COMP Reward Adjustments v2

After many discussions with @getty, I’m pivoting my prior COMP Rewards Adjustments proposal.

The objective of the COMP rewards program was initially to distribute the token to our users. The sad truth is that an overwhelming amount of the COMP rewards are being farmed and instantly sold off, making the rewards program ineffective in achieving the initial goal. I believe there’s a clear need to re-evaluate how best to distribute the token to our community.

When a market first launches, we don’t see much activity until the market receives COMP rewards. There’s little incentive to deposit into it when there’s no borrowing demand (or rewards), and there’s no borrowing demand because there aren’t enough tokens to borrow to make the cost of the borrow transaction worth it.

A market needs to have great enough incentives for depositors to provide enough liquidity to make the cost of borrowing worth it.

The above statements I made in my prior proposal still hold. At the time of writing, I advocated for our users - who are also token holders - to continue receiving a share of the protocol. Decentralization of ownership matters. However, my perspective on the overall issue was slightly incorrect.

Since most COMP being distributed by the current rewards program is instantly sold off, existing users and token holders are at a great disservice. Their share of the protocol is being diluted for nothing other than farming COMP for profit. This COMP farming behavior is not the kind of activity that will bring value to the protocol, or for the existing users and token holders. Incentives need to be used to grow the protocol to the benefit of the protocol itself and its users and token holders.

Going forward, I now believe it’s best to end the current COMP rewards program and to start a new one with the sole purpose of kickstarting new markets. To our users who have faithfully held the tokens received from the rewards program: thank you for trusting and believing in us. Thank you for sticking with us through thick and thin. It’s not fair that your share of the protocol is being diluted for profit while you faithfully hold on to the token. We will fix this.

Here’s my new action plan.

Action plan

1. Cut existing rewards by 50%

On Friday, March 18, I will propose on-chain to cut the existing rewards by 50%. Rather than dropping rewards immediately to zero, we allow some time for adjustment.

2. Cut existing rewards to zero

About a month after, on Friday, April 15, I will propose on-chain to cut existing rewards to zero. This proposal will mark the end of COMP farming for profit at the expense of the protocol, our users, and our token holders.

3. R&D in interest rate models

We’ll let supply and demand take over, rather than COMP rewards essentially dictating market sizes and activity. Now here’s the caveat; we’ve deeply neglected our interest rate models. While our latest version of the jump rate model with its current parameters works okay for stablecoins, they are far from optimal for other markets. These sub-optimal interest rate models can make it tricky to maintain various markets in a way that’s equally favorable for both suppliers and borrowers while maintaining enough liquidity for new borrowers to enter the markets.

I already have a candidate in mind to help us with this who has experience in this area, having worked for one of Canada’s top banks. I will lead this initiative by reading and sharing the latest research papers on this topic (big thanks to my IEEE membership), defining a job scope with expectations and requirements, and coordinating with everyone involved.

4. Introduce the replacement rewards program: kickstart rewards

After upgrading our existing and future markets with optimal interest rate models, we’ll be in a great position to have an effective rewards program to kickstart new markets. Existing markets with low liquidity that haven’t received rewards before may also obtain these kickstart rewards.

This new rewards program addresses the issue mentioned at the top of this post - we need borrowing activity to make depositing activity worth it, and we need depositing activity to allow for borrowing activity to occur in the first place.

The plan is to incentivize $X market size at a rate of Y% annualized APR for three months to kickstart new markets. The X and Y variables are up for discussion. For example, we could incentivize $10M in deposits at an annualized rate of 8% APR.

Conclusion

This proposal aims to end the practice of COMP farming for profit which only hurts the protocol, our users, and our token holders. We need to incentivize behavior that benefits us rather than diluting our loyal COMP holders. This proposal paves a clear path for that, further strengthening the protocol.

12 Likes

Totally Agree!
COMP rewards should be used to stimulate the market when needed. Ends the inflation of COMP, and COMP holders are no longer diluted.

4 Likes

This makes so much sense! Incentivizing the right form of engagement with the protocol is fundamentally important.

4 Likes

In preparation for tomorrow’s proposal, here’s the passing proposal simulation: https://github.com/TylerEther/compound-protocol/blob/halve-comp-rewards/spec/sim/0012-halve-comp-rewards/hypothetical_proposal.sim

All COMP reward rates except for supplying COMP are halved.

