Revisiting The MIN Staking Mechanism [MIP 2.2] - Temp Check

[CWSchub]

Background

On June 28th the Minswap staking program saw its first set of 9-month staking contracts expire. A few weeks later, Minswap V2 started rolling out and there is a high expectancy for higher rewards to MIN stakers thanks to the newly implemented features. Looking back at how the staking model has performed we have identified areas for improvement based on community feedback and our observations over the past few months.

For further in-depth reading, kindly refer to the written report and the mathematical approach that supports the basis for the proposed solutions to improving MIN Staking.

MIN Staking As A Supply Regulation Mechanism

The Minswap Staking Program reached it’s ATH at 442.2 Million MIN staked, representing ~35% of the circulating supply. Over the last year, Minswap has achieved a cumulative trading volume of approximately 2.73 Billion ADA, equivalent to roughly 1,365,000 ADA being distributed back to MIN stakers.

The Current Main Concerns

Based on community feedback and our own observations over the past few months, we have been able to identify the following main concerns:

1. Duration of Staking Periods
2. Negative Reinforcement Loop
3. LBE Rewards Utility
4. Incorporation of Industry Volatility
5. Predictability of MIN Trading Volume and Volatility

These concerns are tightly tied to the soft lock condition imposed on users who must forfeit their rewards in order to withdraw their stake before the contract expires.

Scope of Change

The changes this proposal aims to present as solutions, correspond solely to the rewards accrued from the Fee Switch. Therefore the aim is to answer the question “What can be done to improve the distribution of Fee Switch rewards?”.

This proposal will present three solutions.

Solution #1 - One single 9-month contract.

Solution #1 aims to reduce the amount of staking contracts available, as its design makes all but the 9-month staking contract redundant.

Breaking the negative loop with Milestones

As explained earlier, forfeiting rewards completely due to early withdrawal represents an unnecessarily aggressive approach to incentivise locking up participants’ MIN tokens.
On the opposite side of the spectrum in “staking” models, we can find Indigo Finance’s flat-APR-untiered liquid model, which rewards users each epoch.

In order to strike a balance between the tiered approach used by Minswap and a liquid staking model, it requires introducing the concept of Milestones

Milestones Explained

The current staking contracts are as such:

1. Unboosted contract - 1 month duration. (1)
2. 3x boosted contract - 3 month duration. (2)
3. 6x boosted contract - 6 month duration. (3)
4. 9x boosted contract - 9 month duration. (4)

A Milestone will be considered the amount of months the staking contract has matured.

Therefore

Early Redemptions Explained

If a staker is unable to complete the predefined contract duration of Y months, they will be able to withdraw their stake and receive their early redemption rewards.

This is equal to applying an early withdrawal fee:

Redundancies in the Current Model with Milestones.

If we were to introduce the concept of Milestones in Minswap’s current staking model, the following examples will show that the 9-month staking contract will be the preferred contract by default.

1. Completing the unboosted 1 month lock (1) is equal to achieving the 1 month Milestone in the 9x contract (4).
2. Completing the 3 month lock (2) is equal to achieving the 3 month Milestone in the 9x contract (4).
3. Completing the 6 month lock (3) is equal to achieving the 6 month Milestone in the 9x contract (4).
4. Completing Milestones in staking contracts (1), (2), and (3) provides exactly the same rewards as completing them in contract (4)

Therefore, from a staker’s point of view, implementing Milestones would mean that there is no situation where the 9-month staking contract isn’t the preferred one, as it has a longer duration and performs just as well or better than the other contracts .

With these redundancies in mind, it no longer seems necessary to have an array of options to choose from. Therefore only one single 9-month staking contract is needed.

Changes in APR explained

In order to explain how the staking APR would change, we will take the data available on September 18th 2024.

By shifting to a single-contract model, all participants will enjoy the same APR, as rewards are distributed equally to each MIN token.

At the current values:

As participants decide to withdraw their stake, they receive their Early Redemption rewards. The forfeited rewards are then distributed the following month.

