Simple Summary
A proposal to completely remove MAI from Aave V3 Optimism and Arbitrum over centralisation risks.
Motivation
The objective of this proposal is to improve the quality of decentralised collateral on Aave by removing MAI, even from the isolation mode.
Below is the explanation and the data for the same:
As per this ARFC, MAI is a decentralized stablecoin minted by the Qidao Protocol. However, MAI seems to be controlled by a 2 of 4 multi-sig completely in the control of the team.
Specification
Looking at MAI contract, it can be minted by OnlyOwner here:
and, the owner is a multi-sig
which is controlled by these four addresses, under the 2 of 4 scheme
Out of these four addresses, only one of them seems to be publicly admitted DAO address.
Potential Loss
At combined debt ceiling of $3.1mil, under the hands of only 2 signers, this presents a big risk to the DAO.
If these two actors act not in the best interests of the DAO, Aave can lose upto $3.1mil.
Disclaimer
I am not associated to QiDAO or any other competing DAO. I am doing this to add value to the community.
Next Steps
- Discuss over possible flaws in the proposal.
- Discuss other ways to mitigate this risk.
- Develop a long term strategy to prevent this in the future.
Edits
- Changed from ARC to Temp Check
- Changed Remove to Freeze
2 Likes
Hello @IncentiveLord and thanks for publishing a proposal.
-
ARC process has been deprecated, is this proposal a temp check or a ARFC? following governance guidelines a TEMP CHECK for offboarding would be more fit.
-
Aave is a decentralized & permissionless protocol, an offboarding can’t happen overnight as the protocol can’t (and thxfully) touch user positions, are you advocating for a freeze that will remove the capabilities to have any additional deposits & borrows of this asset? followed by a offboarding plan making deposit unattractive via modified risk parameters and reserveFactor?
-
MAI is present since a long time on Polygon market, are you targeting MAI in general or only MAI on Arb & optimism?
-
The ACI is not in favor of this proposal as we’re supportive of stablecoin diversity and do not consider MAI as a particularly risky asset for the protocol. but if this proposal reach support on TEMP CHECK snapshot stage, we will assist with ARFC & AIP stage following our policy to support aave governance.
Disclaimer: I own QI & vQI representing a small portion of my portfolio.
Thanks @MarcZeller for the reply.
-
This is one of my first governance contributions. Will check now.
-
Yes, freeze followed by offboarding.
-
Though Polygon has a higher security assumption at 3 of 5 multi-sig, I will update the proposal to include Polygon as well.
Here’s the updated temp check.
Simple Summary[Updated Proposal]
A proposal to freeze MAI from Aave over centralisation risks.
Motivation
The objective of this proposal is to improve the quality of decentralised collateral on Aave by freezing MAI, even from the isolation mode.
Below is the explanation and the data for the same:
As per this ARFC, MAI is a decentralized stablecoin minted by the Qidao Protocol. However, MAI seems to be controlled by a 2 of 4 multi-sig on Optimism and Arbitrum, and 3 of 5 multising on Polygon. Completely in the control of the team.
Specification
Looking at MAI contract, it can be minted by OnlyOwner here:
and, the owner is a multi-sig
which is controlled by these four addresses, under the 2 of 4 scheme
Out of these four addresses, only one of them seems to be publicly admitted DAO address.
Potential Loss
At combined debt ceiling of $7.1mil on Optimism, Arbitrum, Avalanche and Polygon – under the hands of only 2 signers – this presents a big risk to the DAO.
If these two actors do not act in the best interests of the DAO, Aave can lose upto $7.1mil.
Disclaimer
I am not associated to QiDAO or any other competing DAO. I am doing this to add value to the community.
Next Steps
- Discuss over possible flaws in the proposal.
- Discuss other ways to mitigate this risk.
- Develop a long term strategy to prevent this in the future.
Edits
- Changed from ARC to Temp Check
- Changed Remove to Freeze
1 Like
Hey, thank you for the feedback.
While I don’t agree with your proposal, I think that your concerns are valid and merit action from QiDao. We should post a proposal on QiDao’s governance forum to address these issues.
From reading your proposal, the main perceived issue is the ownership of contracts by a multisig of a few internal members of QiDao’s community.
