FIP66 - Convex Integration into Curve & CVX Purchase

FIP-XX: CRV Gauge Proposal for Fei Liquidity Pools with Convex Support
FIP-XY: Proposal to acquire CVX for Fei Protocol PCV

Status: DRAFT
Authors: @arcology; @DioDionysos; @fei.saver; @meertitan; @hawthorne
With input by: @eswak; @afirebrand; @incoomer

Executive Summary

This proposal will, in aggregate:

  1. Request Convex to vote on Curve in favor of a CRV gauge for the FEI-3CRV metapool.

  2. Request Convex to vote on Curve in favor of raising A factor in the FEI-3CRV metapool to 500.

  3. Request Convex to vote on Curve in favor of a CRV gauge for a new “Fei PCV 1” pool that will contain stable assets held in the PCV, and where the protocol will be able to supply liquidity and farm CRV/CVX rewards.

  4. Acquire up to 1,000,000 CVX tokens (approx $30.0M) using the ETH from Compound and Aave deposits, to be held in the Fei Protocol PCV.

  5. Reduce TribalChief AP rewards on the FEI-3CRV metapool from 1000 AP to 0 AP over one month and redistribute TRIBE emissions to the following:

  6. Incentivize the FEI-3CRV gauge via Votium with 400 AP (~600k TRIBE/epoch of 2 weeks). With the 1:3 ROI of bribing, this should distribute the equivalent of >1200 AP of rewards on the pool in the form of CRV and CVX tokens. .

  7. Incentivize the “Fei PCV 1” pool with 350 AP (~500k TRIBE/epoch of 2 weeks). The protocol could deploy liquidity there, and capture a ROI of approx 1:3 for DAO revenue.

  8. Incentivize the existing D3 pool via Votium with 250 AP (~350k TRIBE/epoch of 2 weeks).


Fei Protocol is a decentralized, scalable algorithmic stablecoin collateralized by the ‘Protocol Controlled Value (PCV)’. It aims to be the stablecoin for DeFi, with upcoming improvements to the core protocol through its V2 framework, including 1:1 redeemability as the mechanism to hold its dollar peg; a more flexible and robust PCV portfolio management; as well as emergent business adjacencies such as its ‘Liquidity as a Service (LaaS)’ offering in partnership with Ondo Finance and its PCV management system in partnership with Balancer. As of this writing, the PCV holds approximately $1.2B in assets with a collateralization ratio of ~350%.

Fei’s smart contracts have been audited by OpenZeppelin and Consensys Diligence.


Fei Protocol is currently at a high growth trajectory with a robust rollout of new features in its V2 framework, and a potential merger in the works with Rari Capital. Nevertheless, the core product of Fei is its stablecoin and integrations are necessary for success and mass adoption.

Curve is a major player for stablecoin exchanges and for ensuring liquidity to other stablecoin assets and additional yield strategies. Defi Pulse ranks Curve at #2 with over $15B TVL. Fei Protocol deployed a Factory metapool on Curve under its TribalChief rewards contract and now has about $230M TVL and a daily volume of ~$3.5M.

Two Governance proposals for listing Fei on Curve did not meet quorum due to voter apathy. At this juncture, Convex has graciously offered to assist using their veCRV voting power to assist Fei Protocol to acquire a gauge, which allows the Fei pool to acquire CRV rewards, which would attract more liquidity into the pool as well as allowing for more advanced strategies (developed by other protocols) for yield using FEI.

As a participant in the Curve ecosystem, Convex acts as a de facto governance module for curve through its cvxcrv product, which allows vlCVX holders to vote to direct weekly Curve inflation to gauges and vote within the CurveDao. Convex Protocol has distributed over 400m in revenue to its holders over a 6 month period.

Votium is a gauge marketplace built on top of Convex that allows interested parties to incentivize players in the ecosystem to vote to increase their gauge weight, driving $3.35 for every $1 spent on incentives — an especially potent strategy for Fei Protocol’s immense PCV.

To drive more revenue (through inflation) to the proposed FEI gauge through Convex’s Flywheel, defend its gauge weight on weekly votes, and create deep liquidity for FEI through gauge incentivisation, Fei proposes to purchase and lock CVX tokens to boost its PCV and have “skin in the game” with one of defi’s most important primitives.

More information can be found with Eswak’s proposal to Enter the Curve Wars.



This proposal will begin a voting post and ‘town-hall’ type thread on Convex Governance for a minimum period of 7 days. We will then propose a snapshot vote for CVX holders to vote on whether they support using the veCRV holdings of Convex’s DAO to proxy vote on Curve.

