[ARFC] Migrate & Consolidate Polygon Treasury
Vote Result: YES
Rationale
Treasury migration on Polygon is an inevitable step in the transition process from Aave v2 to Aave v3, which is essentially safer for users. The assets being swapped here are all long-tail assets on Polygon Aave v2 and, once transferred to Aave v3, will be more productive. Therefore, we will vote YES in favour of this proposal.
OP Emission_Admin for Optimism v3 Liquidity Pool
Vote Result: YES
Rationale
This is the on-chain vote for this Snapshot that we voted in favour of, according to our rationale posted here. Therefore, we will vote YES in favour of this proposal.
Supply and Borrow Caps Update - LUSD - v3 Ethereum
Vote Result: YES
Rationale
LUSD has almost reached its borrow and supply caps, and the on-chain liquidity levels, current conservatism of the caps, and LUSD being a non-collateral asset indicate a low level of risk in increasing the caps by 2x. Therefore, we think this proposal is sensible and will vote YES in favour of it.
Safety Module Upgrade to v1.5
Vote Result: YES
Rationale
The technical upgrade of the Safety Module circumvents some complexity of the previous design and reinvents some critical mechanics to better protect both the interest of AAVE holders and the robustness of the protocol. In contrast to v1.0 of the Safety Module where ad-hoc development is needed before slashing, v1.5 is clearly more agile and robust in terms of slashing mechanics. The new cooldown mechanics and governance implications incur minor changes that make sense, and the upgrade also paves the way for the GHO project. Therefore, we will vote YES in favour of this proposal.
Activate Emode for cbETH Aave Ethereum V3
Vote Result: YES
Rationale
cbETH has held its peg firmly since launch, making it a good candidate for eMode. In terms of risk, the 93% LT of the ETH eMode category should accommodate the short-term volatility of cbETH, given the historical price range. Moreso, given the highly concentrated liquidity of cbETH, even if eMode is not enabled, most liquidation will occur around the 1% price slippage, and will not add oversized risk to the protocol. Maximising utilisation of LSD/ETH positions will increase capital efficiency significantly and promote AAVE as the leading protocol for LSD utilisation, increasing usage and revenues. Lastly, Gauntlet and Chaos Labs both support this proposal. Therefore, we will vote YES in favour of this proposal.
[TEMP CHECK] Aave Governance Process Improvements
Vote Result: YES
Rationale
The Aave governance process today is confusing and there are no set standards. There is a lot of confusion about the ARC and ARFC terms. This proposal streamlines the governance process and makes it much clearer. Therefore, we will vote YES in favour of this proposal.
Deploy Aave V3 to Metis Andromeda Mainnet
Vote Result: NO
Rationale
Having judged this proposal through a risk vs. reward perspective, i.e., is the risk of deploying on Metis higher than the potential reward, we have come to the conclusion of voting NO.
Before we dive into the reasons for voting NO, we recognise there are a few reasons to vote in favour. Firstly, implementing Aave v3 on Metis would help Aave expand to other chains and build a strong presence on the L2 ecosystem. Secondly, it is a fork of Optimism and hence development effort is not high. Lastly, 30% of the fees accrued from every transaction will be directed towards Aave in perpetuity, which is beneficial for Aave overall.
However, we believe that the reasons to vote against are stronger and counteract some of the purported benefits of deploying on Metis:
Security
Metis has not yet implemented its technical roadmap - data is not posted on-chain, and Sequencer rotation is not yet supported which reduces the security of the system. As pointed out by L2 Beat, without running the verifier dtl, independent observers cannot verify if the posted state root is correct if the data is not posted on-chain. Therefore, if data is missing from the MEMO lab’s storage, then there is no clear plan of action for independent users. When compared to Optimism, which Metis can be closely compared to given that it is forked from Optimism, users can prove that the state root is invalid and notify the community that fraud is happening, and the community can respond in 7 days using social means. In the case of Metis, all users can do is to complain that, according to them, data is missing in the MEMO lab’s storage without knowing if the posted state root is valid or not. This reduces overall security for the system, and for Aave.
TVL, Activity, and Traction
The TVL of Metis is low and market potential is not very high in general, with no discernible reasons to use Metis over other L2s for users. As per this comparison, Metis is slower and more expensive than Optimism, which calls into question its purported benefits and therefore impacts network adoption. If network adoption is low, then revenues that accrue to Aave are low as well.
Metis is the chain with the second-smallest TVL out of all the chains Aave has deployed to, with Harmony being the smallest chain. However, Harmony has faced issues over the last few months, and TVL has dropped significantly, from a peak of $1.82 billion in January 2022 to $7 million today. This drop is almost all of its TVL (reduction of 99.6%), and over the same period, is much higher than the drop in TVL of the other chains Aave is deployed on (Polygon is down 79%, Arbitrum is flat, Optimism’s TVL has increased, and Avalanche is down 86%). This serves to highlight the point that Aave should, as a flagship DeFi protocol, deploy on chains that are battle-tested, more secure, have higher liquidity, and higher potential for growth, with lower probability of bad debt on Aave.
