[PRFC] Fee Allocation - A Proposal for the Prisma Protocol

[wavey]

Title: Fee Allocation - A Proposal for the Prisma Protocol
Author: @wavey and @dudesahn
Date: 18-11-2023

Summary and Specification

Since launch, the Prisma protocol directs all fees to a simple holding contract called FeeReceiver.

This contract is controlled by the DAO and currently holds ~$3.5M in accumulated protocol fees.

This post aims to discuss and build consensus on the following design:

1. Convert all non-PRISMA fee tokens to sp-mkUSD (tokenized stability pool deposits).
2. Establish a 80% / 20% fee split between vePRISMA holders and a new Prisma treasury. This ratio can be configured by the DAO in the future.
3. Implement a smart contract system to facilitate weekly distributions of the 80% to vePRISMA holders (pro rata, or some alternate way TBD based on discussions) in the form of:
   • unlocked PRISMA
   • sp-mkUSD
4. Any fee balance already in the fee receiver at the time of initial distribution shall be distributed over a period >= 6 months to discourage a large one-time lump sump distribution.

Over the coming days, we will modify this proposal to reflect any consensus changes or improvements based on discussions.

Motivation

Like veToken protocols before it, a core part of Prisma governance design is to incentivize users to lock their governance tokens, creating long-term alignment with the protocol’s success. One important mechanism to incentivize continued token locks from users is to grant them a share of protocol-generated revenue.

If adopted, this proposal will require that all tokens in the FeeReceiver contract be distributed in the form of Prisma’s two major ecosystem tokens: sp-mkUSD and PRISMA. The rationale is the following:

• Distributing the many fee tokens as they come can create unnecessary amount of work/gas expenditures for smaller vePRISMA lockers
• This combination of tokens avoids adding sell pressure to mkUSD and PRISMA, while also giving distinct value within the Prisma ecosystem to users who lock.
• Revenue as sp-mkUSD means users are paid out in an asset that is productive for the protocol - it earns interest for the protocol while also keeping funds in the stability pool.
• Revenue as unlocked PRISMA provides lockers with a unique perk that almost no other protocol actor can enjoy.

An initial split ratio of 20% of weekly fees (as PRISMA and sp-mkUSD) will be transferred from the fee receiver to a protocol treasury contract at the start of each week. This amount will be used to for the following purposes:

• fund further protocol development efforts
• build a backstop for possible future bad debt
• provide a budget for bribes, or other incentives deemed important for aligning emissions with the growth of the Prisma protocol.

This split ratio will be the initial value, but shall be configurable by the DAO.

Next Steps:

• Following at least 3 days of comments / suggestions / improvements, this post will be updated and moved to the temperature check phase where it will be posted to Snapshot.
• If 30% quorum is reached signaling support for this proposal, the development team prioritizes building this fee distribution mechanism, committing to deploy and activate on a reasonable timeline.

[Alunara]

@wavey

Any fee balance already in the fee receiver at the time of initial distribution shall be distributed over a period >= 6 months to discourage a large one-time lump sump distribution.

Instead of adopting a somewhat arbitrary timeframe of six months, an alternative approach could be to aim for a specific minimum APR, such as 10% or 15%. Determining a reasonable APR by comparing it with yields from other CDP systems is likely more feasible than specifying a timeframe.

An additional advantage of this method is the need to only distribute sufficient PRISMA to achieve the desired minimum APR. This approach is reminiscent of the initial GLP staking on GMX, where a guaranteed APR of approximately 15% was established (can’t remember the exact amount).

In this case, extra esGMX was issued only when necessary to maintain the APR at 15% if stable yields fell below this level.

In pseudo-code, this can be represented as: if APR < 15% then add PRISMA rewards until APR == 15%.

The benefit of this method is that it would stretch PRISMA emissions as long as possible to reduce selling pressure and increase reliance on more stable yields.

[Contra]

Think it would be more beneficial to holders for fees to be distributed either in mkUSD or ETH/an LST to avoid mercenary farming. Would be nice to receive fees in an LST to incentivize depositing back into a vault to continue the flywheel.

[Sidn3yGottlieb]

The team has a few ideas for this that we will submit in coming days.

