Compound Finance COMP rewards

#2: Compound is a lending protocol… revenue is generated from borrowers, so if anything, borrowing should be rewarded. Rewarding supply just brings in dead TVL.

Currently as far as I’m aware, the only benefits of COMP are minimal voting rights and dumping. So adding value to holding COMP/locking up COMP, from the minting and/or from converting some of the part of the revenue(now going 100% to reserves?) might be useful.

If you really want to make this a community project, maybe use some of the treasury to hire a dedicated team to work on the protocol? Just because it’s decentralized doesn’t mean there should be no leadership…

Anyway, just my 0.02 COMP :stuck_out_tongue:

I believe does the following: you lock for some years and you get on the pool an high interest rates on what you locked. I think it would be good to lock away some tokens while bringing more attention to the protocol and distribute some of the treasury, another interesting thing would be to increase borrow APY on comp distribution and outright remove supply distribution.

Great writeup @getty. My thoughts on each of the ideas:

  1. I’m not in favor of paying people to vote. In an ideal world, every tokenholder is up-to-date on protocol affairs, reads all of the proposals thoroughly, and votes according to her independent view. Unfortunately, we don’t live in that world: it takes a lot of time to get up to speed on governance, and as a practical matter, few tokenholders have the time to vote with an informed view. (I’m not the first to say this; there are always tradeoffs to various governance mechanisms). Paying people to vote may result in incentivizing the wrong behavior: we want people to get informed first and vote second. Paying people to vote for voting’s sake is unlikely to add any tangible benefit to the protocol.

  2. I’m very much in favor of separating supply and borrow subsidies. As a general matter, I think rewards/subsidies/liquidity mining programs should be thought of as growth programs (rather than token distribution mechanisms). As such, they should be managed scientifically. That is, if we’re designing a new rewards program, we should have a hypothesis of what will happen, run the experiment, and compare the test results against the hypothesis. We should do this thousands of times until we are able to understand how users interact with the protocol. Once we have a deep understanding, we can use the data to grow usage, reduce churn, and allow the protocol to use its COMP efficiently.

  3. @getty and I had a few brief conversations about how to turbocharge the Compound Grants Program. One way to do it that is by growing its budget, of course. In my mind, that’s necessary but not sufficient: having the money to fund grants is good, but what’s absolutely necessary is having a well-staffed team of full-time individuals who is able to quickly, efficiently, and dutifully provide grants. One learning we’ve had from the Compound Grants Program experiment is it’s simply not enough to have a part-time grants manager (that’s yours truly). We will share more thoughts on this topic in a separate post in the future.

  4. Presumably, the idea behind subsidizing fees is to grow usage of Compound (particularly with retail users). If that’s in fact the motivation behind the subsidy, then I’m very much in favor. That said, similar to my thoughts in point two, I believe the subsidy should be managed as a scientific experiment. For example, we may have a hypothesis that covering 80% of the fees will grow Compound’s user base by Y, and this user base will churn by Z% over 6 months. To test the hypothesis, we can run a test and compare it against the hypothesis. More importantly, we can run this test hundreds of times until we find the optimal subsidy percentage.

  5. See my thoughts on point one. We want informed voters, not profit-maximizing voters.

I’m glad we’re having this discussion. It’s an important one!


First off, I want to thank everyone for contributing to the topic. This kind of participation is what is so important to developing something the entire community is excited about.

Recap of the ideas:

  1. Reward voters for casting a vote.
  2. Separate supply and borrow side COMP rewards.
  3. Supersizing the grants program.
  4. Transaction fee reimbursement program.
  5. Extra COMP rewards to protocol users who also vote.
  6. Airdrop
  7. Burn

Let’s toss out incentivizing voting (1 & 5), doing an airdrop, and burn. That leaves us with 2, 3, and 4.

  1. Separate supply and borrow side COMP rewards.

@TylerEther is working on implementing separate supply and borrow side COMP rewards. See his post to learn more. In the interest of time, the community should review Tyler’s work and begin the conversation of what speeds to use.

  1. Supersizing the grants program.

Regarding supersizing the grants program. @sukernik is going to work on a proposal for the next phase of the grants program. That should come later this month or in early September.

  1. Transaction fee reimbursement program.

I’ll dig up some of the tools I made for this a while ago and put together a database of Compound transactions for July. I’ll create a new forum post and Google Sheet with that info once I have it, and then we can begin a conversation about what to reimburse.

1 Like

You 4got 6!!! :upside_down_face:

1 Like

If we reduce COMP liquidity mining rewards notably (as has been discussed), but want to keep the rate of COMP distribution high, we have to come up with ways to distribute large amounts of COMP.

It seems like 1, 4, and 5 wouldn’t move the needle much. 3 (beefing up grants program), is the only larger-scale COMP distribution method here, outside of continued liquidity mining.

Growth / biz dev strategies seem underexplored.

  • A referral program to ‘mine’ new users
  • Some institutional business development schemes would be great to fund, maybe even an institutional liquidity mining program or something (to get more potential power users playing with the protocol)
  • token swaps with other protocols and/or distributing COMP to protocols in exchange for parking treasury funds in COMP

One other idea (that admittedly is a bit half-baked right now)…

“Governance Mining” — Governance participation is perhaps the most valuable activity that we want to reward right now. Right now, we’re funding 1 FTE contributor and a couple handfulls of part-time contributors who have formally applied for grants — however, there is a ton of really valuable ideation, feedback, small governance tasks, etc. that is crowdsourced from the community through the forum, in the governance chat, community calls, and other mediums. This activity overwhelmingly goes unrewarded. We could boost governance participation by “mining” valuable governance activity. For example, we could distribute COMP to people who do things like: identify problems in COMP, propose solutions, offer thoughtful commentary/analysis on proposed solutions, etc.