Proposal

Comptroller._setCompSpeeds(
  [0x4Ddc2D193948926D02f9B1fE9e1daa0718270ED5, 0x39AA39c021dfbaE8faC545936693aC917d5E7563, 0x5d3a536E4D6DbD6114cc1Ead35777bAB948E3643, 0xccF4429DB6322D5C611ee964527D42E5d685DD6a, 0xf650C3d88D12dB855b8bf7D11Be6C55A4e07dCC9, 0x35a18000230da775cac24873d00ff85bccded550, 0xface851a4921ce59e912d19329929ce6da6eb0c7, 0xB3319f5D18Bc0D84dD1b4825Dcde5d5f7266d407, 0x6C8c6b02E7b2BE14d4fA6022Dfd6d75921D90E4E],
  [5375000000000000, 33500000000000000, 33500000000000000, 5375000000000000, 4825000000000000, 731250000000000, 731250000000000, 731250000000000, 731250000000000],
  [5375000000000000, 33500000000000000, 33500000000000000, 5375000000000000, 4825000000000000, 731250000000000, 731250000000000, 731250000000000, 731250000000000]
)

I need someone to verify that the proposal actions are correct.

Rewards for supplying COMP are to be halved as well.

Adjusted proposal actions:

Comptroller._setCompSpeeds(
  [0x4Ddc2D193948926D02f9B1fE9e1daa0718270ED5, 0x39AA39c021dfbaE8faC545936693aC917d5E7563, 0x5d3a536E4D6DbD6114cc1Ead35777bAB948E3643, 0xccF4429DB6322D5C611ee964527D42E5d685DD6a, 0xf650C3d88D12dB855b8bf7D11Be6C55A4e07dCC9, 0x35a18000230da775cac24873d00ff85bccded550, 0xface851a4921ce59e912d19329929ce6da6eb0c7, 0xB3319f5D18Bc0D84dD1b4825Dcde5d5f7266d407, 0x6C8c6b02E7b2BE14d4fA6022Dfd6d75921D90E4E, 0x70e36f6bf80a52b3b46b3af8e106cc0ed743e8e4],
  [5375000000000000, 33500000000000000, 33500000000000000, 5375000000000000, 4825000000000000, 731250000000000, 731250000000000, 731250000000000, 731250000000000, 2500000000000000],
  [5375000000000000, 33500000000000000, 33500000000000000, 5375000000000000, 4825000000000000, 731250000000000, 731250000000000, 731250000000000, 731250000000000, 0]
)

agree, nice proposal.

In support of this. Its vital for the protocol to incentivise the holders adequately

Is there a reason the COMP reward rates were also halved for supplying COMP ?
Supplying COMP to the protocol is one of the use cases of the token and giving rewards for this would incentivize people to hold on to the token.

It’s not a great way to reward holders when the function of depositing COMP this way is for lending and borrowing. Not only does rewarding holders this way have an undesirable effect on the market, but it also has additional risks.

In my opinion, there should be staking contracts with lock-up periods and fixed return rates.

1 Like

Sounds good to me, Tyler. How much less COMP is getting sold each day as a result of this adjustment? Does the protocol automatically sell rewards to pay down interest as it accrues?

1 Like

Yes!! This right here is what I’m talking about

Sounds good to me, Tyler. How much less COMP is getting sold each day as a result of this adjustment?

Too soon to say.

Does the protocol automatically sell rewards to pay down interest as it accrues?

Nope. Users must claim their rewards, then they can do as they wish with them.

1 Like

The above statements I made in my prior proposal still hold. At the time of writing, I advocated for our users - who are also token holders - to continue receiving a share of the protocol. Decentralization of ownership matters.

Nope, decentralization matters not. It’s cool story which easy to sell to crowd, but really nothing was done for decentralization. Actually every step so far was in opposite direction. Initial distribution directead most of the tokens to the hands of venture investors and team, with lesser portion to be distributed for a users over 4 years, but, oh well, that distribution was massively concentrated for big capital holders, and now even that going to stop, effectively stealing user portion of tokens and redirecting it into “trreasury” of protocol. Which in turn is largely controlled by those who got
initial distribution, and now they can vote what they going to do with tokens never actually belonged to them. Users on the other hand will have cool story that for their benefit distribution going to improve :slight_smile:

Since most COMP being distributed by the current rewards program is instantly sold off, existing users and token holders are at a great disservice. Their share of the protocol is being diluted for nothing other than farming COMP for profit.

That can’t be further from truth. Share of the protocol for token holder stays exactly same no matter what happens with COMP being farmed. 1 COMP is always 1/10M share of the protocol. Nothing is diluted as COMP is a fixed supply token. As for market
valuation it have literally nothing to do with share of the protocol. It’s just a speculation which matters not in the long run. The only thing protocol ACTUALLY own is reserves, and recursive farming is in fact benefitial to reserves, as it exchange comp tokens, which bear no value, for mostly stable coins in reserves which DO have a value.

