This proposal lays out an overarching framework for setting RWA exposure for the GHO stablecoin. Once the target facilitator and overall portfolio characteristics for RWA are defined, then AAVE should select facilitators and asset types that fit within that strategy.
Facilitators could be other DeFi protocols (Centrifuge, TrueFi, Goldfinch) or non-protocol entities (funds). The important thing is that facilitator strategies are well-defined and that they fit as a puzzle piece in the larger picture in an RWA portfolio strategy.
A portfolio strategy should probably be quite conservative and simple to start. Well-versed credit professionals should define this but it could look something similar to what Rune lays out for Maker in MIP103.
- Maturity & liquidity profiles
- Target at most [ x ]% exposure to collateral that can be converted to Cash Stablecoins within  days
- Target at most [ y ]% exposure to collateral with a duration of  months to  year, with a  month average duration. This can also be substituted for collateral with a shorter duration.
- Target at most [ z ]% exposure to collateral with a duration of [1-3] years, with a  year average duration. This can also be substituted for collateral with a shorter duration.
- Benchmark rates
- Target returns should be informed by defined benchmark rates
- [benchmark return for very low-risk collateral is equivalent to the effective return on Primary Yield Stablecoin]
- [benchmark return low-risk collateral is equivalent to the effective return on the 3-month treasury bill
There have been preliminary discussions from interested RWA facilitators who are looking to be early facilitators for the GHO stablecoin. It feels premature for AAVE to be engaged with RWA facilitators without first defining what they want from RWA facilitators (maturity, liquidity, rates, etc.) and creating a process for selecting originators.
Facilitators onboard RWA collateral to Aave. They are responsible for structuring these collaterals and attesting to their ongoing value and performance. Facilitators may be protocol or non-protocol entities.
- Facilitators are onboarded by polling of Aave governance.
- Facilitators present Aave governance with new collaterals to onboard.
- Facilitators must clearly identify the collateral, how assets will be transferred, and how collateral will be monitored.
- Facilitators must identify material risks associated with the proposed collateral and onboarding or onboarding that collateral.
- Facilitators must identify clear on-chain and off-chain flow for assets, both for onboarding and for liquidation.
- Collaterals are presented by facilitators and onboarded by polling of Aave governance.
- Individual collateral should have a narrowly defined scope, with easily understood properties and clear mechanisms for onboarding and offboarding.
- Collateral will have a clearly identified process and schedule for monitoring.
- Facilitators will identify how this collateral fits within the Asset-Liability Management model.
This model addresses liquidity risk around assets.
Categories of Real World Collateral
Categories will be grouped into different buckets based on duration and ease of liquidation.
These assets can be immediately liquidated by anyone and are the first line of defense for the peg. The protocol should initially aim to hold 30% of real-world collateral as stablecoins. These assets will be managed by the protocol.
- Short Term Real World Collateral
This class of collateral has a duration shorter than a year and highly stable value, such as short-term treasury bonds and money market funds. The protocol should initially aim to hold 70% of real-world collateral as short-term real-world collateral. These assets will be onboarded by facilitators.
- Long-Term Real World Collateral
This class of collateral has a longer duration, with correspondingly higher volatility. These collaterals can be considered as the surplus buffer grows and the market matures. These collaterals will be onboarded by facilitators.
We look forward to a discussion in the comments. We will kick off the discussion in the comments with some “open questions” we came across while researching this framework.