I have submitted a proposal to the Compounds Grants committee about implementing an extension to support real world assets with the Compound protocol. Real world assets could be any asset that exists physically and produces an income stream or gains in value over time such as real estate, art, commodities like gold, etc.
The value proposition is two fold
- enable real asset owners to stake them as collateral in exchange of liquidity
- enable staking of liquidity to borrow real assets and benefit from underlying income streams they produce
The proposed project will involve working on the interest rate model that would factor in income streams underlying real assets staked on Compound, and since real world assets are expected to be staked by asset managers and asset owners, we expect a number of assets to be proposed to be included in the RWA market.
I am writing here to ask for your suggestions on the governance model you feel will work best to review and approve submissions of RWA assets for inclusion in the RWA market on Compound.
An idea we have is to include an asset submission, review, voting, approval process in the RWA extension where information about the asset and parameters (eg, collateral, reserves, interest model, etc) about a specific asset is used to create a DAG node on IPFS and then IPFS pubsub is used to count votes in favour of an asset being listed on the RWA market. This will move the voting offchain but still retain the essential properties of immutability, engage a larger community at no cost (IPFS requiring no gas) and can then post the result on whether to list or not list a RWA asset onchain.
I look forward to your views and ideas.
This is an interesting concept and I am open to the idea of it becoming supported by the protocol some day. If my understanding is correct it is not possible currently.
- How can an autonomous protocol, where software rules have complete control, liquidate a real-world asset?
- How can this be done atomically within one EVM blockchain block which are usually 12 seconds or less?
- How can a liquidation be completed timely without risk of leaving bad debt in the protocol?
- How can a RWA liquidation be done in a trustless fashion?
Also, a Compound Extension is a front-end application that is accessible from the starter UI at https://app.compound.finance/extensions. What you are describing sounds more like a protocol code and likely also an architecture change.
RWAs are actually surprisingly not very uncommon already, I found this interesting map of the RWA ecosystem from a Binance Research Report which shows that there is already a growing number of Defi players in this space
I also saw a very interesting discussion thread in the Maker DAO forum a couple of years back posted by Societe General’s Forge team where they wanted to post onchain collateral (for a debt token, if I remember correctly) and now, it seems that Makero DAO vaults hold $680 million collateral from RWA issuers.
Coming to your specific questions, here is what we think
although a real world asset is physical and offchain, the collateral posted will be onchain. Taking an example, we are currently working on a EUR 5 million real estate backed loan where the asset manager will pay 12% per year in interest and will put up collateral (a redemption reserve) which will be on chain. The redemption reserve is always maintained for the next two quarter’s interest payout. For the principal amount o f the loan, the loan is secured with guarantees by the asset manager. Hence, liquidation of the RWA token issued will still be on chain. The underlying physical asset is liquidated and managed by the asset manager which is usually a large institution with sufficient liquid funds. This is a key criteria for qualifying RWA assets that can be supported by the new proposed market.
The second question is related to the first above. If collateral is onchain, liquidation is onchain as well.
Liquidation can be completed timely without risk of bad debt in the protocol only when those RWA tokens are allowed which are backed by on chain redemption reserves in USDC/some Compound supported stable asset.
RWA liquidation is based on liquid collateral posted by asset managers. The quality and credit worthiness of asset manager is key. In the case of the Societe General example with Maker DAO, Societe General is the asset manager which is managing the borrowing done by an asset owner (could be SocGen’s client). The client owns the RWA and needs the liquidity, and the asset manager arranges the liquidity by guaranteeing it with collateral.
With respect to the Compound extension itself, it will have the web extension which is the front end UI for submitting / qualifying / voting / approving the RWA but it will also have a smart contract operator for managing it on chain. No architectural change of the Compound protocol itself is required.
So from the protocol’s perspective, there will be an ERC-20 token that is collateral for this RWA. Can you describe what that new ERC-20 will be like? Will it be an entirely new token or could it be an existing token like WETH? The on-chain governance system can have a new collateral token added to a Compound III deployment if it has a proper price feed, usually via Chainlink.
I don’t understand why you would need an extension for voting. Compound community governance is on-chain, and there are several voting interfaces available to COMP delegates, why would there be a new one for a single governance proposal?
There are 3 tokens - Real World Assets (RWA), Collateral assets, Base assets.
RWA tokens may not be liquidated on demand, but they may pay off higher interest rates. Collateral asset tokens posted for RWAs can be existing tokens accepted on Compound such as WETH. We may not need a new collateral token to support RWAs.
The Comet extension for RWA Markets could have features
- to let RWA issuers (asset managers or owners) post asset details, post collateral to Compound and borrow from Compound.
- to let RWA extension users view RWAs, put up any collateral that is required to be filled in for such RWAs that is already not posted by the RWA issuer.
The benefits will be
- For RWA issuers : access to liquidity on Compound in the form of loans of base assets.
- For RWA Comet extension users : benefit by higher interest rates paid by RWA issuers (ie, if a RWA is paying 12%, and the base asset cost of capital is 4%, RWA Comet extension users who post any collateral would earn the balance 8% from RWA issuers).
- Compound protocol : more liquidity (ie, more collateral + base assets on chain).
After reading the Comet code and also noting that there is a cap of 15 collateral/base asset pairs (can you confirm this?), its best that we do not introduce any new collateral tokens that would require voting / governance to approve.
We could still have some social engagement on the extension by letting COMP delegates vote on admitting RWAs on the new market, but this is something we can check the community’s views on first before implementing.
Great insights. I think that it is best to use collateral for RWAs on chain that are already supported by the protocol, like the ones you can see on the markets page (Compound Markets). If this is indeed possible, then there is no need for a on-chain governance vote to add a new collateral like the process that is currently going on with stMATIC and MATICX. From the protocol’s perspective, it would not care that the collateral is for an RWA, but simply treat it precisely like any other collateral.
If that is not the case, then there would need to be some analysis of a new RWA ERC-20 being implemented. The protocol needs for it to be sufficiently liquid on the chain so liquidations can be done promptly. Also there needs to be security and economic analysis by community members like OZ and Gauntlet. Yes, there is a cap for number of collaterals allowed but presently we are not close to the cap in any of the v3 deployments.
An Extension for users to interface with the protocol is a great idea. The community is always welcomed and encouraged to create more interfaces for the protocol to ensure that there sufficient is decentralization and redundancy. Would you be the designer and developer of this Extension? I made an Extensions developer guide with resources to get started building one. Happy to help out.
thanks for your inputs. We will then build support for real world assets on Compound using support for existing collateral tokens on Compound. I and a couple of engineers will design and develop the RWA Markets extension. I went through the link to the guide you have authored for new extension development and have also gone through the extension templates and the Comet code itself.
I will then wait for the allocation for this proposal so that we can start work on this.
a description of our work along with the architecture doc is here.
It will be good to get your feedback and also please circulate with anyone in the team who may have an interest in the topic.