FIP-70: Let's get BALsy

gm

With our 4M$ BAL treasury swap executed several weeks ago, the buybacks streaming 2-3M$ per week through Balancer Liquidity Boostrapping pools, and the 100M$ FEI->LUSD auction also concluded using Balancer, now is the time to interface more with the Balancer protocol. Now is the time to deploy some PCV there. Let’s show the world that the Tribe has big BALs.

I have been working on code to be able to deploy PCV on Balancer, but we still have to decide as a community what to deploy and under which parameters. This thread intends to serve as a basis to discuss the initial PCV deployments on Balancer.


Summary:
Create pools and deploy liquidity from the PCV and DAO treasury to Balancer.

The pools will be owned by the Fei DAO Timelock, so our DAO will be able to change pool parameters (e.g. fees, enable the pool to act as an oracle, or pause trading) on each of these pools.

Motivation:
Balancer is an innovative, generic, programmable AMM that allows inequal weights for tokens in a pool, among other things. Weights of 80/20 for instance can be chosen instead of 50/50, to reduce impermanent loss or increase exposure to one of the underlying assets. The 80/20 split is used by numerous pools on Balancer already. See the Balancer documentation of weighted pools.
Balancer supports stable pools (like Curve), and meta-stable pools (that concentrate liquidity like Curve, but works with different-price assets by concentrating liquidity around “oracle” prices).

Balancer, as a DAO, can also co-incentivize pools with BAL rewards. BAL-incentivized pools will attract more user funds than if we incentivize alone with TRIBE, like we are doing now with the TribalChief.

Balancer will be an integral part of Fei v2, it is a very flexible infrastructure for AMM and portfolio management, and is very synergistic with our goals here at Fei Protocol. Strengthening our relationship with the Balancer team/ecosystem is a valuable long-term play in my opinion.

BAL the token also has a lot of long-term potential, the DAO already recognized it by doing an OTC.

Proposal:
Deploy liquidity as following :

  • 80% BAL / 20% WETH : deploy the 200,000 BAL previously obtained by OTC (~4M$), with matching WETH (~1M$).
    • Our BAL are currently sitting idle on the FeiDAOTimelock
    • We will earn ~20% APR in BAL rewards if we provide liquidity
  • 80% TRIBE / 20% WETH : bootstrap the pool with 20M TRIBE (~20M$) from treasury + matching WETH (~5M$) from PCV
    • Liquidity for TRIBE against ETH would de-corelate TRIBE from FEI a bit more.
    • Exposure is 80% TRIBE, so it’s almost neutral for the treasury vs holding plain TRIBE.
    • If the IDO funds are removed (FIP-46), additional liquidity will be useful.
    • This pool could be co-incentivized with BAL and FEI. :handshake:
  • 50% WETH / 50% FEI : our protocol currently holds 300M$ of FEI-ETH liquidity on Uniswap v2. We would move 90% of this liquidity out of Uniswap, over to Balancer. Balancer currently incentivize the DAI/WETH pool with 15,000 BAL per week, and the DAI/WETH has around 110M$ liquidity. We have the opportunity to go big on Balancer and be the biggest stablecoin there. This will naturally attract more FEI pairs on Balancer, and if the Ballers incentivize our pool with rewards comparable to the DAI pool, we could capture a significant stream of outgoing BAL from their LM program (currently 145,000 BAL/week). In doing so, the protocol would build significant voting power in the Balancer DAO over time, and be able to use it to improve integrations of FEI and tokens from partner protocols. :exploding_head:

That would already be a bold first move. But we can open more pools once the relation is established.

10 Likes

Thanks @Eswak for this very interesting proposal.

I’d like to say on behalf of Balancer Labs (one of many actors in the Balancer ecosystem) that we love it. I think your suggestions take advantage of the unique features Balancer Protocol has such as 80/20 pools.

Regarding doubling down on the BAL exposure, we would be super happy to see such an important partner like you (with all the innovation we are building together, with our V2s so connected) holding more and more voting power in Balancer.

I’m happy to help answer any questions the Fei community might have about this proposal or other ways we can work together beyond all that’s already ongoing/planned.

4 Likes

gm @Eswak and Tribe family,

My name is Kia, I’m a product manager at BLabs. I have been a huge fan of the Fei ecosystem and a participant since Genesis.

