The issues presented by the UAC are of extreme importance, and we appreciate the effort of all members involved in bringing this to light and discussing the issue. After reading the original post and comments, we’d like to bring some more context to the discussion, as well as offer our take regarding possible solutions.
blockful
We’ve been to some calls, but it’s our first time in this forum, so for those who don’t know: Blockful is a governance security company, we’ve received a grant from the UF last year to research and build a dashboard to keep track of governance security. We believe that some of the information we are keeping track - publicly available at anticapture.com/uni - can help better evaluate the next steps. Our team counts with stewards, contributors, and delegates in other DAOs.
Approach
As we see, in the original post and other comments so far, there are only 3 presented approaches that at this time would help prevent the decreasing quorum from offering real risks of creating governance paralysis, so we will focus on those.
1. Treasury Delegation
Regarding all approaches that get more tokens from the treasury and delegate them to already active delegates selected by the governance itself.
tl;dr we are against it because benefits collusion to capture, increase and/or perpetuate power and payments in the DAO
- It solves the problem temporarily, giving the DAO room to breathe and solve engagement over time. However, it creates other problems that will likely persist even after it becomes inefficient as a solution.
- We are against it in its current form, as we evaluate that as the least favorable option, given that the current concentration of active voting power coming from treasury delegation strategy is already bigger than we’d see as ideal when compared to current voting activity.
- Although it has been essential for the good functioning of the DAO so far, the lack of expiration periods, the power to self-enforce continuation, and the DAO’s dependence on the program already shows it shouldn’t be increased.
- The original post states that after the quorum is reached, there is a decline in new votes cast on a proposal. It should be considered whether this type of mechanism could be increasing the quorum problem by creating a comfortable dependence on it, while other voters have less vote impact and less pressure to show, and get more distant over time. As @404DAO put it, we also question if this “may unintentionally disincentivize UNI holders from delegating or further engaging in governance”
- We would consider supporting such a mechanism if a large number of votes are distributed among a large number of delegates, instead of increasing “core voter concentration”. Although it will dilute organic delegates more than the original option, it will also help dilute the artificial delegations currently in place.
- We have a deep respect for the (very) hard work being done by delegate teams in Uniswap and other DAOs in the ecosystem, and we hope all parts involved in this system can take this criticism as a healthy security measure to be considered. We can’t support increasing the collusion and capture potential in the DAO based on trust that companies, with many contributors, possibly rotating attributions and with its own internal risks of capture will always do the right thing for the protocol.
2. Quorum reduction
Regarding the reduction of the necessary amount of votes cast in favor of a proposal for its results to be valid.
tl;dr we find it not ideal, but not as terrible as it initially sounds
-
It helps solve the problem by making the quorum more compatible with current voting power; it also increases the proportional voice of smaller delegates in comparison to what’s needed to execute something.
-
It’s important to understand that quorum is a defense against silent proposals and attacks in which the attacker is able to nullify the defenses of the DAO.
a. An attack that could compromise the governance interface can be an issue with low quorums for example
b. An attack during holidays is the most classical example of a problem in low quorums. -
Although partially contrarian to the approach, we don’t see a big risk increase in Uniswap DAO reducing its quorum from 40M to 24M UNI because Uniswap has the contract implementation defenses and the community necessary to defend from its issues.
a. the limit of active proposals from an address and the permissionless proposal cancellation in cases wehere the proposer no longer holds VP above the threshold are examples of those defenses
b. Still, it’d be preferable not to reduce quorum if not needed. More defense = better -
Regarding @kpk comment about quorum vs treasury size, the quorum already sits way below the treasury value, and so does the delegated supply. We believe Uniswap has not been attacked/arbitraged out of its treasury so far because of its treasury composition(99% UNI), not because of the cost of the VP majority or quorum.
-
From proposal to execution, there are ~9 days of delays, and we can assume that FUD would quickly melt those billions of UNI in the event of an attack.
a. There’s no certainty on this, and its better safe than sorry, but we don’t see it as worth holding this as the main worry for this change, considering that the quorum value has already fluctuated between $195M and $790M in the last year.
b. The DeFi calculations of how far the token price could melt in such short notice have too many variables and we’d welcome a well based challenge or support to this thesis. -
The real questions to consider regarding quorum are:
a. Is there anyone that would profit by “risking/burning capital at the quorum’s dollar value” to break Uniswap token or take control of its governance permissions over the protocol?
b. Can delegates coordinate to defend against an attack of 24M UNI at all times? -
If we can defend against 40M VP, then we can defend against 24M VP. So either we are already at risk, or the risk doesn’t change much.
a. If we can’t defend either, then estimating if someone who won’t be willing to lose 40M UNI worth to attack would be willing to lose 24M to do so would be the best questioning. And we find it almost impossibly complex to estimate such things. -
The quorum also presents a previously undiscussed problem of making the current treasury delegations very close to enough for dominance. Assuming the active voting power really follows a patern of decrease but only until its barely above quorum, this could mean a very hard to revert dominance by delegate teams over the next couple of years.
3. Delegation Incentives
Regarding initiatives of token emissions as incentives for holders to have their tokens delegated
tl;dr it’s be the best option presented so far, but will need refinement
@Curia and @1a35e1 made very good additions to this discussion. It’d be very interesting to explore alternatives together as we see the approaches presented being dissimilar in some senses, as lighthouse doesn’t explicitly put it as financial incentives and Curia brings reputation to the table, which we are deeply favorable towards if well executed.
- This is our favorite approach here, and we consider emissions are justified in the short term, with results to be evaluated for continuation. Assuming delegation stickiness, it should be worth doing with or without continuation after the reevaluation.
- The issue with Uniswap DAO’s quorum is not a lack of delegation as in other DAOs, but a lack of active delegates with VP. Currently, only 3 of the top 20 delegates are active voters.
For that reason:
a. We believe there should be an incentive for active delegations, independent of alignment or quality of the delegate.
b. It’s not the ideal governance’s role to enforce a distinction between “good” and “bad” delegates for holders to choose from to get yield from vetted vaults. It severely hurts holder’s choice and power.
c. If you want people to participate, holders should have power over governance, not be limited by it in their choice of who to delegate to i they want yields. - Incentives for delegation are our favorite option. But we disagree with the IDVs approach presented because of the pre-selection of vaults. It can and likely will still be good if done this way, but we see a creation of incentives towards active delegates as a better option to solve the issue in the long term, as it’d favor a body of delegates that’s more resilient to 1 or 2 big of them becoming inactive in the future.
- We would prefer to see an option where all delegators get yield:
a. proportional to their voting power delegated
b. when the voting period is finished
c. as long as their delegate has voted. - There is a problem of an incentive for proposal spamming to farm more yield here, which can be prevented with a time-based mechanism to calculate the total value of the distribution to be made in each vote. We’d love to discuss this further and refine it.
The other options presented are (very) interesting mechanisms to experiment with and can bring positive results. Still, given the state of quorum and activity, we don’t believe they would be beneficial at this point or are prioritary.
Particularly for Optimistic Governance, given that we are struggling to maintain quorums above the necessary level as they continue to decline, we don’t want to create an option for any proposal to pass instead of fail in case the quorum is not met, even with a smaller quorum for vetoing.
In summary:
- Treasury Delegation will lead to centralization and open doors for “internal” capture
- Quorum reduction is possible given Uniswap’s resilience, but it is not preferable
- Delegation incentives are the best approach, but need to focus on rewarding delegation to active voters, not delegation to a pool of voters chosen by the inside group