The External Governance Team of ChicagoDAO has reached a unanimous decision to vote against the COMP Reward Adjustments v2 Proposal with our 50k delegated COMP tokens from A16Z. The reasons for this vote are as follows.
There is no clear alternative for a reward mechanism. Although there are obvious inefficiencies in the current system, ending the current incentive plan would be an act of self destruction – especially when considering the initial benefits from the reward mechanism. Additionally, a period with no reward mechanism could initiate a run from Compound to other competitors, permanently hurting the community.
Ending distribution of the COMP token would impact the quality of decentralization in the Compound community. By capping the production of COMP, the community is more vulnerable to centralized overhauls. Thus, the current reward mechanism is not simply an incentive for further contributions, but it also provides a way of stability for the decentralization of the community.
In conclusion, ChicagoDAO does not oppose the idea behind the proposal. The current reward system is outdated and needs further attention to promote the growth of the community. But, complete removal of the current structure would be premature causing fractures in the reputation of the community.
We will be sharing further reasoning, qualifiers, and rebuttals to common For arguments on our Twitter later this evening—find us at @chicagodao_io!
Madhav Vats | University of Chicago
On Behalf of Chicago DAO - External Governance Team
It would be logical to actually present his new vision of new rewards. There is more than enough COMP available already after first 50% reduction to start anything he might have in mind.
And when comminuty could see results of new program we can revisit COMP distribution again but that time with much more data to back the proposal. There never was a reason to rush into anything. COMP rewards even before fist 50% reduction were scheduled to go for years. After reduction it can run for years AND have additional incentives implemented also.
There is very minor difference if the rewards stop already or stopped several months later. If new distribution model will be so good it would be easy to convince everybody with amazing results.
So far it’s really not that obvious and voting results reflect that. It’s very big difference in comparison with voting for something with what majority agrees. You can see that some chose to abstain, which kind of reflects lack of backing behind pitch for new model. Nothing stops Tyler from presenting his vision currently. I’d say actually failure to present it at this point just reinforce argument that there was nothing really, and the whole point was just to stop existing distribution “because we don’t like it” and there are farmers who make money on it. Well, wasn’t there analysis which pointed that vast majority of recursive farming positions are not in profit after first reduction already.
This is an extremely strong way to state this. Gauntlet is being paid to run some simulations to give a more statistical basis for determining risk of insolvency, and where there is space for parameters to be tweaked without increasing risk they make a proposal.
I would expect a statement from them if reserve factors were to be changed in a way that created risk of insolvency, but they do not control the reserve factor, nor are the tasked with finding creative new uses for it.
Here’s my summary of why this second proposal failed:
There hasn’t been enough time to evaluate the effects of the first proposal and voters are fearful of the negative consequences of completely removing the existing rewards.
Voters don’t see a clear path to decentralization without the existing program.
The action plan for kickstart rewards is really simple and I expect the first proposal for them (with the first new market) to achieve consensus fairly easily. It goes like this: incentivize $50M (or 10% of circulating supply, whichever is smaller) at an annualized rate of 8% over 3 months, then cut the rewards. We assume at least 20% (or $10M of that initial $50M target) will stay in the supply market (this is our starting hypothesis).
Factors influencing the correctness of our assumption:
Interest rate model(s)
Reserve factor
Usefulness of borrowing the newly supported token
Collateral factor (and borrowing factor when it’s implemented)
We’ll continually analyze the correctness of our hypotheses and will continue to test new ones, making adjustments to IRMs, RFs, and market addition priorities as warranted.
The action plan of the COMP rewards adjustments as per kickstart rewards isn’t exactly linear - we can and will go ahead without cutting the existing rewards program to 0. While everyone is paying attention to this thread, please speak now on the numbers I presented; otherwise, I’ll go ahead with those.
I’ll loop back to cutting the rewards (or halving them again if that’s what the majority wants) in a matter of months.
As for additional rewards programs, those will take time to develop as they’ll change how we organize ourselves as well as the direction of the protocol. I have ideas in mind, but we’re subject to the inefficiencies of bureaucracy. This process can be expedited by having our voters (more importantly, our largest voters) provide quick and thoughtful feedback (which has proved to be very challenging).
Glad the thread will continue; the kickstart proposal here already goes a long way toward addressing some of the earlier questions about ambiguity of the plan forward.
Quick question to check my understanding: if a new market attracts $200M in supply, the COMP kickstart reward to suppliers in this program would be diluted to 2% (4x dilution of the kickstart rewards on the target $50M); am I interpreting that correctly?
Kickstart rewards will still likely attract mercenary capital to farm-and-dump COMP, but given the evidence we’ve seen of stickiness in supply markets, this could still be an effective way to kickstart them.
There is at least one key difference between currently established markets and new ones that could affect stickiness: the currently established markets never had (and still don’t have) a pre-determined end date for their COMP distributions. I wonder if baking the 3-month timeline into the proposal in advance will make this capital less sticky by allowing suppliers to plan their next moves in advance? It feels like it should be possible to guess the answer with data from other protocols that have offered fixed-term rewards, but empirically it seems like every case has unique features making it difficult to generalize.
I’m an observer of this extremely close round of voting regarding the Rewards incentives and I would like to suggest an alternative way of creating perhaps a more sustainable rewards solution. I can understand the limited (some people have commented that this is negative) value a continual rewards program brings; and that some are concerned about the system shock of an abrupt cut in rewards.