It’s amazing to see this proposal!
In short, I want to express that there are lots of synergies between the two communities and we’re here to support you guys will all your AMM needs for a long lasting partnership. :handshake:

To expand, we’re thrilled to be working on a new version of our Managed Pools which will be the home of Fei v2. Given the move in Fei v2, it is very beneficial to make Balancer the premier home of liquidity for FEI and TRIBE to increase maximum efficiency.

To expand further, Balancer is an AMM platform for DAOs and DeFi protocols. We are a DAO-to-DAO protocol and our mission is to continue to create the most flexible AMM with the largest design space for other protocols to build on top of. You guys have been shipping innovative products that we are very excited about, such as the Fei<>Ondo LaaS and Fei–>Fuse liquidity service. As you guys continue to build products that will establish FEI as a dominant and DeFi native stablecoin, we are here to help you build exactly the right type of AMM pools that contribute FEI’s success.

Looking forward to a long lasting partnership and to working with you all more closely!

5 Likes

In favor of all this! I’d love to see the WETH/FEI pool use a slightly more ETH weighted pool like 60/40 (same as WETH/DAI) or 70/30.

3 Likes

Im a huge fan of balancer and 100% in favor of strengthening ties. I support all of these measures to some degree but I would like to discuss the third one a bit more

in deciding where to deploy pcv I think we need to strike a balance between earning yield that can be directly used in protocol operations (e.g. earning ETH and stables) versus earning yield in the form of gov tokens that will presumably hodl’d long term (e.g. BAL/AAVE/COMP/FLX/LQTY/TOKE/LDO/…)

I think migrating the FEI-ETH liquidity from uniswap to balancer would be an improvement from the current situation. however I also wonder if some portion of that massive $140M chunk of ETH should be deployed somewhere else, earning tokens that can be used for protocol operations. earning yield that we wouldnt care about selling (we dont want to have to sell any BAL obvs)

there’s a very good chance that the long term strategic value of these BAL rewards would exceed the short term tactical value of earning something else like stETH. but I would like to raise this point because I havent heard it discussed very much so far. I think we should be a bit more intentional with our high level yield strategy because, the faster fei grows, the sooner we will have to make decisions about what to do with the yields that we’ve earned when they need to be used for protocol operations

1 Like

That is a very good idea. I’ll use 65/35 to meet you in the middle in the rest of the post. I think we should do it.

That is definitely a concern, let’s look at the numbers here. Maybe you can produce a more detailed analysis than me, but I’ll start :

Amount of fees

  • Looking at the Uniswap v2 pair, we are doing ~15M$ of volume per day. At 0.3% fee, that is ~300k$ of revenue per week (we own most of the liquidity).
  • I’m assuming trading volume will remain constant when we move to Balancer. I have no idea how we can model that to know whether this is true, but at first glance I think people trade FEI “because they have to”, and not “because it’s on Uniswap”, so wherever the liquidity is, trades will happen.

Protocol fees

  • Uniswap-v2 could in the future charge protocol fees (16.67%), but currently does not.
  • Balancer currently does not charge protocol fees, but on the Balancer forum post about this partnership, the ballers suggested a fee split in the future, potentially as high as 50%. Under this scenario, we would be earning 150k$ less fees per week, and earn instead almost all of the 6000 BAL allocation (90k$ per week at current price), so we’d be losing money. Protocol fees won’t be active at first, so these 90k$/week will be pure bonus at first, but as @storm puts it, we don’t want to earn significantly less “hard assets” to earn more assets which we can’t sell. If the fee split does not exceed what will eventually come on Uniswap (i.e. 0.05% for protocol and 0.25% for LPs, i.e. 16.67%) we’d be giving 50k$ of fees per week to Balancer in ETH, and receive 90k$ of BAL.

Impermanent loss

  • The Uniswap pool is 50% ETH / 50% FEI, so if ETH price moves 300%, we get an impermanent loss of 20%.
  • In a 65% WETH / 35% FEI pool, on a similar ETH price move, we’d only get 16.5% of impermanent loss.
  • Assuming this happens over 1 year on a pool with 300M$ liquidity, that 3.5% difference is worth 10M$, or 200k$/week. Having an overweight-ETH pool is a very strong advantage to avoid leaking PCV through impermanent loss.

Additional points
The ballers also proposed 1,000 BAL/week on our new 80% TRIBE / 20% WETH pool. Pools with similar incentives on Balancer attracted anywhere between 2M$ and 40M$ to the pool. This incentive is not very high, but it is not their governance token, so it is already a pretty strong gesture of good will.

