EDIT: removed TRIBE derivatives and TribalChief tokens per discussions
EDIT2: Subject FEI derivatives to $5m global supply cap and increase seed max to 2.5m FEI
Proposal: FIP-39
Status: Last Call
Summary
Expand Optimistic Approval Charter for Fuse to add frameworks for:
New assets
Verified vs Unverified pools
Adding/Removing FEI
Liquidity mining rewards
Motivation
In FIP-13, Tribe holders authorized Optimistic Approval to deploy newly minted FEI to lending markets following a framework. This only included the minting of FEI, without any provisions for removing, or more granular instructions for which pools can be supported. With the recent launch of permissionless pools, many include FEI and this introduces some arbitrariness into which pools get seeded.
Rari Capital DAO holds snapshot votes to determine which DAOs and users get the whitelisted ability to create verified pools. This adds a layer of protection around fraudulent or immature pools for prospective users.
Many discussions in the community have revolved around adding new FEI to markets with high demand and removing them from markets with low demand.
FIP-29 granted Optimistic Approval power over the FeiRari pool, including the ability to add new assets and change parameters.
Proposal
Expand FIP-13/29 charter to authorize OA to do the following:
Add any new FEI derivative to the FeiRari Fuse pool subject to:
5m supply cap
Sufficient on-chain liquidity at least 1x amount in Fuse
A manipulation resistant oracle (e.g. Chainlink or Uniswap TWAPs)
Verifiably cannot be minted arbitrarily
No recursive fuse derivatives
Add FEI up to $2.5m FEI in any verified Fuse pool but NOT unverified pools without a snapshot vote (DAOs behind unverified pools can petition on forum)
Increase FEI in a pool by enough to move below target utilization if:
above target (the “kink” in the interest rate model, usually 80%) for more than 1 week straight
whitelisted by DAO for increases (only FeiRari to start)
Decrease FEI in a pool to up to 50% of the current liquidity if:
below 20% utilization for more than 1 week straight
The points regarding Increasing/Decreasing FEI in pools also relate to non Fuse markets such as Aave, CREAM, Kashi.
Add derivatives to FeiRari
Whitelist only verified Fuse pools
Increase FEI in a pool if above target (including FeiRari)
I’m dumb, can someone please explain what a derivative is in this context: “Add any new FEI or TRIBE derivative, or TribalChief reward token to the FeiRari Fuse pool.”
And if I could get a real world example of when/why we would do this, that’ll help my wrap my smooth brain around it.
Essentially we want to allow users to borrow against any assets that have good oracle prices, have certain properties that make them secure as collateral, and are staked in the TribalChief such as FEI/TRIBE LP tokens, FEI+3CRV tokens, and G-UNI-FEI-DAI LP tokens. That way, you can earn yield on your assets in those protocols and then leverage them to borrow more of any asset.
At this stage, I would only approve OA to add/remove FEI to whitelisted pools + whitelist verified pools on Rari. Adding derivatives and adjusting TribalChief rewards are more contentious issues that potentially effect lots of users and create emergent behavior. I would hold off on those until OA can demonstrate autonomous judgement for a longer period of time (sorry guys). In that vein, I would like to see OA start to hold regular meetings to discuss operational aspects of the expanded charter (eg reviewing liquidity deployed on Rari to see whether adjustments are needed)
As for the add/remove FEI operation, I presume its subject to the monthly FEI allocation, and if FEI is removed from one pool it can be deployed elsewhere immediately?
Lastly is there a limit to how many OA actions for add/remove liquidity in a certain time period? While I trust the authority and the competency of the OA members, there should also be some guardrails to not needing to make adjustments impulsively.
To be clear, adding TribalChief rewards to FeiRari is NOT intended behavior. Only the underlying tokens. e.g. G-UNI FEI-DAI.
The intention is to eventually potentially allow OA to transition existing approved rewards to FeiRari from TribalChief similar to fTRIBE, but this is not part of this proposal.
I think adding derivatives to the markets that fit the criteria does not meaningfully increase the OA’s power to act autonomously, as this is a relatively algorithmic set of possible tokens. Perhaps we can limit to FEI derivatives only, as these are generally more useful and increase the native utility of the FEI market.
Any FEI removed would still be available for deploy.
I think there shouldn’t be a limit. If OA ever begins to act frivolously, the DAO can veto and consider restricting OA powers.
One of the advantages of OA is for the group to be a lot more nimble in reacting to market conditions. Currently OA has been doing a good job at executing snapshots, so no issues there. The OA also consists of a multi-sig so there would be some dampening of activity there. My only concern was about overreacting to certain events and not necessarily actions by potential bad actors. As long as there is recourse to challenge OA actions then I think the current policy is fine.
I would be in support of a phased charter expansion starting with FEI derivatives and seeing how that goes for a quarter.
Increase FEI in a pool by enough to move below target utilization if:
above target (the “kink” in the interest rate model, usually 80%) for more than 1 week straight
whitelisted by DAO for increases (only FeiRari to start)
Decrease FEI in a pool to up to 50% of the current liquidity if:
below 20% utilization for more than 1 week straight "
I would say these perimeters are too rigid and potentially prone to exploits if they became well known. It should be a “substantiality test”, whereas a set number of hours, say 120 hours (consecutive or not) within a window of one week/10 days should suffice.
Otherwise users can manipulate a spike either way to derail further injection/withdrawals. The OA should have some measure of discretion on liquidity movements until we develop a more sophisticated set of parameters. (such as cross comparison with the average interest rate of a basket of other stablecoins at leading lending protocols)
This is why in addition to an expanded fuse charter the OA needs a formalized framework in reviewing market conditions for further intervention. In other words, the OA should be given responsibility in meeting regularly and discussing/reviewing actions.
I support this. It would be good to make the framework @arcology mentioned after this voting to consolidate all the FIPs, the scope and the procedures of the multisig. What are you thinking to add to this framework, @arcology?
I agree that OA scope is becoming more complex and requiring more structure. Some additional ideas: a dashboard where we could view all the results from OA actions. Also, for transparency, it would be good to include in Fei Newsletters, a summary of the actions executed by OA on that month. Maybe something in Fei History, specific to OA history.
That would be really cool, and fitting. I think logging the history of the OA is like logging the protocol’s transition towards further decentralization, as the community grows in responsability.
After the CREAM incident, can we wait a bit before introducing derivatives in our lending markets ?
A more simple analogy is that the OA is now becoming the central bank executors of monetary policy (the Fed, ECB, PBOC) with recommendations to/from Fei Labs.
Under this role, I would suggest the OA setting up criteria on when and how they are to intervene (for example, if Rari borrowing rates spikes to 80% plus, is it after a sustained period? What happens if there is an event which pushes it to 99%? Do they inject liquidity right away?). I would also defer to the OA committee to introduce and set up regular sessions to discuss such criteria and current lending market conditions.
My intention was that decreases for inactivity could be the default. We actually could have gotten a lot of FEI out of CREAM if this were a policy.
I agree this is a great idea, but lets keep it high like 90% of the week.
A dashboard may be a bit heavy, but we can easily make a discord bot which sends any on-chain transaction through #optimistic-approval.
Yes, we can make derivatives on a case-by-case basis. I do think we should consider adding them. We can use borrow caps to protect the pool.
I think formal sessions would be a bit of “power creep” for OA. The intention is to still enact only the will of TRIBE holders via snapshot frameworks. There is some interpretation to be had, but not a ton. If OA ever does something TRIBE holders don’t like it can be veto’ed or at minimum discussed and reversed after the fact.
Not opposed, but not really in favor of that at this stage