The following graph shows how APR would change over the course of 48 months, considering the amount of staked MIN remains the same and the rewards from the Fee Switch remain the same to, at 90k ADA.

Solution #2 - Liquid un-tiered staking.

This solution eliminates the concept of tiered MIN Staking. As a result, Fee Switch rewards which have accumulated on a monthly basis are distributed equally to all participants based on their weight in the Staking Pool.

Therefore, the total APR (tAPR) calculation becomes:

At the current values:

Solution #3 - The Hybrid Solution. Two Contracts.

The Hybrid solution strikes a balance between both solutions #1 and #2. Allowing Staking participants to choose between two different staking contracts:

1. A liquid staking contract which ideally yields ADA APR equal to the ADA delegation rewards + MIN rewards + LBE token rewards.

Rewards would be distributed so as to have a competitive untiered MIN staking option (besting ADA delegation thanks to additional MIN and LBE Token rewards).

2. The 9-month tiered staking contract, described in Solution #1.

All the remaining rewards would be funneled to the 9-month staking solution, increasing the APR by a heavy margin during periods of high trading volume.

This solution must contemplate two different case scenarios:

1. The flat ADA APR from the Fee switch is higher than the average APY from ADA staking on Ouroboros PoS.

2. The flat ADA APR from the Fee switch is lower than the average APY from ADA staking on Ouroboros PoS.

Case A

This scenario is a positive one for Minswap as a whole. As it indicates that sufficient trading volume is flowing through Minswap so as to beat ADA delegation.

In this scenario, the Hybrid solution would perform the following actions:

1. Calculate the weights of Staked MIN in the Liquid Contract (LC) vs the Tiered Contract (TC)
   
2. Calculate the flat staking APR (sAPR) for ADA-only rewards and obtain the average ADA delegation rewards (dAPR).
   
3. Distribute staking rewards as follows:
   

It must be noted that the Tiered Contract would still be functioning as described in Solution #1.

Case B

This scenario is suboptimal for Minswap, but must be taken into account.

The only potential issue here is when gamma >1. This means that the ADA rewards distributed to MIN Stakers via the Fee Switch has a lower APR than the APR obtained by delegating ADA.
In that case, the unwanted behavior can be bypassed by setting gamma = 1, which will reduce all staking contracts to Solution #2, where the same flat APR is shared.
This could seem unfair for the users who selected the Tiered Contract, as they would be seeing the same rewards as the Liquid Contract stakers. But it would be a necessary adjustment to avoid breaking the Staking mechanism.

This would effectively nullify the benefits of locked staking. At least temporarily, if APR goes down, the amount of staked MIN would also naturally go down, pushing APR back up. Also, any withdrawals in the locked staking would also push the APR back up due to the early redemption fees.

Calculations Using Current Data

Kindly refer to the written report and the mathematical approach that supports the basis for the proposed solutions to improving MIN Staking.

Implementation

After compiling feedback with the Minswap Labs technical team, if any of the proposed solutions were to be passed, once developed, they would be implemented as following:

Solution #1

If solution #1 were the most voted option:

1. Disable 1,3,6 month locked options and maintain “old” APR.
2. Enable Early Redemptions to all active staking options based on contract maturity.
3. Once the last 6-lockup expires, rewards in the 9-month lockup will be set to the “new” APR.

Solution #2

If solution #2 were the most voted option:

1. Disable 1,3,6,9 month locked options and enable Liquid Staking Option.
2. All active staking positions are set to receive the same APR.
3. Disable early withdrawal penalty.

Solution #3

If solution #3 were the most voted option:

1. Disable 1,3,6 month locked options and maintain “old” APR.
2. Enable Early Redemptions to all active staking options based on contract maturity.
3. Once the last 6-lockup expires, rewards in the 9-month lockup will be set to the “new” APR.
4. Enable Liquid staking option, with fixed APR (as defined in Solution #3) and no early withdrawal penalty.

Proposed Compensation

As detailed in the written report, the proposed compensation is 720$ in MIN Tokens to the Author and 110$ in MIN Tokens to Elder Millenial.