[CaptainRational]

Revenue for vePrisma owners should primarily come from Hidden Hand & Votium vote rentals. Market buy prisma, twap or otherwise strategically, and allocate as vePrisma to users that have opted into the rolling auto lock feature. This incentivizes users to auto lock, possible the prorata distribution done to auto lockers with a min time period that the auto lock has been applied. This ensures that distributions are sent to users that have max alignment with the protocol.

[secondsetmaze]

I do like the idea of making a determination in allocation somewhere for those who freeze (autolock) vs those who don’t have this feature switched on. Yes, freezing the lock keeps vote-weight from decay and that’s the inherent benefit, but a bonus in the fee allocation for those who freeze would be a strong incentive… Maybe on a tiered scale, like you get x% more every y weeks your lock is kept frozen. I’m reminded of multiplier points type model over on GMX/GLP.

[Sidn3yGottlieb]

Thanks @wavey for getting the convo going.

Some general notes:

• Annualized revenue seems to be around $5-10m right now (excluding PRISMA)
• Would like to propose, LSTs from redemptions be sold for mkUSD
• 80% of mkUSD be distributed as stabilitypoolmkUSD (yield bearing ERC4626)
• 20% of mkUSD to be added to the PRISMA/mkUSD (balanced using some of the PRISMA there)
• We are against using the PRISMA for now (seems we have enough to have good fee accrual, it would be interesting to release further down the line though)

An added feature we’d like to play with:

• we’d like to separate boost and fee distribution so that each locker can decide to allocate their voting weight to receive either more boost or more fees (much like cvxCRV option to receive CRV or fees).
• boost calculations use your allocated %, but the math is still based on the total lock weight (we don’t subtract all the weight allocated to fees). net effect, there’s just less boost to go around. boost becomes a more valuable commodity and there’s less PRISMA being released

Things that need to happen:

• Add PRISMA/mkUSD as a receiver (snapshot in progress)
• Release and audit of stability pool module
• Build and audit the logic to do the above
• All in all, I wanna stress to the community this isn’t a quick endeavor and that there are some things that need to be done to get there.
• It’s also important we focus on peg keeping mechanisms as mkUSD is yet again trading at 0.995 (and we’ve had a few redemptions earlier).

Other options we like:

• Only distribute some of the PRISMA to those with frozen locks (lower dilution to those most committed to the protocol (credit to @CaptainRational )

[daesu]

imo the goal of the fee receiver should be to alleviate the selling pressure of the supply that is released each week, ahead of rewarding prisma lockers.

Prisma dao should target a percentage of the prisma/mkusd curve pool liquidity as protocol owned liquidity.
To do this pair the earned prisma in the fee receiver with a maximum cap around 40% of earned mkusd (based on target of veprisma apy), given that prisma accrues at a faster rate than mkusd.

In the span of 6 months, 40% of earned mkusd represents 3m mkusd distributed to 25m locked prisma (50% of the circulating supply at this moment), ie. ≃15% apy for veprisma at a prisma price of $0.80.
So having POL doesn’t compromise offering a competitive yield to veprisma holders.

Applying this would deposit 690k mkusd and $690k in prima into the prisma/mkusd curve pool.

goal:
-establishing a robust and “locked” liquidity base that defines a price impact floor to enhance overall prisma price stability, particularly crucial in prisma’s scenario where supply surpasses demand due to valuable emissions.
-earned prisma yields positive effects without being transferred to potential paper hands.

A target portion of the total pool’s tvl as POL should be defined. Given the recent pool’s launch, a very high ratio could be achieved. A bonus being the trading fees earned.

If the POL exceeds 50% of the pool’s tvl, it could positively influence market sentiment. Investors would perceive a higher level of commitment, knowing that the majority of liquidity is effectively “locked,” providing more serenity in buying prisma (perception that it’s harder to decrease the price).

Adjustments to this strategy can be considered once the POL no longer constitutes a significant portion of the prisma/mkusd pool’s tvl due to external deposits.

i’m not convinced of spmkusd rationale as it could lead to the creation of a money market where spmkusd is leveraged 4x, resulting in sp emissions falling into the hands of mercenary farmers.

In return this will generate buying pressure on mkusd, and finally align the SP yield to the market rate (dsr). But is it truly beneficial to have additional idle liquidity in the SP, considering the high TCR that prevent liquidations?