However, it’s true that COMP distribution mechanism isn’t that great. It worked somewhat, and users were actually able to get a little bit of it, though scraps.

As for your plan i don’t have a goot feeling about it, if only for a reason that normally you don’t dismantle anything, which works, to build something which might work better in future (or might not). You build something FIRST, and then you transition from old to a new model.

As for COMP tokenomics, as it was first, it obviously wasn’t ideal. First of all initial distribution had not created treasury of protocol, opting to allocate all COMP supply. And treasury is what should be used for bootstrapping new markets.

Bootstrapping is a good idea by itself.

But treasury isn’t the only thing DeFi discovered through it’s existance. Of course, treasury in protocol tokens isn’t really a treasury just like it also isn’t in traditional finance. Aside of having protocol native tokens for some initiatives treasury should mostly hold actual reserve assets, like stable coins, eth, wbtc maybe several other tokens. Compound protocol actually fits well for having that, as it does naturally collect some reserves from it’s pools, which could be managed by protocol instead of just sitting in the pools for kind of nothing.

And another good discovery made by DeFi on the way, that it’s actually great to have protocol-owned liquidity. Not in kind of scammy way ohm forks did, but as a concept. For example, protocol can and should provide deep liquidity for COMP-ETH pairs
on important chains. Like it’s great to pay for that liquidity in protocol tokens like some dao do, but it’s even better when protocol actually owns it itself and instead of paying for liquidity, recieves trading fees, which slowly grow it’s treasury.

All that things are quite well-known by that point:

  1. Having treasury supply of native tokens (COMP) for incentivize programs

  2. Having deep treasury reserves, nominated in non-protocol tokens like stable-cons, eth, wbtc, possibly some others.

  3. Having deep own liquidity of nativetokens pairs on importand DEXes on important chains. At least on eth, maybe some others, depending on presence/planned presence of protocol.

It’s controversal if protocol tokens should be distributed for locking COMP tokens, there is not much value for protocol when someone just holds tokens, especially if at the same time they neither can use it as collateral, nor even vote in governance. It’s a waste in a long run, incentivising liquidity pairs is by far better spending (and even that is debateful,
as if protocol owns deep enough liquidity itself, there is no need to pay for external suppliers of liquidity, as they will come for trading fees anyway)

There might be some benefit in locking tokens for years and recieving weighted voting power for that, but we all know what it creates. CRV is good example. It’s just going to create another Convex. Which might be not bad.

The biggest question is while COMP model could be improved, should it? Do those VCs actually that useful for community to drag them with their bags, or it’s better to dump them and just use the code? Compound works fine as a protocol, price appreciation of COMP token isn’t really needed for anything and even if it plummet to zero pretty much only VC will be hurt, it’s irrelevant for small users, who hold pretty much nothing in vast majority.

That’s the thing: wide distribution of tokens is needed more for major bag holders, rather than for small guys. And yet they not even were able to push airdrop through governance.

It’s scientifically interesting to see what will happen when distribution will stop, but it’s not a big deal. Liquidity for COMP tokens provided by speculators, farming is done by speculators. There is nothing fundamental there, just one traders try to benefit at the expense of other traders. And VC mostly sit on their tokens and only marginally care about price at least mid-term. To have something more solid, Compound should be more of a DAO, which owns value, manages it, and is profitable in growing protocol-owned funds. Then shares in such enterprise will grow naturally.

Using reserves to provide COMP-ETH liquidity is a great idea.

Bancor is an good DEX for providing protocol owned liquidity. You can provide one-sided liquidity and you are the Bancor protocol protects you from impermanent loss.
See blog post below.

Passing proposal simulation for step 2: https://github.com/TylerEther/compound-protocol/blob/halve-comp-rewards/spec/sim/0013-cut-comp-rewards/hypothetical_proposal.sim

Proposal actions

Comptroller._setCompSpeeds(
  [0x4Ddc2D193948926D02f9B1fE9e1daa0718270ED5, 0x39AA39c021dfbaE8faC545936693aC917d5E7563, 0x5d3a536E4D6DbD6114cc1Ead35777bAB948E3643, 0xccF4429DB6322D5C611ee964527D42E5d685DD6a, 0xf650C3d88D12dB855b8bf7D11Be6C55A4e07dCC9, 0x35a18000230da775cac24873d00ff85bccded550, 0xface851a4921ce59e912d19329929ce6da6eb0c7, 0xB3319f5D18Bc0D84dD1b4825Dcde5d5f7266d407, 0x6C8c6b02E7b2BE14d4fA6022Dfd6d75921D90E4E, 0x70e36f6bf80a52b3b46b3af8e106cc0ed743e8e4],
  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
)
1 Like

Sorry for being late to the discussion, but I believe there is a lot to unpack here. Let me start by saying thank you to Tyler and the team for working on these proposals and trying to move the community forward. I generally try to stay to the side-lines on these types of conversations, but I think there are some aspects worth clarifying for this proposal.