Perhaps one way to redirect and focus rewards is the usage of KPI Options. This community values organic growth and behavior that is beneficial to the community, so perhaps if rewards can only be paid out upon additional criteria beyond the staking of capital, maybe this can also be considered. The payout/reward will be variable, but directly tied to how much of the KPI the community has achieved, which makes this an efficient form of incentivization. I envision that an application of KPI Options can replace the existing program, with each Option having an expiry of 1-3 months depending on the KPI measured, and addressing the potential need to wind down the existing rewards program and for bootstrapping new initiatives where a clear need for capital, governance, or even any sort of pain point has been identified.
The specific KPI can be further explored, and we will be more than happy to brainstorm with you on defining this further if there is initial interest among the community to proceed. A very simple example would be a TVL of a specific market or pool.
KPI Options can also be used to mitigate whale effects on the ecosystem by considering a community-wide distribution in addition to the KPI Option itself being a source of emissions. In this case, the effects of community participation is partially socialized. This system is being trialed on other project integrations we are currently working with.
Finally, KPI Options can also be used as a method of compensation for contributors. This is the model that the contributors at some DAOs are exploring or using.
Happy to take any questions/feedback and appreciate everyone to read through this post.
Disclaimer: I am a contributor at the SuperUMAns, a sub-DAO of UMA.
This was really close. I wonder what would have happened if cCOMP token holders whose COMP is locked in the protocol as collateral could have voted? I know there was some talk of this long ago but not sure what ever came of it. It makes sense that a lot of COMP longs use the protocol to buy more COMP and also to borrow against gains tax free.
Quick question to check my understanding: if a new market attracts $200M in supply, the COMP kickstart reward to suppliers in this program would be diluted to 2% (4x dilution of the kickstart rewards on the target $50M); am I interpreting that correctly?
That’s correct!
Kickstart rewards will still likely attract mercenary capital to farm-and-dump COMP, but given the evidence we’ve seen of stickiness in supply markets, this could still be an effective way to kickstart them.
Possibly, but it depends on various factors: exchange liquidity and slippage, holder distribution, ability to borrow the token elsewhere, etc. Buying large amounts of a token to temporarily farm COMP with may become unprofitable when it comes time to sell the token after rewards dry up.
There is at least one key difference between currently established markets and new ones that could affect stickiness: the currently established markets never had (and still don’t have) a pre-determined end date for their COMP distributions. I wonder if baking the 3-month timeline into the proposal in advance will make this capital less sticky by allowing suppliers to plan their next moves in advance? It feels like it should be possible to guess the answer with data from other protocols that have offered fixed-term rewards, but empirically it seems like every case has unique features making it difficult to generalize.
Yeah, this will be hard to generalize. It really depends on capital efficiency in two regards: the ability to borrow against it, and the interest rate on depositing compared to other avenues. This creates the need for a quality asset listing strategy.
Thanks for keeping the momentum going here @TylerEther. I want to boil down my understanding of your kickstart proposal to make sure I’m on the same page.
new market launches
the first $50m of (supply? supply and borrow?) capital into that market earns an 8% annualized COMP incentive and anything over $50m dilutes the incentive linearly
after 3 months the rate of the incentive will be adjusted down to somewhere less than 8% and greater than or equal to zero
can you clarify what you mean by:
We assume that at least 20% (or $10M of that initial $50M target) will stay in the supply market
the first $50m of (supply? supply and borrow?) capital into that market earns an 8% annualized COMP incentive and anything over $50m dilutes the incentive linearly
Correct. Supply will be incentivized and borrowing activity will have to develop organically.
after 3 months the rate of the incentive will be adjusted down to somewhere less than 8% and greater than or equal to zero
I’m thinking immediately down to zero after the 3 months. Then supply and demand will take over.
can you clarify what you mean by:
We assume that at least 20% (or $10M of that initial $50M target) will stay in the supply market
So assume that by the end of the three months reward period that there’s $50M in supply. The goal is to have at least 20% of that supply stay in the protocol.
$50m seems like a reasonable target based on the current market stats. Given the compressed yields market-wide, 8% should be enough to make lending appealing. And cutting to zero is a good idea from a discovery perspective - it’ll be super interesting to see attrition rates.
Looking forward to voting for this one eventually.
A big argument against this proposal was “AAVE does it - so should we.” I hope this proposal gets revisited this year. The longer we wait the more the treasury drains.
Yes, we should stop aimless liquidity mining and turn to incentives when necessary.
For example, to stimulate future multi-chain liquidity growth, provide more compound interest for COMP holders, veToken model, and protocol income distribution.
At the same time, the community should pay corresponding rewards to Tyler, who has worked hard for this.
In order to be able to open the switch for distributing the profits of the protocol, the switch and ratio must be designed first.
It is easy to think of designing a veCOMP model for COMP, providing lock-up for 1 week to 4 years. (like Curve)
veCOMP gradually replaces the voting rights of COMP, which is more effective in preventing governance attacks.
veCOMP holders share the profits of the protocol, which is expected to exceed 30% APR in the early stage, and gradually decrease to 7% as veCOMP rises.
Obviously, the market circulation of COMP has decreased and the demand has increased.