Once this snapshot vote passes, a new sCIP and CIP (snapshot signal vote and then DAO vote) will be created for the purpose of activating a CRV gauge for the FEI-3CRV pool and the Fei PCV 1 pool. The amplifier (A) parameter of the FEI-3CRV pool will also be increased from 10 to 500 to optimize for swaps/slippage given the deep liquidity in the pool.

From Fei Protocol’s perspective this strategy will override the existing TribalChief rewards into the current FEI-3CRV pool as a far superior use of capital.

The protocol currently deploys 1000 AP on incentivizing the FEI-3CRV pool which amounts to about 750k TRIBE per week. The proposal will gradually reduce these rewards to 0 AP, and replace this reward stream by Votium bribes at a 3:1 ROI. This will drive more rewards than are currently distributed on the Fei metapool, and we will also be able to incentivize 2 more pools that increase FEI liquidity and strengthen our partnerships with FRAX and Alchemix.

Subsequently the rewards will enter one of the following streams using Convex’s Votium platform for incentivization:

  1. Fei will use approx. 600k TRIBE / 400 AP per Votium epoch to incentivize the existing FEI-3CRV metapool, given that a gauge can be acquired. However, Fei will not be able to deploy liquidity there as there is a risk avoidance with centralized stablecoins making up a majority of 3CRV. It is calculated that the ROI will be about 1:3.
  2. To counteract the effects of #1 above, Fei proposes to create a new ‘D3’ pool of decentralized stables comprising FEI, DAI and LUSD. Fei has proposals put forward to acquire (or has already acquired) positions in both DAI and LUSD. The protocol aims to deploy a liquidity of at least $250M due to these positions. Fei will use approx. 500k TRIBE / 350 AP per Votium epoch to incentivize the creation and maintenance of a gauge. It is calculated that the ROI will be about 1:3.
  3. The existing D3 pool will continue to be incentivized by deploying approx. 350k TRIBE / 200 AP per Votium epoch.


This proposal will spend up to about 3% of the PCV, drawing from the Compound and Aave ETH reserves to acquire 200,000 / 500,000 / 1M (based on DAO vote) CVX tokens of an approximate value of up to $30M as of this writing.

The method to acquire CVX to be finalized, but a preference to be via an OTC swap due to the high value amount of tokens being exchanged. In the scenario that the lowest amount is chosen, the required tokens can be purchased at spot price from the market.

CVX can be locked for 16 weeks to be able to vote on gauge weights, and drive even more CRV/CVX incentives to our desired pools. In the last Convex snapshot vote for gauge weights, 22.3M CVX voted.

The vote-locked CVX could be used as a diplomatic tool (we’d usually vote for pools with FEI in them, but we could also offer to vote for a partner protocol’s pool from time to time).

Voting Options

  1. In Favor of FIP-XX and FIP-XY with 1M CVX purchase
  2. In Favor of FIP-XX and FIP-XY with 500k CVX purchase
  3. In Favor of FIP-XX and FIP-XY with 200k CVX purchase
  4. In Favor of FIP-XX, Against FIP-XY
  5. Against all (do nothing)

The vote will be a single-choice voting. If no option receives the majority of votes (>50%), the amount of CVX purchased will be determined by the dot-product of the total number of each votes for each option by the list of options: A(1,000,000) + B(500,000) + C(200,000) + D(0) + E(0). If an option earns the majority, the winning option will be put to an on-chain vote.

Would you Support this Proposal?
  • A - In Favor of FIP-XX and FIP-XY with 1M CVX purchase
  • B - In Favor of FIP-XX and FIP-XY with 500k CVX purchase
  • C - In Favor of FIP-XX and FIP-XY with 200k CVX purchase
  • D - In Favor of FIP-XX, Against FIP-XY
  • E - Against all (do nothing)

0 voters


I made a spreadsheet to calculate the effective APY that a DAO earns by voting with CVX. I believe purchasing CVX is a good strategy for Fei Protocol.

Why is CVX underpriced

Like CRV, CVX is a token that is worth much more when held by DAOs than individuals. The effective APY of vote-locked CVX that is held by a DAO is currently as high as 128%, and is projected to remain above 50% if the current price levels persist. On the other hand, APY of CVX that is held by a small individual investor is 5.62%. And yet much of CVX is held by individuals. This is the fundamental reason why CVX is underpriced, and why it represents a good buying opportunity for Fei Protocol.

Fei should bribe, but it’s not enough

Bribing is a low risk opportunity for Fei Protocol to make a significant profit and should definitely be used. However, the opportunity may not last. There is a scarce amount of veCRV, and as more DAOs compete to bribe them, Fei Protocol will eventually end up getting slightly more than a dollar per each dollar spent on bribes. In fact, the current ROI of 3 will immediately be diluted as soon as Fei Protocol starts bribing.

