Prisma has now launched its DAO around a month ago and has been on chain for three months. TVL today reached a new all time high of $400m. Growing Prisma hasn’t been without pains but the interest raise appeared to have helped the peg stabilize and the FeeReceiver has been growing rapidly.
Currently Prisma’s annualized fees are around $48m though we expect the true figure excluding PRISMA from penalty fees to be around $10m ($6m in interest rate and $4m in mint fees) or roughly $200k a week.
Prisma is currently the third biggest CDP protocol behind MakerDAO and Liquity.
80% / 20% Split
Any LST fees will be sold for mkUSD (our preference would have been to distribute the LSTs direclty but gas on mainnet makes that prohibitive)
Up to 100k mkUSD from the feereceiver will be distributed weekly (this is a parameter that can be modified)
Another 25k mkUSD will be paired with $25k PRISMA and added to the mkUSD/PRISMA pool weekly (the LP position will be owned by the DAO in perpetuity)
Remaining PRISMA in the fee receiver will remain there for now
If there is less than 125k mkUSD in the fee receiver, 80% of the mkUSD would be distributed to vePRISMA holders and the remaining 20% will be added to the mkUSD/PRISMA pool
The 100k cap can be adjusted. If the protocol kept growing rapidly and its revenue exceeds expectation, the DAO could change it.
The above achieves several things:
Does not make emissions of PRISMA higher by redistributing penalty fees
Guarantees vePRISMA holders are incentivized
PRISMA locked for a year would yield approximately 38% APR at current prices
Prisma in fee receiver can be used as a backstop or to bribe in the future (with the DAO’s approval)
The idea is to sustain and stretch the amount accumulated in the fee receiver for years to come and being able to counter balance emissions lowering in comings weeks/months.
My understanding is there are over 4,2 million $ in fees already in the fee_receiver.
In addition it generates 200k $ / week. In adition the DAO is building quite a fund, with 25k mkusd going into the curve pool.
Therefore the cap of 100k $ for vePrisma holders seems too low. It would just increase the millions in the fee_receiver.
I also agree it should be set somewhere around 400k range. Assuming 200k per week, this will let the backlog clear out over a few months avoiding the 1 week of super high returns but still properly flushing things out.
Distribitute 100% of the fees earned each week. (20% with a maximum of 25k$ for the mkUSD/PRISMA pool), the rest for the vePrisma holders
Distribute the fee_receiver buy spending it linearly over 52 weeks for curve bribes to incentivise the stability pool, to attract a higher TVL.
We need to get the flywheel started:
High fees => higher TVL, => higher emissions => higher bribes => higher APR => higher TVL…
A cap is wrong:
It stops the flywheel before it takes off and does not fit in my ponzinomics excel sheet.
I dislike very strongly the idea: “The cap can be adjusted if the DAO agrees”. We should have clear fee distribution rules, so investors know what they buy into. Prisma is uninvestable for fundamental investors, if you can not calculate the fees without politics.
The point of the POL not generating revenue is to not dilute PRISMA/mkUSD LPs. We can change it no problem but I don’t think it’s a good idea.
Same as the calls to immediately make the cap higher than $100k, the whole point is to make the fees collected early on from the launch to stretch and be distributed to those locked for the long term.
It’s also evident it’s better for Yearn and Convex who could suffer if the APR was to spike massively and then drop if revenue is lower (if IR was to be lowered for example) ( cc @C2tP unless you are happy for prisma lockers to be extremely volatile assets which we are not).
More more more now isn’t the long term vision that the DAO should have.
Rasing cap was just saying if you were to try to fit everything in this cap system to distribute the current built up fees in the receiver then you would have to raise to something more than weekly intake or else it’s not distributed (fee receiver would grow instead)
This doesn’t really require raising the proposed cap per say. If you want to keep some cap on weekly fees but still distribute the launch fees slowly you probably need to pull those fees out and then linearly vest them to the distribution logic. Thus 100k(weekly cap) + vested. This can be whatever timeframe, like a year etc.
So I guess it’s more of a question of do you want to make sure the launch fees are distributed along side some weekly fees? Or do you think we should hold them as a buffer to be used if we ever dip below the weekly fee cap?
The logic would just up to 100k mkUSD for fees on a weekly basis from the fee receiver. There is no difference between fees previously accumulated and fees that will accumulate after fee distribution.
If revenue becomes lower (less mint fees or lower IR which is highly possible), we can still have strong fees being distributed. If revenue keeps getting stronger, we can have a higher weekly cap in a few weeks.
I agree with @Blue that it’s hard to trust the “vote to change it” approach, but I also understand @Sidn3yGottlieb’s reasoning for sacrificing higher distributions now to build a strong balance sheet. It’s wise to be cautious with the fee receiver at this early stage, given the volatility.
@Sidn3yGottlieb, it would be helpful if the framework or revenue milestones for the fee cap were clearly defined. This would elucidate how the DAO is thinking about the issue and make any future cap changes less contentious.
MDR has a point.
If you want to cap the fees in the beginning, it is better to define the future fee changes already.
Say for how long there will be a cap and when the cap will be gone.
No cap is the best cap and maybe really invest the fee_receiver funds into CRV to incentivise liquidity long term.
It’s difficult to model future fees. I see the current interest raise as provisional because it was to address the peg struggling which did so fairly well so far. A vast majority of fees going forward could then come from minting fees which are fairly high right now but a good share of them also aren’t organic and we’d like incentives for minting to be lower.
The stance we’ve taken is simply a cautious one in response to user demand that those fees be distributed.
There was a request for comments from @wavey and we’ve coded something in the past weeks based on that and community feedback. We can go back to the drawing board if you aren’t satisfied if what’s being proposed.
If revenue is consistently high in coming weeks then I see no issue in raising the cap. For now, I think it’s wiser to let emissions and the amount of locked PRISMA before distributing fees too quickly.
Understood. I am not suggesting a different course of action. You all are managing the situation honorably. I am just trying to understand how the DAO thinks about these things. Your answer is beneficial in this regard.
I agree with your conclusions and would have made similar decisions in your seat.
I am one user who gave his five cents, maybe this is helpful.
I will stay a user if you go ahead with your proposal, because I trust in your execution. No need to go back to the drawing board.
Would like to say that my comments above about cap was just me trying to process how current fees in receiver would be processed. I think its also a valid choice to process them along side weekly and adjust as we go.
With that point out of the way, as far as distributor mechanics go I support the proposal and its current settings.
In favor of the proposal but I don’t understand the idea of capping the fees distributed if this money is not used for anything else specific.
Why not distribute all the mkusd earned every week from now on? Except save 200k mkusd over a year to close the gap between prisma and mkusd balances in the fee receiver.
If fees and yield go down, that’s not bad, it’s just the reality of the state of the protocol, that transparency is required. Long-term lockers have conviction, they already tolerate price volatility, they can tolerate potential yield volatility. Would-be lockers also have to factor this. It is better to attract lockers with a higher real yield which may eventually decrease naturally rather than to cap this yield and take the risk of attracting fewer lockers in exchange for a stack of mkusd with no defined use case.
So imo veprisma holders should receive all mkusd fees collected every week. Prisma and mkusd in the fee receiver are used to fund the POL in the prisma/mkusd pool in increments of $50k once a week as described. It will already take around 100 weeks to add the $5m in the pool.
As the POL is added weekly and not all at once, fees already collected can still be used in case something goes wrong.
