FEI Deployment Strategy

This forum post kicks off a discussion for a more structured approach to PCV exposure to minting and deployment of FEI for various applications, such as Laa$, Turbo, and yield generating use cases. It would also end the FEI contractionary monetary policy and propose an initial deployment strategy.

Going forward, we should enable a policy with guidelines for the amount of mintable and protocol deployed FEI based on market conditions, until a more automated approach is developed for management of expansionary and contractionary policies.

The ability to quickly remove FEI from various deployments is just as critical if we have to pursue a contractionary policy again. This makes it very important to identify a suitable amount of FEI as well as appropriate deployment strategies.

The Fei Labs team performed a baseline PCV Stress Analysis. The analysis examined various market scenarios and characterized FEI’s on-chain liquidity depth. This analysis suggests that up to 90M FEI can be deployed across Turbo, Fuse Convex Farming, Arbitrum, and Compound without significant risk to FEI’s peg or collateralization. Currently, there are 64M FEI in OA Timelock, as well as additional 26M FEI from 50M FEI Laa$ allocation, which could be utilized for a guarded reentry into expansionary PCV management mode.

This represents a fairly conservative position which allows the Tribe DAO to effectively launch Turbo, add FEI liquidity to Arbitrum, provide additional borrowable funds on Compound, while continuing with Laa$ and further supporting Fuse pools.

Proposed deployment strategies:

Turbo: $40M FEI (estimated 5% APR): the initial deployment for 40M FEI, but with the planned expansion of Turbo to Aribtrum, Tribe DAO should pre-approve additional 10M FEI to be minted as needed

  • Double collateralized
  • Strengthen DAO-2-DAO partnerships
  • Increases Fuse TVL and FEI utility
  • Technical requirements:
    • Nothing new to develop (just whitelisting collaterals & setting their caps)

Compound: $5M FEI

  • Blue chip lending platform
  • Visibility and utility for FEI
  • Technical requirements:
    • Nothing new to develop

Arbitrum: $10M FEI

  • Layer 2 presence and adoption
  • Technical requirements:
    • Need new contracts to interact with the Arbitrum L1ERC20Gateway (L1->L2)
    • Need new contracts to interact with the Arbitrum GatewayRouter (L2->L1)
    • Need new contracts to propagate L1 PCV_CONTROLLER role to interact with contracts on L2 and handle PCV securely

Minting FEI to Convex Fuse Pool: $10M FEI

  • Increases Fuse TVL
  • Increased FEI utility with long tail assets
  • Technical requirements:
    • Nothing new to develop

Laa$ ($25M)

  • Currently deployed $50m
  • Generating 5-7% revenue on deployed FEI
  • Expected deployments
    • SYN (estimated 14M)
    • NEAR (estimated 3M)
    • FOX (estimated 3M)
    • POOL (estimated 1M)
    • UMA (estimated 4M)
  • Improved returns and lowered exposure

Look forward to the participation of the community in this discussion and working with various partner DAOs to launch Turbo and other products.

4 Likes

Cool analysis! I’m sure that the framework laid out here will prove useful for many protocol decisions down the road, as well as for informing the current deployment strategy. I’m assuming Delia led the charge?

I just have two minor comments. In Section 3, as Turbo FEI gets redeemed against the PCV, I think the amount of circulating FEI should decrease by the same amount as the PCV is decreased. For example, moving from Case 3 to Case 5 in the ETH=$1700 case, PCV drops by 50 M, but so should FEI. Generally speaking, FEI mints and redemptions change the PCV and circulating FEI by the same amounts, so they don’t change the Protocol equity (although it does change the collateralization ratio).
Also, the new FEI would count as protocol-owned, so I don’t think it should be counted as circulating FEI in calculating CR.

Second, I am not sure of the exact amount, but the PCV has some LP tokens and I believe there is around 60M of ETH-FEI in Balancer. These positions are subject to divergence loss (impermanent loss) and thus more exposed to downturns compared to hodling ETH. It would be nice to incorporate this in the stress tests, though I’m not sure if it would change the qualitative results of the analysis.

2 Likes

In Section 3, new FEI drops by 50% since I assumed that only 50% gets redeemed against PCV. Same as PCV

1 Like

The overall analysis present here is quite sound, and these allocations are reasonably well adjusted for risk. Though I do wish to speak out against reinstating TRIBE incentives for borrowing FEI on AAVE, or any other platforms.