Conclusions

This report has presented 3 different improvements to the current MIN Staking Mechanism.

Solution #1 rewards all users equally and enables a scaling solution for Early Redemptions, incentivising longer staking periods.

Solution #2 rewards all users equally and allows participants to withdraw their stake without an Early Redemption solution. This lowers the APR and is more risk-averse.

Solution #3 appears to be the most rewarding product given the fact that it allows participants to choose their favourite contract. The hybrid option manages to balance risk aversion vs. risk appetite while rewarding both in proportion.

Voting Options

The on-chain governance proposal will allow for the following voting options:

1. Keep MIN Staking as is (No Changes)

2. Solution #1 - One 9-month contract with Early Redemptions and scaling rewards.

3. Solution #2 - One liquid staking contract without Early Redemptions and flat rewards.

4. Solution #3 “Hybrid” - Two contracts: a liquid staking option with bound APR and a 9-month option with Early Redemptions and scaling rewards.

Special Thanks

The author of this proposal would like to thank @long.ngn, @Ch_cken and @richard.minsw for their insight and contributions throughout the writing of this proposal, and Elder for reviewing the code written for the analysis on Solution #1.

Author: CWSchub

Image Edit: The image in Solution #1 showcasing the change in APR has been updated to reflect current values

Edit 23/09/24: Updated Proposal to include a compensation plan for the Author and collaborators

[Ch_cken]

Well done brother! I know you have been at this for a while. My feedback.

1. This is the average hourly for a research analyst. Please add a comp for yourself and Elder if this passes.

https://www.ziprecruiter.com/Salaries/Market-Research-Analyst-Salary--in-New-Jersey

2. The ideas and supporting research make sense. I am curious about the dev work required to do this and whether this is feasible to implement. It would require a noticeable change to the staking contracts that house 35-40% of the supply of the MIN token. I would not be willing to use them unless they have undergone rigorous testing. 1 small exploit and it can lead to a catastrophe. I believe it also requires a UI/ UX change so we should factor the dev time for that too.

3. It looks like both solutions allow liquid staking. The aprs increase as stakers leave. So it becomes easier to market as people leave vs people looking at the 9 month contract rn and seeing it pays a higher apr. So just wanted to see if (depending on the dev complexity) we can have an interactive chart that lets users define amount of min staked to see how the rewards change with time and milestones achieved.

Eitherway I see that it is a price worth paying to have a more flexible staking option.

Will add more thoughts as I think of them.

[marco_112358]

Thanks for this sir. I wouls be in favor of looking at option 3 as a new potential implementation

[CWSchub]

Thanks so much for your support and the comments.

1. I will shortly edit both the proposal and the report to reflect the work put in and the requested compensation.

2. I believe solutions #1 and #2 are technically uncomplicated, while #3 would require real-time balancing the APR. Hopefully we get a comment from MinLabs clarifying these points.

3. Yeah, completely agree that by reducing the amount of contracts, more emphasis can be made on the UI, providing more historic data, etc.

[richard.minsw]

I take the responsibility on the MIN Staking implementation. Here are few things on technical implementation I can break down at the moment:

• Cut off 1 month, 3 months & 6 months lock : We can prevent new participants from staking in these options, but we must continue distributing rewards to the existing stakers until all their positions reach their respective end dates. This process will take up to 6 months, in the case where a user stakes their MIN with a 6-month lock-up just one day before the new mechanism is applied.

• Introduce Early Redemption Rewards: This mechanism can be implemented as soon as possible and can function alongside 1, 3, 6, and 9-month lock-ups.

• Liquid Staking : This mechanism poses no significant technical blockers, aside from maintaining a consistent APR. The challenge lies in the fact that APR is influenced by the MIN price and the number of MIN being staked, both of which fluctuate constantly.

[Swoops]

Random question: Wouldn’t early redemptions actually reduce APR? Normally that yield would go to stakers if people pulled out early. It supports min price if they hold until the remainder of lock term.