Metis’ Current State of Development
As explained by @Peter88, one of the co-founders of Metis, there is only a high-level roadmap that can be provided by Metis at the moment. This indicates that Metis is still very early on in its development. Aave is a flagship DeFi protocol and should only deploy on chains with the highest level of security and user activity. At the moment, it seems as if Metis is a level below this and it may be more prudent to wait for Metis to further fulfil its roadmap.
Increase Supply Cap for cbETH Aave Ethereum V3
Vote Result: YES
Rationale
Previously, it was not clear how being incorporated in emode will affect the risk profile of cbETH, given that there was an ongoing proposal on enabling emode for cbETH. Due to this reason, in an earlier discussion on the forum, Gauntlet expressed their standpoint and suggested that the supply cap increase should be postponed. Now the emode for cbETH has been enabled, and the utilisation is almost at its maximum. Therefore, we support the proposal to increase the supply cap to 30k for cbETH.
Supply and Borrow Cap Updates V3 Polygon and Arbitrum
Vote Result: YES
Rationale
Given that wstETH is currently at 100% utilisation of its supply cap on Aave V3 Arbitrum, increasing the borrow and supply caps by 2x under the direct-to-AIP framework is sensible.
The core assumption made by Chaos Labs is that it has simulated a 15% drop in the price of wstETH to evaluate the amount that would be liquidated, and assessed the on-chain liquidity that would allow for profitable liquidations under the 2% Liquidation Penalty configured for the emode. The 15% drop in price is for stETH only when compared to the price of wETH on Arbitrum, i.e., a stETH depeg. For stETH to fall 15% at once (compared to all other tokens, including ETH) is rather extreme. At current liquidity levels, the recommendation holds for a 30% drop. If both ETH and stETH drop by 15%, the impact on liquidation is lower since there isn’t a depeg and a mismatch, and as the value of the collateral vs. the borrowed amount stays constant, there are no or lower liquidations. Therefore, the 15% depeg of only stETH is a fair assumption to make.
Therefore, we will vote YES in favour of this proposal.
Aave V3 Interest Rate Curve Changes
Vote Result: YES
Rationale
This is the on-chain vote for this Snapshot vote that we voted YES for, according to our rationale posted here. Gauntlet has reassessed its rationales for MAI, FRAX, agEUR, EURS, WETH, and GHST, which still make sense, even taking into account the ongoing efforts to disable agEUR borrows and freeze GHST. Lastly, the delay USDT experienced in most markets in returning to lower levels of utilisation following its uptick in utilisation during the USDC de-peg event further supports Gauntlet’s recommendation to move USDT to the more conservative interest rate curve.
Given the above, we will vote YES in favour of this proposal.
AAVE on Harmony Recovery Plan
Vote Result: NO
Rationale
The opinion on the forum is largely divided. From our perspective, the only argument that supports the proposal is that the recovery will save Aave’s reputation and allow it to continue to work on Harmony. Much of the posts on the forum that support this argument and thus the recovery plan, especially the aggressive ones, are from ONE holders, who bear losses due to the exploit on Harmony. We do not view them as objective opinions on the issue. We are in agreement with @sakulstra’s comment above, and there is not enough depth in this proposal regarding key implementation details such as the impact on Aave’s treasury. Therefore, we will vote NO on this proposal.
Risk Parameter Updates for Aave V3 ETH Market (2023-03-18)
Vote Result: YES
Rationale
Gauntlet suggests that wstETH and rETH borrow caps are above the 75% threshold. According to their borrow and supply cap methodology, Gauntlet recommends the changes mentioned above in terms of the borrow and supply cap, which we believe is reasonable. Therefore, we will vote YES in favour of this proposal.
Freeze DPI on V2 Ethereum
Vote Result: YES (vote did not go through)
Rationale
Given usage and liquidity conditions of DPI, migrating the asset from v2 to v3 allows for more granular control over risk parameters, which is important in light of declining market cap and liquidity for the asset. Conservative caps allow for organic borrowing activity while removing all profitable attack vectors, mitigating risk for Aave. Therefore, we will vote YES in favour of this proposal.
[ARC] Gauntlet Risk Parameter Recommendations for V3 AVAX (2023-03-20)
Vote Result: YES
Rationale
There is a need to proactively de-risk the protocol considering ever-changing liquidity conditions, especially as the result of an event like the USDC de-peg, and the proactivity by both Chaos Labs and Gauntlet should be commended. Given that both risk providers are in agreement over the changes, we will vote YES in favour of this proposal.
[TEMP CHECK] Aave Chan Initiative 6-Month Budget Proposal
Vote Result: YES
Rationale
The ACI’s work on Aave to date has been impressive, thorough, and extremely valuable to the ecosystem. ACI has led on a lot of proposals and has helped drive a lot of the operational efficiency present in Aave today. Compensating the effort is only fair, and if this is not done, then it acts as a disincentive for other contributors to work on Aave. The $250k budget may seem excessive, but the stated aims are to also grow the team, for which this budget seems fair. We agree with what @OriN and @dtalwar have said - Marc’s active participation in Aave governance has been great and very helpful.