TRIBE rewards for borrowing inflate the borrowing rate on that platform and disproportionately attracts arbitrage borrowers. As mentioned by SebVentures here, there is empirical evidence that AAVE borrowing is heavily redeployed into other TRIBE yield programs, rather than any organic usages.

Finally, in the same thread , I have argued in depth why it is more preferable to subsidize borrowers by lowering the interest rate, instead of giving out TRIBE rewards outright. I’d like to raise some of them here again: foster organic usage growth by non-arbitrageurs; equalizing FEI borrowing rates across different platforms, eventually allowing the rollout of an unified policy rate across the entire platform; amongst other reasons to not reinstate TRIBE rewards for borrowing.

2 Likes

After additional analysis, minting additional FEI for Aave is not in consideration at this time. However, extending incentives for borrowing is a topic of its own here: Extending Aave TRIBE Incentives for FEI Borrowing - #2 by nikhil

1 Like

Great analysis @Delia and @Brianna!

I liked the deployment proposal. I agree on Compound, we need to kickoff the market there. Makes sense to also provide TRIBE incentives there?

For the Laa$ allocation, I would suggest that before we move with new deployments on this front, we make a “report” with the results of this 1st phase and think about the lessons learned.

These deployments will be made through OA or with the new governance model (Tribal Council + Pods)?

1 Like

Upon review of the community feedback and analysis updates to existing FEI deployments, and the impact of minting additional FEI with the existing market conditions, we have identified additional optimization steps:

  • recover 10M FEI from Angle protocol deployment
  • reduce Laa$ deployments to 20M

These actions will allocate Tribe DAO up to 104M FEI to deploy across broadly supported initiatives, while adhering to the FEI minting policy recommendations.

Proposed deployment strategies:

Turbo: 40M FEI (estimated 5% APR): the initial deployment for 40M FEI, but with the planned expansion of Turbo to Aribtrum, Tribe DAO should pre-approve additional 10M FEI to be minted as needed

  • Double collateralized
  • Strengthen DAO-2-DAO partnerships
  • Increases Fuse TVL and FEI utility
  • Technical requirements:
    • Nothing new to develop (just whitelisting collaterals & setting their caps)

Compound: 5M FEI

  • Blue chip lending platform
  • Visibility and utility for FEI
  • Technical requirements:
    • Nothing new to develop

Arbitrum: 10M FEI

  • Layer 2 presence and adoption
  • Technical requirements:
    • Need new contracts to interact with the Arbitrum L1ERC20Gateway (L1->L2)
    • Need new contracts to interact with the Arbitrum GatewayRouter (L2->L1)
    • Need new contracts to propagate L1 PCV_CONTROLLER role to interact with contracts on L2 and handle PCV securely

Minting FEI to Convex Fuse Pool: 10M FEI

  • Increases Fuse TVL
  • Increased FEI utility with long tail assets
  • Technical requirements:
    • Nothing new to develop

Laa$ (20M FEI)

  • Currently deployed 50m
  • Generating 5-7% revenue on deployed FEI
  • Expected deployments
    • SYN (estimated 9M)
    • NEAR (estimated 3M)
    • FOX (estimated 3M)
    • POOL (estimated 1M)
    • UMA (estimated 4M)
  • Improved returns and lowered exposure

Volt (12.4M FEI)

  • Tribe DAO incubated product

  • 10M FEI PSM backstop

  • 2.4M FEI for Volt LBP seed their initial liquidity

Balancer Fuse Pool (5M FEI)

  • Strategic partnership with the Balancer DAO

With this update, I propose to move this to Last Call.

2 Likes

I think this list looks great, we should consider sizing down on the convex pool to 5m initially and only raise to 10 if the demand gets filled. Aave/d3 as well if we need more room.

Just wanted to make two notes/clarifications about the Volt proposal for those reading –

The 10m FEI for the PSM backstop will end up in Fuse pool 8, or possibly in Turbo or other Fuse pool that is onboarded in the future, so it remains within the Fuse ecosystem. At launch all Volt PSM deposits will be in Fuse pool 8 (FeiRari dao pool). The VOLT minted may be redeemed by TRIBE DAO in the future for FEI.

The 2.4m FEI received in the VCON token swap (if confirmed by the tribe dao and executed on chain) may be used by the Volt DAO for purposes such as seeding initial liquidity for the VCON token, and as a default (or any remainder if used elsewhere) will also be included in the Peg Stability Module. This FEI will be held by the VOLT DAO in the long term as part of the system surplus.

1 Like