would forego a percentage of income to ensure a backstop fund exists.

Please feel free to start a governance thread on Discord to read what people think (or if you can’t start one because of permissions, ping me and I’ll create one).

3 Likes
#7

While I wouldn’t consider my position a hard no, I do feel that the value of mSOL was not harmed greatly and materially I don’t think any one person noticed a difference. However a growing organization, MDAO, would greatly notice 2647 sol. I think having a strong treasury and using the DAO to build and fund community tools would be far more valuable then ever so slightly increasing the value of mSOL over time. This is coming from someone who holds far more mSOL then MNDE.

I agree with Ben’s opinion.
I think this 2647 sol does not harm mSOL values so hard.
APY of mSOL is not fixed. it fluctuates epoch by epoch.
How strongly this 2647sol affects that APY. I guess its within a range of fluctuations.
Probably the reason why I think this way is I am a long-term staker.

If this is for short-term stakers, I don’t think this kind of compensation is good for them.
This is because it takes for a few months and it happened a few months ago.
Imho, I think its a little bit too late to compensate for short-term stakers.

In addition, I agree with Socrates idea. Imo, Marinade should focus on how to prevent that in future. I like the idea of funds for the incidents.

2 Likes
#8

Great thread and conversation so far. I wasn’t around for the bug I guess so don’t have a first hand memory on the impact and reaction that users had. I do agree with @socrates that the need to stick to your word is priority #1.

Is there a public dashboard where we can view the treasury already? It would be helpful to have an idea of the impact of 3 months without revenue would do. Particularly from an expense perspective, not sure if anything like that has been disclosed either. Without knowledge of the internal finances of the DAO/Protocol I would still wonder if this repayment could be made over a longer period allowing at least some stream of revenue still going to the treasury? Would this redirection include early withdrawal fees?

I also think that coupling this slower repayment while also looking into some insurance could keep the lights on while also maintaining, if not improving Marinade’s brand. This DAO represents a crucial intersection of the entire Solana ecosystem supporting both Validators and DeFi. I think it would be extremely beneficial to establish a formal insurance policy to protect against future bugs or black swan events.

2 Likes
#9

Marinade will still receive share of liquid unstake operations (25% of that 0.3% liquid unstake fee).
This means, that there will be some stream of income for treasury.

1 Like
#10

Okay good deal, that makes me feel better about the redirection some.

#11

Counter proposal here… Consider this as a maximizing shareholder value of Marinade Chiefs

Increase the fee from 2% to 10% as for benefit of staked MNDE holders, both octopuses and sharks in (80-20% ratio).

Make the mSOL drop to the holders on monthly basis.

In case this brings to trying to escape from the mSOL - make the 700m TVL hostage of a 99% withdrawal fee.

In order to this proposal to be declined + mSOL holders should rush to buy the Chefs nft and vote against it…

And that will drive the MNDE token to go up + all the NFT ownership mandatory to secure themselves.

Remember that we are moving toward decentralized society and this opinion has to be told too…

I know that it’s too bad… But just think about it. How fun it would be

1 Like
#12

I’m leaning towards agreeing with this proposal, as I believe keeping our word and Marinade’s reputation is the number 1 priority. I would also support a separate proposal of gradually building up an insurance fund so such future events do not cause a financial shock to the treasury and to provide prospective stakers with extra confidence in the protocol.

Our TVL has been fairly consistent over the last 6 months, which I roughly calculate makes our monthly income from staking rewards fees around 850 SOL, x3 months is 2550 SOL. I’m undecided whether I would agree with capping the repayments at a max of 3 months if we get so close to fully repaying the target amount. But I can see if the TVL unexpectedly drops, we might end up paying for much longer than we anticipated.

I would also like to ask, is there a valid reason or security concern why it’s not currently possible to be able to deposit SOL without minting mSOL that I can’t think of? It would make life easier in the future if we can make a single payment from the SOL treasury (or potentially insurance fund in the future).

2 Likes
#13

I agree to this proposal.

#14

First of all, shoutout to the team for actually bringing this proposal forward and not just trying to sweep it under the rug like many others would. Like @socrates, I also think that both Cerba’s and Congent’s proposals are great ideas.

Given the team’s desire to make the community whole in the best possible way and the substantial amount of SOL that could potentially be diverted away from the DAO’s treasury at an important time, what do we think about letting the community decide how to proceed?

There are some really awesome suggestions out there IMO (more grants & insurance fund are two that stand out to me). What if we did a quick poll to get a temp check on this, leave it open for ~ 7 days, then take the most voted option to a formal vote?