[ARC] Isolation Mode Recommendations for Aave V3 Avalanche, Polygon, and Arbitrum
Vote Result: AGGRESSIVE
Rationale
We support the aggressive recommendation since the underlying assumption of the debt ceiling at 25% of the circulating market cap is still safe while allowing for enough growth, and the aggressive recommendations would still de-risk the protocol.
There is a question around EURS on Polygon, for which borrowing would cease against the asset. There is one whale account that supplies 62% of EURS on Polygon, for which there would not be enough liquidity on Polygon to liquidate the user in a worst-case scenario. According to Gauntlet’s rationale, which we agree with, this account is supplying $1.7M EURS and borrowing $700k USDC, which suggests a relatively low risk from this account. In the event of liquidations, it’s likely that liquidators would redeem the liquidated EURS for EUR through Stasis rather than swapping EURS for other assets through DEXs. As an initial step, reducing the debt ceiling to limit further borrowing against EURS makes sense.
Gauntlet has also clarified that for USDT, the available liquidity of assets that can be borrowed is the limiting factor for both conservative and aggressive recommendations, rather than the circulating market cap. Therefore, the conservative and aggressive recommendations are the same.
Therefore, we will vote YES for this proposal.
[ARFC - Temp Check] “MVP” V3 Deployment on ZkEVM Mainnet
Vote Result: YES
Rationale
Polygon’s zkEVM is one of the flagship zkEVM L2 chains and it is sensible and important for Aave to deploy on zkEVM, especially given market trends. Moreso, Aave v3 has been successfully deployed on the ZkEVM testnet for approximately 6 months (since October 2022). Since ZkEVM is EVM-equivalent, deployment would require minimal development work, and risk is mitigated by onboarding only strictly limited assets (only wETH, wMATIC, and USDC as collateral assets and USDC as a borrowable asset). One point of concern is that Gauntlet and Chaos Labs have not vetted this proposal and the risk parameters. However, given that this is a temp check, there is enough time for approval by Gauntlet and Chaos Labs on the risk parameters, and their validation is needed before the on-chain vote.
Therefore, we will vote YES in favour of this proposal.
[ARC] E-Mode Parameter Changes for V3 Avalanche, Optimism, Polygon, and Arbitrum
Vote Result: Option 1A
Rationale
We recommend Option 1A, since we agree with Wintermute that for users to use e-mode, it needs to offer more capital efficient parameters than regular borrowing, which is achieved under Option 1A while still offering various trading strategies. Option 1A is simply a more conservative version of Option 1, which we think is fair at this point given market conditions and the wider macroeconomy. Another positive is that Chaos Labs is supportive of Option 1A and has laid out a plan of implementing 3 consecutive 3.5% reductions to mitigate the impact on liquidations and user experience. Gauntlet has also implemented Chaos’ suggestion of reducing LTV to 80% which will further protect retail users from liquidations.
Regarding the other options, Option 2 limits e-mode to mostly ‘USDC’ based stablecoins, while Option 3 is just a more conservative version of Option 2. This reduces the benefits of e-mode significantly, one of which is arbitrage between stablecoins. Lastly, Option 4 does not make sense since stablecoin e-mode is still valuable.
Therefore, we will vote YES in favour of Option 1A.
[ARFC] - Chaos Labs Risk Parameter Updates - Aave V3 Optimism - 2023.03.22
Vote Result: YES
Rationale
Increasing LTVs and LTs for wBTC and DAI on v3 Optimism leads to a projected borrow increase of ~$155,000, with no increase in VaR and Extreme VaR, making this proactive move beneficial for Aave. Reducing the liquidation bonus for wBTC to be more in line with assets of a similar level of risk and liquidity (USDC and ETH) makes sense. A potential point of concern is the risk in ~40% of the on-chain supply of WBTC on Optimism being currently held on Aave. However, Chaos Labs’ analysis indicates that there is adequate on-chain liquidity to support significant liquidations.
Therefore, we will vote YES in favour of this proposal.
Risk Parameter Updates for Aave V3 ETH and V3 Arbitrum (2023-03-27)
Vote Result: YES
Rationale
AAVE, LINK, and WETH supply caps are above the 75% threshold on v3 Arbitrum, and Gauntlet proposes to increase WETH’s Supply Cap from 35,280 to 45,000 while leaving the supply caps for AAVE and LINK the same due to on-chain liquidity conditions. Moreover, rETH supply cap is above the 75% threshold on v3 Ethereum. They propose increasing the rETH supply cap from 10k to 20k. Moreover, they recommend decreasing the supply cap for CRV from 62.5M to 51M; current supply is only 108k, so 51M should be more than enough.
We think the above is sensible and will therefore vote YES in favour of this proposal.