On the other hand, competition to bribe will increase the APY of individual CVX holders and boost CVX price. That is why the DAO should move now and buy CVX. Bribing is like renting CVX - the DAO pays CVX holders to use CVX for two weeks, and then return them. Renting is extremely profitable now at an ROI of 3, but that means the price of renting CVX will soon go up. The way to capitalize on this is to buy CVX.

Risks of buying CVX

The PCV will be exposed to CVX, which is highly correlated with CRV. Given that Curve is a DeFi blue chip that is not included in DPI, being exposed to CRV is aligned with the current PCV management strategy.

Another major risk is that more CVX can be vote locked and dilute the APY. However, this is likely to happen when the demand for CVX grows and thus CVX increases in price. Also, as more CVX gets locked, the APY of CVX that has not been locked (such as staked CVX) will increase as a response and dampen the increase in locked CVX.

Bump - with Fei V2 and the Rari merger passed, let’s revisit this proposal.

All comments welcome.

@Eswak this seems to dovetail with your votium bribe proposal - please clarify whether your proposal is contained within this one. Thanks.

Hi @arcology,

Yes, to be clear my proposal is repeating the parameters you described here (step 8).

I think we should work on / snapshot this proposal in multiple steps, because if this proposal stays monolithic, it will be hard to agree on everything and it could take forever. It’s also hard to digest such a big proposal.

I suggest these snapshots on our side :

  • CVX purchase + voting strategy
  • Bribe on d3pool
  • Bribe on FEI metapool
  • Bribe on “FEI PCV 1” pool (don’t think we should do it anymore, see below)

Snapshots to be worked on the Convex side (Do one of you co-authors own this ? It doesn’t seem like any input is needed from the Tribe for these, so you could start discussions in the Convex governance for these) :

  • Raise A factor in the FEI metapool to 500
  • Gauge for FEI metapool

ps: in the last 2 weeks since this has been posted, I think I changed my mind on something: a “FEI PCV 1” pool just to farm CRV/CVX is probably a bad idea, we can deploy stablecoins on Balancer instead and build something much more exciting potentially, their new boosted pools are very interesting, and the more we concentrate liquidities in Balancer, the more they have synergies together. I’d also like to think more about how we can integrate the metapool & d3pool LP tokens more into Fuse, and maybe offer TRIBE rewards to users that stake LP tokens but the protocol would keep CRV/CVX, or something like that.

Bumping this thread to move it forward. I think we can proceed to last call and execute 3 snapshots:

  1. CVX purchase + voting strategy
  2. d3 pool bribing
  3. FEI metapool bribing

Per @Eswak 's suggestion, lets start threads on Convex for the Fei metapool gauge and A factor change.


I kicked off a gauge vote:

I am not in favor of market buying CVX at this stage.

Very in favor of ramping up bribes and incentives to the FeiRari pool. Now that the protocol benefits from the TVL in those pools by re-staking in Convex, I believe this should be a primary emissions strategy

Assigned FIP-66

Can you please elaborate on reasons for not doing so? Some of us discussed the merits of this and we agreed that acquiring CVX for the treasury is a good strategic move that will pay off. It may be even better when everything is at a discount right now.

I’m open to acquiring some but want to make sure PCV is in liquid tokens to the extent possible. If we don’t need to vote escrow our CVX I’d be more open to it. Smaller sizes can be cool too but 30M is excessive imo.

We can snapshot the lower end of the spectrum as a start. 200k CVX can be readily acquired using market prices without an OTC strategy and assuming $38.30, will set us back about $7.67M. We can start building a position from there.

Moving the discussion to last call!

7M still too much for my comfort if we’re doing this in a public market buy. I’d love to get more tribe holder opinions here

We’d incur 5.2% slippage with a $7M market buy, which is not a financially sound choice imo. I’m fine with primarily bribing and using PCV to farm d3. I think the best choice for purchasing CVX is a <$10M OTC.

Thats fair.

We will delay a snapshot for a CVX purchase for the time being, and allow more discussions to be had.


I think a Balancer LBP for 2-3m FEI would be noncontroversial if we bought over a couple weeks

1 Like

Snapshots are now up for bribing 2 pools using Votium!
They will begin Jan 17 and end Jan 24.

Bribe the FEI-3CRV pool

EDIT: D3 Pool is already bribed under FIP-59. FIP-66.2 Snapshot retracted.

I like this. Does anyone from the Convex community or others here in favor of this as a starting point?