So I guess my real question is what this change is looking to solve?

I get wanting the staking system to seem less punitive but the soft lock always seemed like a fair compromise there.

If we are budgeting funds to increase adoption of $Min is this what we want to spend dao funds on?

I just wonder if the current staking is a real barrier for people or if this would just be a nice thing to have down the line?

Huge gratitude and appreciation for everything you put together here. Curious to see what others think about this as well

[CWSchub]

Hi there! Thanks for your questions.

In order to answer your first question, I’ll try and explain Solution #1 a bit differently, in order to clarify how it works.

Solution #1 proposes eliminating the full penalty to stakers as well as unifying all staking contracts to a single 9-month contract. Therefore all stakers will be enjoying an initial flat APR, which will be equal to all stakers (as calculated in the examples, current data would set APR at 7%).

This means APR will initially be reduced slightly, from 7.3% to 7% based on current data. The premise of rising APR is based on the idea that not all participants will complete the 9 months.

If a user doesn’t complete the 9 months, they will still be able to receive their Early Redemption Rewards (proportional to the maturity of the contract), while the forfeited rewards will be redistributed among the rest of stakers. This redistribution is an increase to the base APR.

Therefore solution #1 doesn’t guarantee an increase in APR if ALL participants follow through with the 9 months.

On the other hand, solution #3 guarantees an increase in APR to participants who choose the tiered contract over the liquid contract (the tiered contract functions as defined in sol. #1). While the liquid contract guarantees higher APR than the first few months of the tiered staking contract.

As per your question regarding investing dao funds to this. The dao funds spent on the development of this proposal will only be those requested in the compensation section.

Lastly, the issues this proposal aims to solve have been outlined in “current concerns” and detailed in the written report.

[CWSchub]

Thank you so much for your support! I too believe that Solution #3 offers the best conditions for an attractive staking product.

I can also see improvements to this staking model happening further down the line. Where new sources of rewards can be introduced, and a better functional dependency between liquid and tiered staking can be established.

[JBZ]

Very thorough and well
written

Im going with option 3

[JBZ]

Very thorough and well written

Option 3 all day

[PurritoGeneral]

Thank you @CWSchub for the great Proposal and apologies for not having provided feedback before.

Have to say, at the beginning, I was skeptical. My main point is that the current Staking model doesn´t seem bad. It optimizes for rewarding long term committment and those who provide something to the protocol (taking coins out of circulation) and without punishing too heavily those who exit early. This is something “liquid staking” (LS) doesn´t do. But LS does represent an attractive offering. One of the biggest issues with $MIN is the disparity between risk/reward for staking vs LPing. LS may help converting MIN/ADA LPs to MIN stakers. Framing the proposal with that background and aim, it makes much more sense to me.

To summarize, Option 3 seems to be the preferred one. So it comes with 2 options:

• A liquid staking option: with lower APR but ideally ADA APR equal to the ADA delegation rewards.

• A 9-month tiered staking option: with higher APR but where rewards are not fully forfeited for early withdrawal and instead depend on how long MIN has been staked.

This proposal is great on a theoretical level. But let’s get to the practice. In terms of the technical implementation, it looks to me like we need to solve for how we determine rewards for MIN liquid stakers. 2 options come to mind:

• Set a fixed ADA APR for Liquid Staking MIN: which is tough because this APR does depend on MIN Price. And we can´t predict Trading Volume either. We could set a range, let’s say 2% - 6% $ADA APR. We could also stop accumulating Batcher Fees for the DAO Treasury (which has millions of $ada already) to supplement this yield (but that is a separate discussion)

• Set a fixed ADA amount for Liquid Staking MIN: such as a X% of all ADA accumulated that month. But its hard to determine this number and fix it because we cant predict MIN staker behaviour.

Thoughts?