The positives: this would speed up problem identification and solution ideation, it would get more people meaningfully involved in governance, and it would distribute COMP to a group strongly aligned with the protocol.

The negatives: might dilute the signal of discussion (gov spam), it’s really hard to come up with a quality & scalable heuristic for assessing gov participation (requires a lot of human input)

I think there’s something here. Maybe this just means the grant program (or some group managing the governance mining rewards) gives small monthly grants/rewards to those who it deems have contributed valuable activity in governance, and even if it’s not a perfect distribution method, it’s at least a step in the right direction.

SourceCred is the only model I’m aware of for something like this, but curious if anyone knows of others.


I like the direction of @JacobPPhillips’ thinking around growth / biz dev strategies as untapped potential. That said, I also share the concerns raised about potentially incentivizing un/counterproductive behavior through governance mining. Compound is an innovator in on-chain governance, but I’m definitely not convinced that we’ve found the secret sauce that will make incentivized governance a seamlessly productive use of COMP.

Zooming out, there are currently three major ways for people and organizations to procure voting power in on-chain protocol governance:

(A) Earn COMP directly from the protocol
(i) Earn as a founding team member or early investor
(ii) Earn through liquidity mining
(iii) Earn through governance for an implemented protocol development or through the Compound Grants program

(B) Acquire COMP tokens through secondary markets
(i) Buy COMP on an exchange or via p2p trade
(ii) Borrow COMP through a lending service (CeFi or DeFi)
(iii) Receive COMP from someone as a gift (I’d lump Coinbase Earn in here)
(iv) Steal COMP (obviously this is terrible and not condoned, but to leave it out would be to ignore the very real problem we have with scammers, e.g. on our own Discord server)

(C) Others can delegate their COMP to you

In this thread we are exploring tuning and broadening category (A) in ways that benefit the safety, health, accessibility, and overall mission of the protocol. I’m framing it this way because I think it’s worth emphasizing that this discussion is focused on strategies for apportioning future governance rights that are currently under the full purview of governance.

I support all of the tweaks to (A)(ii) and (A)(iii) that have been discussed (e.g. compSpeed split, supersizing the grants program).

Regarding category (A)(i): the folks listed here were absolutely critical to the funding, development, and bootstrapping of the protocol, so it is fitting that they are entrusted with a significant role in protocol governance. Also critical to the bootstrapping stage was the community of early (pre-COMP) users, would-be stakeholders who were in most cases diluted to oblivion since COMP launch day by the deluge of capital that came to yield farm professionally. Early protocol users should hold a meaningful fraction of collective governance power: perhaps not the full 5% level initially floated by a16z, but at least 1% of collective governance power. What is the rationale for holding this number at 0%?

I also want to signal-boost @rleshner’s suggestion of rewarding on-chain activity in an asset proportionally to its effect on increasing protocol reserves in the relevant asset, with greater rewards for boosting reserves of assets in which the protocol is least well-capitalized. I can’t think of anything more clearly aligned with the goal of promoting protocol health and safety. And for those who envision governance eventually rallying around a periodic disbursement of reserves deemed “excess” to COMP holders, well, this idea could put the feasibility of such a concept on a faster track (fwiw, I would not support such a proposal for quite some time to come, if ever).

I would love to see this “COMP for contributing to protocol health via effects on reserves” idea explored further and ultimately formalized (ideally by someone with more of a risk management background than me!) so we could think about adding it to the RFPs.


Another model to crowd-source incentives is Coordinape. (developed in the Yearn community)

I mostly agree that burn mechanism isnt best option in this moment but I would not reject it for the reason that a well-implemented burn function can make the tokenomics system more flexible and sustainable through longer period. As for airdrop I understand the views of users with a large amount of COMP tokens, the thing is quite subjective.
Compound was the first to implement the farming function (relatively) much earlier than the first real airdrop (Uniswap) so it is not necessary to specifically filter early users because realistically no one knew that the DeFi sector would move in this direction. How can someone play a system that doesn’t exist?

High gas fees and pointless impact given the current governance structure was discentivized small voters to participate in governance. If you analyze voters structure on past proposals, you will get it. Users who voted with a negligible amount of COMP tokens either “donated” protocol support or expected a reward for the move. If there is a third reason please state it, because I am blind.

I can’t understand this part. What is the purpose of the lending and borrowing protocol? How does Compound make revenue? Given the very low lending rate, why do USERS put funds into the protocol? If we omit the borrowing function, will the Compound business model be sustainable?

Supersizing in terms of a horizontal hierarchy is a good idea

Great idea

COMP distribution goes in pockets of big holders. Who have and who have not incetives to participate in governance. Maybe we could reconcile this issue with the airdrop issue - users will vote if they have voting power. We can compare two types of users:

  1. Long-term users (2-3 years) with less capital without farming attention
  2. Two weeks active stablecoin farmers with huge amount of capital
    Guess who has a bigger voting power?
    Who is “more important” for protocol growth?

@getty I appreciate your work in the community, I just don’t understand what the point of some opinions and ideas above (1,2,5 + intro). Please accept my criticism as something constructive, as I maybe have the wrong view

1 Like