I abstained from voting on Proposal 32, and voted FOR Proposal 59. This proposal isn’t perfect, but I believe that the advantages outweigh the disadvantages.
While there are strong arguments on both sides of the debate as to whether the November DAI price spike warrants compensation, it is the only event in Compound history that comes close to a “reserve worthy event”, and @kybx86 / @arr00 have done a remarkable job in the technical implementation of a solution (following the work by @getty and Chainlink to add preventative measures against the root cause).
This proposal demonstrates how reserves can be utilized by the community to protect users of the protocol, and I firmly believe that its passage sends a message to current and prospective users that the community is capable and willing to tackle adverse events (rare as they should be).
@monet-supply, @pyggie and others have pointed out that the allocation of reserves is inefficient; that it rewards farmers, and that the community should pro-actively select which users deserve compensation; I completely agree, but don’t think we should seek to punish (or not-compensate) users retroactively for activity we wish wasn’t occurring–instead, we should use this as a catalyst to calibrate COMP distribution and incentives going-forward, to disincentivize farming or non-useful activity.
Thank you @kybx86 for putting together this proposal and thank you to @arr00 for your contribution in the technical implementation of this proposal.
a16z voted yes on this proposal, for the following reasons:
The protocol behaved in a manner that significantly diverged from users’ expectations: The price provided by the Compound oracle was significantly divergent from the rest of the global market at the time. While the specific oracle design was publicly known, users generally expected the oracle (and the protocol more generally) to reflect general market prices for Dai, and this expectation was not met. Given that the losses stemmed from the protocol behaving in this manner, it’s reasonable that the community should step in and make users whole.
Previous concerns have been addressed: @kybx86 has taken the time to directly address the criticisms and concerns levied against the previous reimbursement proposal.
There has been successful testing for the technical implementation: If this proposal were to pass, the mechanism for executing the reimbursements have already been created + successfully tested on the kovan testnet. The code is a fork of the Uniswap merkle code used for their airdrop with a few small modifications, so there is a high degree of confidence that the implementation will work as expected.
Finally, while we supported this proposal, we do not think it should establish a broad precedent for backstopping any manner of user losses going forward. The losses in this case arose from the protocol behaving in a manner that materially defied user expectations (e.g. the oracle mispricing Dai by ~30%). While it’s reasonable for the community to cover losses of this sort, the community should generally oppose attempts to stretch this logic to cover other (less valid) claims for reimbursement going forward (e.g those arising from general market volatility, or minor deviations in oracle pricing, etc).
The most important thing in this situation is that the community agreed that the users are not guilty for that liquidations (there was talk of poor asset management, which is a ridiculous argument).
I agree with the members who claim that compensation is the minimum possible and that it barely covers the gas fee cost in an attempt to defend the position.
I think that users who were not involved in stablecoin farming should have been compensated more, because for example in my case (as with a large part of the damaged users) Ethereum positions were liquidated (at that time about 350$/ETH). It’s been almost a year and ETH has gone 10x, so the losses are huge.
It is understandable that the protocol simply cannot cover amounts based on market changes, but it must be understood that these losses would not have occurred if users had used another protocol (i.e. Aave). Losses were incurred solely through the use of the Compound protocol and I think all big COMP holders should have been more committed to this case in early stage after that event.
The positive thing is also that their comments about this case are now appearing - 9 months later, but better late than never.
Compound Governance is not disorganized, it just suffers from too many political games. Realistically, this situation would be resolved by the state administration within 9-10 months. @kybx86 all praise for the engagement in this case