Compound

Given that Compound V3 USDC Comet has been live for several months now, and a wETH Comet is upcoming, Gauntlet would like to help kickstart the community conversation around migrating users from V2 → V3.

Gauntlet’s scope on Dynamic Risk Parameters is focused on areas in market risk; however, we thought it would be helpful to have one forum thread to conduct community conversation and decision-making around V2 → V3 migration. We defer to the community on strategic decisions but wanted to share some initial thoughts to help kickstart and organize community discussions to drive progress.

V3 offers capital efficiency and risk management enhancements compared to V2, including additional parameters such as supply caps. As such, it is valuable for the community to align on the goals of migration and the risk tradeoffs.

V3 currently only has 1 instance, which is USDC as the base asset (borrow asset), and the only collateral assets are ETH, WBTC, COMP, UNI, and LINK. Because V2 is multi-asset borrow and multi-asset lend, the market structure differences add additional complexity to any migration plans.

What should the community define as the goals and non-goals of migration?

• Is the goal to incentivize user position migration to V3 as much and as fast as possible, or should V2 co-exist as an “equally attractive” protocol?
  • Due to external protocol integrations with Compound V2, a total migration may not be immediately possible
• Should the protocol care about losing recursive borrowers on V2?
• It would be a sub-optimal outcome if the protocol’s actions result in just losing V2 TVL instead of accelerating V2 → V3 migration. These outcomes are difficult to predict, but how can the community best minimize the chances of this risk of losing TVL?

Strategic Considerations:

• From our understanding, Compound already launched a tool where it is easy for users to migrate positions. An easy UI and education on its use can be helpful in accelerating migration.
• How should the community handle recursive borrowers that won’t be able to migrate to V3?
• How should the community handle assets that are listed on V2 (or even frozen on V2) that are not listed on V3 Comet? Should any of those assets be added to the existing V3 USDC Comet? Should other USDC Comet markets be spun up?
• Should a DAI Comet market be spun up? This will help users that borrow DAI on V2 migrate.
• Increasing supply caps on V3 adds risk to the V3 markets (mitigated by decreases in risk in V2 markets). However, increasing caps more proactively can allow for faster migration.

Potential Paths for Encouraging Migration:

• Note that many of these should be done simultaneously to be most impactful
• Reduce collateral factors on V2
• Incentives on V3
• Interest rate curve changes on V3 to encourage more attractive borrowing terms
• Cut all incentives on V2
• Raise supply caps on V3
• Increase RF on V2 to disincentivize V2 supply
• Freeze all existing V2 markets

Potential Risks:

• Note that this is not an exhaustive list, but is aimed to kickstart community conversation
• Liquidity Crunch risk
  • If we encourage supply to migrate from V2 to V3, and if borrowing levels on V2 don’t change (e.g., if borrowing is inelastic to interest rates), there is the risk of 100% utilization on V2 markets. Suppliers won’t be able to withdraw due to the lack of liquidity. In addition, 100% utilization increases the chances of insolvency if they interfere with atomic liquidations.
• TVL loss in V2
  • User experience risks, confusion from users
  • In addition, shorts are currently not supported on V3 (as USDC is the only borrowable asset)
• Overspending on incentives

Next Steps:

We welcome community feedback and look forward to discussion.

By approving this proposal, you agree that any services provided by Gauntlet shall be governed by the terms of service available at gauntlet.network/tos.

Hello @pauljlei,

Thanks for opening this discussion here on the forum. We (Morpho Labs) would like to insist on this 2 points:

• As you mentioned, the V3 is not offering all the use cases of the V2, due to the low number of borrow markets opened at the moment and the restricted list of collateral. It is not clear that the community wants to open markets other than ETH, and the biggest stablecoins. Forcing the shutdown of the V2, which has proven to be resilient to various market conditions, could make Compound loose its organic users, which could move their positions to other markets such as Aave (or Morpho-Aave).
• The Compound ecosystem is large in terms of on-chain and off-chain integrations, dashboards, analytics… These networks take a lot of time to build, and pushing migration to Compound V3 could also fade this ecosystem away.

That’s why we are more in favor of a slower but natural migration. This means doing no change to Compound V2’s parameters for the sake of making liquidity migrate to V3, but instead gradually pushing the V3 to its full potential. We recognize the all the benefits of the V3 for the use cases it covers and are confident in the fact that concerned people will progressively migrate their positions. This is also a strategy that seems to minimize the risks associated with this migration that you mentioned.

About COMP incentives, we don’t have any strong opinions. It should be an effective way to accelerate the organic growth of Compound V3. Reducing them on Compound V2 will certainly decrease TVL, but will not necessarily “kill” the organic use.

I think launching DAI Comet market would be very critical before aggressively pushing users to migrate.

Completely agree. I believe that incremental migration is better, especially for key assets. For example, DAI market can be fully deactivated on V2, once 80% of the borrow volume (as of a certain cut-off date) has moved to V3.

This is an exciting and important conversation. To add to the initial context, there is also a Twitter Spaces with OpenZeppelin/Gauntlet/myself on 1/10/23, and the Community call on 1/11/23 where the migration path has been discussed.

To echo what others here have mentioned, because V2 and V3 don’t completely overlap in usecase, V2 should be given the opportunity to run in perpetuity. Rather than taking steps that deprecate V2 (such as lowering collateral factors or freezing markets), we should strive to transfer users and risk for V3’s core use-cases (e.g. borrowing stablecoins).

I also agree that this process should be gradual or phased, which will allow Gauntlet and the community to monitor the liquidity and market risk of each version and recommend changes (or pause) if necessary.

Here is one potential draft approach for consideration and fine-tuning:

• Each week, for 8 weeks, there is a linear parameter adjustment proposal intended to migrate users and risk between versions
• These proposals would migrate the COMP Distribution from V2 markets to V3, increase V2 stablecoin reserve factors, and raise supply caps on V3 to accomodate new users
• Migrate reserves from V2 to V3 corresponding to the quantity of active borrowing volume between versions (weekly, or at the end of the migration)
• If needed, adjust interest rate models in either V2 or V3

For the remaining strategic considerations, here are my personal views:

• Agree with Gauntlet that the Position Migrator should be given additional educational resources and awareness
• I don’t think that recursive or non-organic use of V2 matters, and that if this is eliminated the protocol & ecosystem will be left in a healthy position
• By preserving V2 use-cases (such as using BAT, ZRX, SUSHI etc as collateral), they don’t have to be immediately added to V3; the community can decide which assets to add to V3 markets as collateral
• Launching a DAI market is up to the community; that being said, V2 once re-upgraded to support the DSR, might be a superior market to V3 (unless V3 is modified to support the DSR)