VIP XX: Liquid Venue Onboarding

Background

why new venues?

While technically two separate venues, Morpho-Compoundv2 and Compound v2 “vanilla” represent limited diversity for the most liquid portion of the VOLT backing. After gaining a deeper understanding of the Compound v2 codebase through our review, it’s also clear that integrating markets with fewer long tail assets and more robust safeguards on oracles, per asset borrow caps, etc is highly desirable. There is good reason to hope for a steady hardening and ungoverning of the Compound v2 market, such as the recent proposal to pause supply of assets potentially vulnerable to price manipulation. at the same time, as COMP incentives are reduced yields there may decline.

VIP Structure

Venue onboarding should always be done one at a time, given the meaningful time commitment required for the Volt Protocol team to conduct and publish economic and technical reviews prior to integration, as well as to write (and have reviewed/audited) integration code.

In this thread, we’ll be working to break down the pros and cons of each venue, considering factors such as:

  • previous security incidents and how team responded
  • security diligence done such as audits
  • collateral quality
  • oracle system
  • system lindy, how many upgrades have occurred on the code
  • TVL over time
  • bug bounty programs and previous payouts
  • historical and current yields
  • anticipated near term changes, whether incentives programs, collateral onboarding or offboarding

Below, I will start to make a reply for each venue and edit it with more details.

1 Like

Compound v3

Market Overview

While it doesn’t rank high on the Lindy factor, Compound v3 represents a strict improvement on Compound v2 from the perspective of safe handling of collateral, notably adding a per asset supply cap and removing the Uniswap fallback oracle which in v2 potentially introduces risk over the Chainlink oracle. Compound v3 markets have only a single lend/borrow asset, and the current instance supports USDC. Given the current proposal to deploy an ETH lend/borrow instance, DAI and other versions may not be far behind.

Collateral Risk

The collateral assets present in the market are currently limited to ETH, wBTC, COMP, LINK, and UNI, with reasonable per-asset supply limits (will go into specifics on this soon). There is an upcoming proposal to increase supply limits for ETH, wBTC, and COMP, all to what seem like reasonable values. There are no active proposals to onboard new collateral assets.

Yields

The Compound v3 USDC instance, which we could call cUSDCv3, is currently offering slightly greater yield than the blended lending + incentive yield available in the Compound v2 DAI market. I expect a transition of COMP rewards away from the v2 market before too long, with the result that yields are generally higher in the v3 market or via Morpho peer to peer matching than in v2.

Conclusion

I currently feel that Compound v3 is the best candidate for the next liquid venue to onboard for Volt Protocol PCV deposit, pending the results of smart contract security review.

Aave v2

Market Overview

Aave v2 is the largest and most liquid lending pool on Ethereum. Compared to Compound v2, it is much more active in terms of potential new collateral assets and functionality. There was originally a plan to directly upgrade the v2 market into Aave v3; fortunately the decision was made to pursue a separate deployment as Compound did. Aave has removed its incentives and so has lower yield currently than Compound.

Collateral Risk

In the Aave system, there are active efforts to onboard more exotic collaterals like LP tokens. The Aave v2 market has substantially more long tail assets than Compound v2 or Compound v3, such as REN, ENJ, or MANA. While there are steps being taken to mitigate long tail asset risk, these fall well short of what I would consider adequate.

Yields

With no ongoing AAVE incentives, yields are well below those of the Compound v2 or v3 markets.

Conclusion

I believe that Aave v2 is not suitable for Volt Protocol PCV deposit, primarily due to excessive risk from long tail collateral assets.

1 Like

DAI Savings Rate / Current Thoughts

With a poll indicating MKR holders will support activating the DSR to as high as 1% in the near future, it’s a good time to prioritize direct integration of the DSR as a PCV Deposit. Compared to lending on Compound or other similar markets, this is a meaningfully lower and uncorrelated risk profile. We are working on upgraded versions of system components like the PCV deposit to support the goals of market governance, such as on chain profit and loss tracking in venue deposits. After this upgrade is complete, we will be able to onboard a series of new venues. Expecting this to be the next audit round at the end of January or early February.

Since the start of this thread, there have been meaningful steps to reduce long tail collateral risk in the Aave v2 market. We’ve also been working on the next step towards market governance which will support per venue allocation caps.

As a result, my current expectation is that we can integrate a large batch of venues with appropriate supply limits once the next iteration of the PCV deposit and other core development work is complete.

These would include:

  • DAI Savings Rate
  • Compound v3
  • Aave v2
    • correspondingly, Morpho-Aave
  • Euler
  • Aave v3 to be evaluated based on its supported collateral set

In all cases, the Volt Protocol developers will conduct a security review of the associated protocol and adjust recommendations for allocation limits accordingly.

The riskiest venues should have a deposit cap equal to the size of the surplus buffer. This prevents VOLT holders from being exposed to loss even in the event that 100% of the deposit in a given venue is lost. The most likely reason for such a total loss is smart contract or market manipulation risk. Any venue that includes a high risk collateral asset, for example Aave v2 which includes SNX collateral, should be subject to the risky venue cap, ~293k based on the current size of the surplus buffer.

Currently, I am in support of no-cap for the DAI Savings Rate, and imposing caps of 25% of the PCV on both the current venue Morpho-Compound v2, as well as new venues Euler, Compound v3, and Aave v3, and the lower cap of ~293k on Aave v2 / Morpho-Aave.

Once we have completed integration of this range of venues, it will also be feasible to integrate a few cross market arbitrage strategies in the market governance menu. For example, borrowing ETH on Morpho-Aave against DAI deposit, and lending it on Euler.