5 Likes
#15

I think it’s an extremely good idea and I’m really happy of the alternative proposals that have been shared on this forum post. Would you include the original proposal in the temp-check vote or only the insurance funds and the redirection of those funds to grants?

I’m definitely behind this idea :pray:

1 Like
#16

Hi, let me try to answer some of the questions raised here

We at Marinade try to be open and transparent about all our financial numbers, however it is a long road to having a nice updated easy to read dashboard. :slight_smile:

For the first time we have published a lot of numbers, links to treasury wallets etc. in this recent article Marinade Kitchen Stories: March monthly update | Marinade.finance Let me quote:

"

As of April 1, 2022, Marinade holds in its treasury a bundle of mSOL, MNDE, and USDC accounting for $650,000 (at current prices)

For the month of March, Marinade monthly expenses were $118,691, or approximately 929 mSOL at current prices. $28,040 were also used in MNDE (unlocked by the executive team from the operational DAO treasury)

For the same month, management fees for the delegated SOL to Marinade and instant unstake fees brought in 690 mSOL, or $88,119 at current prices. Resulting in a net negative of 239 mSOL for the month of March, or approximately $30k at current prices.

"

So reading these numbers we can come to conclusion that treasury currently holds about 5 months expenses buffer if no income, and also that the compensation amounts to about 45% of Marinades treasury. So yes, at this moment it is a lot while it should not lead to destabilization as we have more mechanisms (cutting expenses, selling MNDE) that we can use in case the token exchange is delayed.

Instant unstake fee is unpredictable and lately has not been much. On the other hand we rely also on the token exchange program which should quickly increase our treasury.

I also want to react on the proposal of “security fund”. So generally we aim to grow our treasury (currently via the token exchange) and also aim to have strategies on diversification where the majority of treasury would be placed in DeFi and in theory the return would be enough to cover operating expenses of the DAO. Minority would be used for various purposes, one of them would be storing a part of the treasury conservatively so that we are hedged against crypto/solana risks and also so that we are always looking on a buffer for at least 24 mths (ideally 36mths) of operating expenses that cannot be affected by “crypto winter”.

This treasury may allow us to buy back MNDE when we consider the price to be low and see ourselves enough solvent to do so. Having such treasury that we can rely on, also means being able to cover potential future need for compensations. But, these are impossible to predict. They always require a decision on what is fair and doable. I do not see the benefit to somehow change our treasury strategy for a special fund. We will try to grow the treasury anyway.

4 Likes
#17

That is certainly our intent and the whole point of governance. :slight_smile: This is only how the team proposed to address it.

Sure, I like this idea. If the community decides to go with a different approach, I am happy to help structure the winning suggestion into a formal proposal.

It should be included, given that we have formally introduced it. Otherwise, we’re effectively saying it’s a no-go before we actually ask the community which one they prefer.

3 Likes
#18

That would mean to significantly change how the Marinade main contract works and after programming we would need to feel safe. And because Marinade contract takes care of millions of SOL, it is something we want to avoid.

On the other hand, changing one constant (percentage of rewards) is pretty easy, can be easily checked by holders of the multisig (which are currently various actors around Solana) and updated on chain.

2 Likes
#19

I’m undecided on the necessity of this proposal but, if enacted, I’d recommend something like 25% or 50% of fees be allocated to the catch-up reserve. Losing 100% of reward revenue for any period of time seems unwise to me. A fractional split would enable the protocol to reduce deficit spending during the catch-up period.

I also don’t think there is a strong sense of urgency - the loss happened over a month ago and we don’t owe any duty to short-term mSOL holders.

I understand the 3-month vs. 6-12 month payback period is somewhat superficial but it ensures Marinade has steady income flows and gives the protocol more flexibility if things change in the future.

3 Likes
#20

As agreed here and told on Discord, let’s put this discussion to a vote :slight_smile:

What would you be your prefered option between:

  • Cut the fees to 0% for 3 months or until 2646 SOL are earned and redirect them to the SOL staked pool to increase mSOL price
  • Cut the fees to 0% for 3 months or until 2646 SOL are earned to build a grant treasury to be distributed to community builders
  • Cut the fees to 0% for 3 months or until 2646 SOL are earned to build an insurance for eventual similar events in the future

0 voters

Depending on the most voted answer, a proposal will be written to formalize the terms and put through an on-chain vote soon after :slight_smile:

2 Likes