Tl;dr:
- The Safety Module (SafM) is being re-evaluated as the expense to fund insurance for depositors on Aave is both expensive and ineffective.
- A primary concern about changing the assets in the SafM is reduced utility for AAVE, which currently earns nearly 10% in the SafM.
- We believe that attempting to balance AAVE utility and depositor safety in a single mechanism is not a viable long-term solution for the DAO.
- We propose transitioning the current SafM into a Staking Module (StkM) where AAVE holders can stake AAVE or AAVE/wstETH BPTs in exchange for voting rights, and interest rate discounts. Using some of Llama’s recommendations, the Aave DAO can also allocate some treasury assets to earn veBAL voting power and allocate rewards to AAVE pools.
- The new SafM should prioritize liquid assets. We recommend further research on stablecoin and ETH-related BPTs, as they could provide the best balance between sustainable funding (from BAL rewards) and liquid assets to protect depositors in a time of need.
Recent Safety Module Discussions:
Technical Update: BGD shared the latest updates on the SafM recently, with the new code passing audits and going to vote for on-chain implementation soon.
Design Discussions: Llama recently posted a 6-part update temp check, following earlier discussions in February here and here. Their proposal includes new ideas around adding assets to the SafM that are more easily funded by the DAO using metagovernance, namely Aave’s partnership with Balancer and Balancer’s veBAL rewards system (including leveraging boosted yield strategies through Aura).
We applaud these designs for their alignment with the DeFi ethos, their composability, and their intentions to improve the sustainability of the SafM.
Our primary concern with these designs is they’re likely ineffectiveness as a safety valve for depositors should the protocol accrue bad debt. Research from Gauntlet and Xenophon Labs demonstrates the uncertainty of liquidating assets quickly at a fair market value. The performance of EUL during Euler’s recent incident demonstrates how a protocol token might trade during times of crisis. These examples demonstrate that if the SafM’s underlying assets lose value due to illiquidity in a crisis, the cost analysis of depositor protection is likely underestimating the true cost of SafM protection during such times.
Safety Module Recommendations:
We have consistently pushed for the most liquid assets to be included in the SafM instead of AAVE or AAVE BPTs.
The most important function of the safety module is depositor insurance. Using the native token is a bad design choice for that purpose. Stablecoins and ETH-based assets are significantly more liquid and better suited for insurance assets.
In terms of cost, we would encourage the DAO to ask existing service providers or propose grants to research the efficacy of more liquid-asset BPTs (like stablecoins and ETH-assets), so the DAO can achieve both effective protection for depositors and sustainable expenses for the DAO.
AAVE Utility and Governance:
The DAO can create a Staking Module (stkM) for AAVE that provides AAVE or other permissioned asset stakers (like AAVE/wETH BPTs) to receive a Medallion Token representing their governance power. We would recommend having a BPT option to ensure the liquidity of the native token even if adoption is high.
If the value of one AAVE token today is determined by the value of its governance power or claim to DAO assets, that power is 1/total AAVE supply. With a staking module, AAVE’s value is 1/total staked supply. At worst, the value will be the same between the two, but more likely the staking ratio will be less than 1. Note that there are no unsustainable games here around long-term incentives from inflation or lockups. Simply a separation between active participants and passive holders. Although we love the Variant idea of crypto as “The Ownership Economy” because users own their data (great for the web3 consumer sector), we think DeFi is better thought of as “The Participant’s Economy” because active community members and users are first class to passive owners and rent-seekers.
Much of the work @Llama has done can also be applied to the StkM. The DAO can allocate some treasury resources to a metagovernance committee, suggested in part V, which is responsible for voting and reporting the results. This can help drive sustainable value and revenue to AAVE stakers, while supporting its partners.
stkAAVE holders will also have extra utility like discounts on borrowing GHO or possibly other assets. Alignment between governors of the protocol and the success of the protocol can be powerful but also needs to be checked as a risk; Aave should not be useful only to active stakers.
A final risk might be the increased governance participation cost for smaller AAVE holders. We recommend an oracle solution that allows AAVE holders on roll-ups to stake and get equal governance rights. @BGDLabs is currently working on the new V3 governance module which enables cross-chain governance. Synthetix and others have successfully implemented similar solutions so this could be a viable option for Aave as well.
Corollaries to Existing SafM:
The SafM today addresses utility as well as depositor insurance. It asks AAVE holders, who by their holding AAVE have skin in the game, to take more risk in the form of slashing. In exchange, they get AAVE rewards. The issues with this setup are mentioned above, namely it is an ineffective insurance pool and unsustainably expensive. Paying stakers an exorbitant budget for a poorly provided service is bad business.
Conclusion:
The Safety Module is critical for Aave users and it should be optimized for their protection while being sustainably funded by the DAO. AAVE as an insurance asset is an irresponsible choice for depositors, and the DAO should research the use of more liquid assets in the SafM. AAVE utility and governance can be amplified by implementing a Medallion Token model and creating a Staking Module that mirrors the current benefits to stakers in the Safety Module. This design improves the Safety Module for depositors, value for token holders, and sustainability for DAO spending. There is more research to be done, particularly around the potential costs required to simultaneously incentivize both the SafM and the stkM, but we appreciate all feedback and hope this discussion is fruitful in driving the best outcome for AAVE token holders and Aave users.
edit: changed to discussion