First, Proposal 092 was executed less than a month ago. There simply has not been enough time for community members to analyze the effects of that proposal, and thus I believe it is too soon to make another significant change like this to the Protocol. For a proposal of this magnitude, I would expect significant analysis on the effects of that proposal, on the market health, etc. I see many thoughts on the genesis for making this change, but little on the analysis of the change, or a clear action plan on the future replacement.

To follow up on that, there is insufficient conversations from the community on this greater plan (discussed originally in this thread and now here). We should spend more effort bringing in a larger set of voices from the community (e.g. suppliers, borrowers, liquidators), and not mistake a limited response for tacit agreement. This forum post, and its precessor, have garnered less conversation than many less signficant proposals. I do not believe that the community has been given sufficient time to grasp the change and contribute to the conversation. As far as I can tell, there is no reason to push this change now, as opposed to after more time with more people contributing their own thoughts and ideas.

Finally, there are a lot of thoughts here that address fundamental questions for the existence of COMP and the distribution. From my personal view, I want to be very clear: decentralization of the Protocol has been, and should continue to be, the primary target of governance. I can understand the arguments raised by Tyler in the beginning of this conversation, but I do not see a concrete plan to achieve similar means of decentralization. Thus, until a future plan is hashed out and built, this proposal may actively work against continued decentralization. We should take more time to discuss and review these concepts, and I would suggest we table this proposal until after those conversations occur.

12 Likes

Thanks for sharing this perspective @hayesgm . I am also planning to vote against proposal 100. My rationale was the practical issue that removing the first 50% of rewards and removing the second 50% of rewards is a far more asymmetrical proposition than it sounds. A regular schedule of slowing COMP emissions seems more likely to support healthy liquidity in the markets than a sudden curtailment (even though 50% of rewards have already been cut).

I also think that the point about decentralization is an underappreciated one. Compound Labs’ deployment of COMP as a valueless governance token distributed pro rata to users is at the center of its identity as a non-security. I appreciate @TylerEther and others’ push to maximize the development and community-building work that the community can secure with its remaining COMP, but pro rata emissions to users is doing a different kind of work, call it political or legal, that still has value to the protocol. Retaining some pro rata distribution to users ensures that no one can claim control of the protocol isn’t being incrementally handed over to users (even if they farm and dump; we have no control over secondary markets).

4 Likes

Thank you @hayesgm and @allthecolors for sharing your views.

Geoffrey has highlighted a larger issue - it’s very challenging to get input from the various groups of stakeholders, and this has a big impact on decision-making and execution. @getty has been my biggest resource in planning this set of proposals as well as other proposals. Some Discord channels specifically for each stakeholder group would be a great first step at tackling this issue.

I agree that there should be more focus on decentralization of the protocol. So far, most tokens have gone to LPs with a few very large accounts with recursive positions harvesting and selling their rewards, which doesn’t exactly lead to a higher level of decentralization. We must not let the reservoir dwindle while we come up with new decentralized distribution plans.

We need to put the tokens in the hands of people who provide value to the protocol, those who’ll hold the token, and those who’ll participate in governance. Rewards for community members, for those who add to discussion, rewards for those providing help to new users, rewards for data analysis, rewards for innovative ideas, rewards for development, rewards for business development, etc.

The community has been neglected as far as rewards go, and we lost some good people after the temperature check for retroactive COMP distributions was shut down. I cannot stress enough how important a community is in the longetivity of any project. This is one area I think we should focus on.

As for the effect of this proposal will have on the markets, we do have Block Analitica’s impact analysis. I’d be surprised if market utilizations or interest rates change significantly.

2 Likes

I see no reason for pushing this either. I don’t believe even first part with 50% slash was great idea, as i believe that first new plan should be introduced and THEN old system might be deprecated.

But what is done is done. It expected much higher impact for TVL of stable coin markets, but so far they kind of holding relatively well. However as for second part i want to point one specific thing. After 50% slash there is now plenty of COMP released for presenting to community new and improved Kickstart rewards.

And until they surface i see no reason for slashing current rewards further. Actually since most distribution was concentrated in stable coin markets we very possibly could have same or better results with leaving distributions for all markets intact and slash 50% only stable coins markets.

Anyway i believe that phase two should be presentation of new rewards rather than removing remaining ones.

Also i don’t like potential collapse of TVL of stable coin markets, as these markes are pretty much sole actual contributors for protocol reserves with all other markets combined not contributing even 1/3 of reserves created by usdc and dai markets.