Also, you noticed I didn´t talk about Launchbowl tokens or MIN rewards for MIN Stakers. Those are separate topics that we will have to work on to address the larger issue I mentioned above (MIN Staking vs. MIN/ADA LPing)…

[PurritoGeneral]

Just a small correction. When you use the word “contracts” that is actually not correct. These aren’t/wouldn´t be different contracts. All the rewards calculation/calculation is done off-chain (very common on Cardano). So the “contract” is just a wallet, that allows us to see which wallets stake how much MIN and then we distribute rewards according to the calculations.

So not really possible to audit it, but that’s why we have a 3 person QA team for.

Correct me if I am wrong @richard.minsw

[Ch_cken]

@CWSchub glad you added the comp

One additional thing I would do is modify the scope of change to include “what exactly the proposal would impact for voters if passed” and summarize each option into a bullet point for anyone who wants a tldr.

Also add MIP - 2.3 as a subheading so we can keep track of the proposals. I believe the last one in governance was 2.2, but just check once more

[CWSchub]

Thanks for the support and comments!

I think the current approach to MIN Staking is reasonable, given that it’s the first iteration since its inception.
After having done the research, the result is clear. The product which is MIN staking must not rely on the price performance of MIN itself. This is because MIN Staking has delivered at the highest 14% APR and at the lowest 5% APR, while the token itself has depreciated over the course of 12 months. Therefore the staking product must offer some early redemption rewards to compensate the losses.

You have pointed out an interesting comparison between LPing and Staking. But LPing does expose one to impermanent loss, and that requires research which has not been performed for this proposal.

Regarding the practice, the formulaic approach to Solution #3 covers all the possible scenarios. The only issue is the difficulty in terms of technical implementation.
Solution #3 fixes the value of gamma on a periodic basis and calculates the distribution for liquid rewards depending on the value of gamma. The rest of the rewards get distributed evenly in the tiered rewards.

You have mentioned my excluding MIN rewards as well as LBE.

1. Since MIN rewards still come from fee switch, I don’t think they need special treatment.

2. I think a future proposal will be required to separate LBE rewards from being slashed by the Early Redemptions fee. I realise it’s an issue, but treating LBE rewards separately ought to be discussed too in a separate proposal.

[Swoops]

Thank you for the reply.

@CWSchub

If a user doesn’t complete the 9 months, they will still be able to receive their Early Redemption Rewards (proportional to the maturity of the contract)

In the existing system if a person removes stake prior to the maturity date they forfeit earned rewards and those are redistributed back to stakers (generating an improved yield)

Introducing early redemption rewards reduces that improved yield correct?

Only a portion of the forfeit rewards is being paid to stakers who remain until maturity. If this plan is accepted it would be a reduction to the current apr stakers enjoy (assuming early withdrawals are constant).

@CWSchub

As per your question regarding investing dao funds to this. The dao funds spent on the development of this proposal will only be those requested in the compensation section.

I was actually concerned about the cost of developing this new staking system, not the work you did to submit this proposal. You have done a great job and I will vote to comp you.

[CWSchub]

@Swoops

Introducing early redemption rewards reduces that improved yield correct?

It really depends on what the expected withdrawal rate is in the new staking model.
For instance, in my example, the flat APR is 7% while the APR with the current staking mechanism is 7.34% (corresponding to last month’s data).

We don’t know how many staking positions will withdraw early. The model I used to predict the increase in APR is a simplification, but it assigns a random number (1,9) to each active staking position and evaluates the increase in APR over time.

This model will be inaccurate to a certain degree, as there could be a tendency towards longer/shorter staking periods depending on market conditions.

But in any case, the minimum APR Solution #1 would assign is equal to the flat APR, that is, if all stakers withdrew their stake after contract-expiry.

What this change provides is for participants to choose whether they want to continue staking for 9-months or withdraw if they’re satisfied with their rewards beforehand. This is something that is not possible currently.

@Swoops

I was actually concerned about the cost of developing this new staking system, not the work you did to submit this proposal. You have done a great job and I will vote to comp you.

Thanks a bunch. Regarding costs of development, I’m afraid I made the mistake of not including this. But at the same time, Minswap Labs has a source of revenue for development & maintainance which are the batcher fees. But I’m not able to answer in more depth.