Keeping consistent with the concept of providing the most robust and accurate access to the value of Fei’s peg to $1 USD, the goal of this idea is to start a discussion on why having an active & decentralized fallback oracle in place for Fei DAO is important, and why Tellor is specifically well-equipped to fill this role.
The Problem:
Fei’s DAI PSM (Peg Stability Module) aids in keeping Fei’s peg to the USD by allowing users to deposit DAI and receive minted Fei at a 1:1 ratio. If the price of DAI falls outside the $0.98-1.02 range according to the oracles, the PSM contract halts its DAI/FEI exchanges. However, at present, Fei’s DAI PSM points to a constant price oracle as its backup oracle, which puts Fei at risk during a distressed event where its primary oracle fails. In this scenario, pointing to a constant price of DAI at $1 exposes a potential vulnerability where:
-
DAI loses its peg
-
The attacker mints Fei at a 1:1 exchange rate
-
Attacker exploits the arbitrage opportunity
-
Fei’s peg gets exposed to resulting sell pressure
While this could be perceived as a small & unlikely scenario, oracle safeguarding is becoming more commonly implemented, as the fallout of negative consequences (loss of peg, dev effort to fix, downtime, project reputation, etc.) are exponentially larger than the minimal proactive effort required to have a sound contingency plan in case of failure.
The Solution:
We are proposing the idea that Fei governance change the backup oracle in the DAI PSM contract to point to Tellor, replacing the current ConstantOracle contract. Integrating Tellor as Fei’s provisioned backup oracle for the DAI PSM will not only keep Fei’s backupOracle DAI/USD feed in a live state with no downtime, but provide a proven backup solution in the DeFi space.
Proven How?
Tellor’s reputation is established as the backup oracle for Liquity, as well as a median input for Ampleforth. These use-cases are especially relevant to Fei’s needs as there are several similarities. Liquity is an immutable contract and Ampleforth required a governance vote to implement the further decentralization of their oracle.
We want to know what the Tribe thinks here so questions/concerns are encouraged to help determine if this solution is a good fit and worth moving forward as a FIP.