I strongly agree with Robert that the v2 protocol should be kept running in perpetuity. This informal proposal is intended to suggest a few changes that will leave Compound v2 in an optimal state for long term safety and support of its existing integrators. We look forward to integrating with v3 in the future, but feel the current v2 market has a valuable role to play as one of the most liquid existing DeFi markets with a relatively more conservative collateral set than Aave.
I propose that Compound go further by eliminating all remaining collateral assets whose liquidity-adjusted market capitalizations are lower than the combined value of borrowable stablecoins and ETH in the v2 market. Namely, I suggest offboarding all three of UNI, COMP, and SUSHI as collateral. UNI and COMP are both available as collateral in the v3 market where more risk controls are in place. SUSHI is not a desirable collateral in the long term in my opinion, but could be added to the v3 market if desired.
Notably Euler has already removed the Collateral factor of LINK and UNI to reduce systemic risks.
Euler is the first major DeFi lending market to enable USDT as collateral. While a contentious topic, I think we can all agree that 1 USDT is not worth zero. Even a lower and more conservative collateral factor than proposed for Euler, say 50%, would add a major new use case to Compound, allowing users to either long or short USDT according to their preference.
We at Volt Protocol expect to integrate with Euler in the future and to be able to conduct cross market arbitrage between Morpho-Compound and Euler, for example. Extending collateral status to highly liquid assets will help increase efficiencies in the system and lower costs for users.
It’s probably best not to add many new assets to v2 in the future. I would adhere to a principle of not even considering assets with less than 1b market cap as collateral given the current size of the Compound market.
The asset most sorely missed in the v2 market today is wstETH. Even though a v3 ETH market is planned, it doesn’t change the fact that adding wstETH collateral to v2 would unlock new opportunities at minimal risk.
In the future, BUSD, GUSD, cbETH, and rETH all have potential to onboard.
It’s a nice idea to use a decentralized price feed as an “anchor” or reference point while using Chainlink as the main feed. The problem is that this can actually introduce additional vulnerability, as discussed in our security review. The Compound v3 market uses Chainlink exclusively, and I feel v2 should do the same for now by getting rid of the anchor entirely.