We can do very cool things like a Fuse plugin where users deposit their TRIBE/WETH bpt token in Fuse, and earn dual incentive on it in BAL and TRIBE. This would be like a supercharged fTRIBE-8 incentive that we are doing now, but where the TRIBE deposited on Fuse also provide liquidity on Balancer, and get dual incentives :grinning:

Personal thoughts

It is a big change for both of our protocols, so perhaps it would be better to get there gradually.

On the Balancer side, this would represent a Core pool allocation, so it is a long term commitment, and a large one. We would be the 2nd largest pool there. On our side, it would involve moving the main source of FEI liquidity owned by the protocol.

The Balancer LM currently allocates 15,000 BAL to WETH/DAI, and 2,500 BAL to WETH/MKR. Their current proposal allocates 6,000 BAL to WETH/FEI, and 1,000 BAL to WETH/TRIBE. I think this is a good start, and we should migrate, provided that their protocol fees does not get activated too soon. We should negociate lower fee sharings, like we negociated with Rari on Fuse. In the longer term, after we have a significant chunk of PCV there, we want to become the dominant stablecoin in the Balancer ecosystem. Balancer also supports metastable pools, so we can redesign some of our Curve plans to be on Balancer, too. We can create a strong “d3pool” on Balancer, we can have a FEI metapool, etc. This should reflect in the BAL LM allocations, and FEI/TRIBE incentives should equal/exceed DAI/MKR incentives, because it is going to be a true DAO to DAO partnership.

We are earning a lot in trading fees on Uniswap. That makes me want to keep Uniswap liquidity at first, and allocate ETH from Aave/Compound to start on Balancer. We have over 500M$ on Aave and Compound, maybe we could take 65M$ out to seed a WETH/FEI pool on Balancer with 100M$ liquidity. On the Balancer side, they can start with smaller incentives that do not completely change their plans, and after we have more real-world feedback on how trading volume and impermanent loss behave, we can re-discuss the big migration ?

1 Like

I know governance is busy lately with the merge, so I won’t propose a snapshot soon, but I’d like to use this baseline for the following discussions. If there are no objections / good ideas to improve this plan, that’s what I’d like to submit to snapshot :

  • 80% BAL / 20% WETH : we deploy the OTC’d BAL with matching WETH
    • Earned BAL rewards will be compounded by re-depositing in the pool
  • 80% TRIBE / 20% WETH : we deploy 20M TRIBE from treasury + matching WETH.
    • Balancer will offer 1000 BAL/week of incentives
    • We will allocate 50 AP to offer dual-incentives on this pool.
  • 70% WETH / 30% FEI : move 90% liquidity out of Uniswap v2, over to Balancer.
    • Balancer will offer 6000 BAL/week of incentives
    • We will allocate 100 AP to offer dual-incentives on this pool.
  • Pools will be owned by the OA Timelock, so we can snapshot changes to fees and have the OA change them. The pools will have 0.3% fees initially.
  • The TRIBE that the protocol will earn (because it will also be providing liquidity in the pools and earn BAL and TRIBE incentives) will be sent back to the LM rewards dripper.

I dont think creating more TRIBE liquidity makes sense until FIP46 is resolved

the other initiatives sound great

while we’re at it, what do you think about transferring the DPI-FEI liquidity from sushi to balancer? network effects will make it so that each additional liquidity pool on balancer adds more and more value

2 Likes

Pools should be owned by an admin wrapper so the DAO can control as well. We use a similar pattern elsewhere.

Agree with @storm that we should wait for FIP-46 and move DPI-FEI as well.

I also think TRIBE from protocol should go to treasury

1 Like

Hi guys,

Reviving this thread now that the merge is done and I’m back from vacation :slight_smile: let’s get the code for this done !!!

Our 200,000 BAL tokens have already been moved to Balancer in late December (FIP-33b).

Let’s move this proposal to last call under number number FIP-70 & snapshot the following :

  • 70% WETH / 30% FEI
    • Move 50% liquidity out of Uniswap v2 (we currently have ~250M$), over to Balancer
    • Move the other 50% one week later (to make sure Chainlink & other oracles don’t break on the sudden liquidity migration)
    • Pool will have 0.3% fee initially, and will be owned by an admin wrapper contract
    • DAO and OA will have the rights to change the pool’s parameters

Let’s have the TRIBE/WETH pool & incentives discussions separately, as this is related to FIP-46 & needs more discussions in general (we now distribute rewards through Fuse).

Voting will be a simple YES/More discussions needed.

1 Like

Snapshot is live : Snapshot