435.17%
Collateralization Ratio
I was studying [Gyroscope design] and find interesting the idea to relate the redemption price to the redemption level (% of supply). Maybe a combined metric of CR and redemption level would be good.
Source: https://docs.gyro.finance/learn/how-the-pieces-fit-together/gyro-amms/p-amm
The point is, in case we are undercollateralized, how can we avoid a bank run and have time to sell TRIBE at reasonable prices to increase CR? The problem is TRIBE price going down and making more difficult to recover and restore the confidence in the system.
The limitation of amount of PCV on the redemption contract could help somehow?
If it is not too complex, another idea would be to consider a dutch auction mechanism in this stress scenarios (sth as used by Float Protocol). It was an idea we discussed in the early days of Fei.
When Fei Protocol is undercollateralized and Fei is below $1, the minted Tribe could be used in a dutch auction to defend the peg. If CR is 80%, we could use 80% of reserves and 20% of minted TRIBE to buy Fei.
The auction advantage is the capital efficiency and paying only what the market requires to restore the peg.
Redemption is currently rate limited by how much PCV can enter the stabilizer (5000 ETH per hour), we’re also looking into dynamic redemption fees similar to LUSD and UST.
We’ve actually developed a dutch auction contract using Balancer V2 smart pools for the buybacks! It should be possible to repurpose for the backstop, although we’ll have to have more discussions on if that is the right path. We may only want to backstop as-needed to defend the peg and not push TRIBE out into the market when the peg is fine even if undercollateralized.
Happy to hear about this developments with dutch auction buybacks.
For the backstop, we could do dutch auctions every x hours when we are below the peg. Minted Tribe would only be used when we are below the peg AND undercollateralized, otherwise just PCV would be used for the backstop dutch auction.
One of the things I’ve been hearing recently is DAO Treasury Management. It is important to note that for Fei, the Treasury is not really the PCV - currently the lack of activation of the Treasury is a common challenge for other projects. Perhaps this is outside the scope of V2 but do you have any thoughts on how to better use the capital there other than a potential backstop?
This is a good point and something we need to discuss.
My mental model is that the DAO “owns” the PCV, and all non-TRIBE assets are simultaneously collateral for FEI and operating capital for the DAO (such as using ETH for gas refunds).
In this light, we can always sell DAO TRIBE for more PCV and vice versa. Not sure if we need anything more explicit.