I’ve spoken to many active members of Aave’s governance community in order to arrive at a solution that best addresses the concerns you’ve mentioned. The general consensus is that QiDao should make certain changes, such as moving to a system similar to Aave Guardians. This of course needs to go through the proper governance process at QiDao. We should see a proposal posted on the QiDao forum sometime this week.
If you have any other feedback for QiDao, I would encourage you to visit the governance forum on QiDao’s Discord. Community members often propose ideas for governance approval and it’s a great way to improve the protocol collectively.
3 Likes
Following up on this, the proposal has been posted on QiDao’s governance forum: https://twitter.com/Benjamin918_/status/1660590473583493120?s=20
2 Likes
I begrudgingly agree. I was just looking at the implementation of QiStablecoin on Polygon, which is the same token listed as MAI on Aave’s polygon deployment. (0xa3fa99a148fa48d14ed51d610c367c61876997f1)
The contract has a burn function, shown below:
function burn(address account, uint256 amount) external onlyOwner() {
_burn(account, amount);
}
The owner of this contract is a Multisig (0x3FEACf904b152b1880bDE8BF04aC9Eb636fEE4d8), which is a 3/5 multisig. This should not be acceptable.
Summary
QiDao, owner of MAI Finance, is adopting QiDao Guardians, synonymous to Aave’s own Guardians. Per the Aave docs, the Guardians represent a 6/10 multisig and hold responsibilities allowing them to intervene during times of crisis in order to protect the protocol.
Both QiDao Guardians and Aave Guardians have similar roles:
- Guarding the respective protocol against possible governance takeovers by having a veto power on any deemed malicious AIP.
- Acting as failsafe emergency actors, if needed.
However, QiDao Guardians have additional duties:
- Managing the minting and burning of MAI to regulate debt ceilings for collateral vaults.
- Overseeing bridge diversification allocations for both MAI and QI.
Furthermore, they are expected to actively engage in multisig operations, staying alert, responsive, and prompt in action when necessary; and uphold the protocol’s best interest, promptly revealing any potential conflicts of interest.
Based on QIP197, although a Snapshot hasn’t been established yet, the following timeline is suggested:
- Activation of the QiDao Guardians: May 31st, 2023
- Addition of the Timelock to the multisig: June 15th, 2023
QiDao currently utilizes a 2/4 MultiSig on Optimism and Arbitrum, and a 3/5 MultiSig on Polygon. The proposed changes aim to address the existing situation where QiDao members alone could manipulate the supply and distribution of MAI. The new setup would require 4 of 6 members to pass emergency actions, including minting and burning MAI. The six delegates are composed of two QiDao team members and one QiDao community member:
- Marc Zeller from Aave Chan Initiative
- Hamzah Khan from Polygon Labs
- Weso from Beefy Finance
- 0xNacho from QiDao Community
- Benjamin.lens from QiDao Protocol
- Pablo the Penguin from QiDao Protocol
Market Risk Overview
From a market risk standpoint, Gauntlet has recommended the removal of MAI from the Stablecoins eMode category across all chains. An analysis of user positions reveals that few complex positions are collateralized by MAI (added to Optimism and Arbitrum ~2 weeks ago):
- On Optimism, only $105 of MAI is supplied, and $27 is borrowed.
- On Arbitrum, only $469 of MAI is supplied, and $345 is borrowed.
- Roughly half of Polygon users merely supply miMATIC to generate interest, while the other half borrows various volatile assets.
Hence, it does not seem that MAI is being utilized by users in the same capacity or with the same diversity as other stablecoins, such as USDC or DAI.
1 Like
MAI actually has very healthy usage on the chains where it has been onboarded. On Polygon, for example, utilization is at 42%. This is a higher utilization than most borrowable assets on Polygon.
It should be noted that both Arbitrum and Optimism markets only started recently, and without e-mode. that has contributed to low usage. As a result, it’s too premature to judge those markets.
Following up on the e-mode consideration, MAI has one of the lowest standard volatilities in the stablecoin market. As a result, I would say that it’s a great candidate for e-mode.
In Q1 of 2023, MAI saw lower volatility than most decentralized stablecoins (source).
