FEI<>WETH Liquidity and strenghtening ties with Fei

Dear Balancer Community,

Greetings from a rainy Sardinia, Italy.


Balancer<>Fei partnership has accelerated significantly over the past few months.

From the launch of FEI V2, which algorithmically manages volatility risk through a custom Investment Pool on Balancer, to the more recent treasury swap.

With the intention of further strengthening collaboration and aligning the interests of both protocols, I propose an incentive initiative for a FEI<>WETH pool on Balancer. This proposal, coordinated with FEI, intends to capture 90% of the liquidity currently deployed on Uniswap (300M$), and transfer it to our Protocol.

This operation would make the FEI<>WETH pool, the second largest pool on Balancer.


I see the importance of consolidating strong and long lasting relationships between DAOs. With this in mind, I consider extremely beneficial seeking and extracting any value that could potentially help the Protocol secure its position in a very competitive market. Today, more than ever, strong partnerships can help us navigate the ever evolving and highly competitive DeFi landscape.

Fei is already a strategic partner, it is our intention to consolidate this cooperation even more.


The current Tier system does not allow the Ballers (and the Liquidity Mining Committee) to make any significant change to BAL allocation, especially to Tier1 pools. These pools, also called CORE, are considered strategic and receive 15,000BAL per week in distributions (the BAL<>WETH pool has two Tier 1 allocations).

Current Tier 1 pools are:

WBTC<>WETH - 50/50 15,000BAL
DAI<>WETH - 60/40 15,000BAL
BAL<>WETH - 80/20 30,000BAL
USDC<>DAI<>USDT 33/33/33 15,000BAL

The introduction of the FEI<>WETH pool would require a redesign and recalibration of the Tier 1 slot system.

This is based on the intention to allocate 6,000BAL tokens and make the FEI<>WETH a Core pool (long term, in principle untouchable, allocation)

With this in mind, I propose a drawing from the following pools:

DAI<>WETH 4000BAL 15,000BAL 11,000BAL
BAL<>WETH 2000BAL 30,000BAL 28,000BAL

Additionally, I propose a further allocation of 1000BAL to a new pool TRIBE<>WETH 80/20, the deal is based on a two months renewable allocation and should be coincentivised by both protocols. This allocation falls within the LM competency, but helps you have a better understanding of where we want to go with this.

The negotiation involved a preliminary conversation regarding swap fee split (50/50) between the two Protocols to help alleviate our BAL emission. This has to be discussed in more depth and will not be activated immediately. Regardless, I wanted to give you the full picture of my thinking.

So, data from the past fourteen days show the following (FEI<>WETH Uniswap trading pair Uniswap Info):

52,285$ Average Daily Swap Fees
365,995$ Week
19,031,740$ Year

The proposed BAL distribution (6000BAL per week) would represent an expenditure of 92,000$ p/w at current market valuations.


We want to strengthen the relationship with Fei. The question is, at what cost?

There are a few point I want to bring to your attention:

  1. The pool on Uniswap is mainly controlled by Fei, this means most of the BAL distributed will go to their treasury once deployed on Balancer. Under present conditions (Fei treasury owning 200,000BAL) it will take just over two years for them to double their voting power. This is something that goes against the principle of voting distribution and decentralisation. We need to think long term and the implications this would mean for us as a community;
  2. In financial terms, we are operating at a significant loss. This is in line with other pools at the moment. A revenue split on the pair has been proposed to alleviate this expenditure and to make the pool sustainable for us. Further discussions will take place around this;
  3. Drawing from Tier 1 pools follows the principle of a long term allocation strategy. Making the FEI<>WETH a core pool will protect it from sudden reductions in allocations, even if market conditions evolve. If we use discretionary BALs (those under the discretion of the LM Committee), we are effectively reducing our freedom to allocate BALs to where we think is best at a given time.
  4. A 6000BAL allocation would allow a smooth transition without impacting our current distribution schedule. APRs would suffer a reduction but not dramatic.
  5. The 6000BAL is an initial allocation, emission could be increased over time.

A short SWOT analysis of the deal:

Strengths: We are strengthening ties with one major Protocol. We are strongly aligned and this proposal should be seen as a natural evolution on an already strong collaboration;

Weaknesses: We are going to give Fei (single entity) an increased voting power over time;

Opportunities: Balancer will be the new home of a significant amount of liquidity and we are betting heavily on the success of the FEI stablecoin. If FEI gains sufficient adoption and market share, Balancer would benefit immensely in both reputation (main source of liquidity) and market position;

Threats: Known unknowns and a more broad market competition could make this deal an expensive exercise. Time will tell.


I want to bring you data and numbers so you can make an informed decision on something very important.

We are friends with the Fei Community. This deal should be seen as a part of the more general strategy of collaboration between Fei and Balancer. As a member of the Community, and with the future of Balancer and the DAO at heart, I negotiated conditions that would appear sustainable and sufficiently efficient with the sole intention of generating long-term sustainability.

I am personally very excited about this, it is a huge shift for both Protocols. But these shifts could bring amazing opportunities for both communities.

Looking forward to your comments.


so they are going to manually send us 50% of the swap fees every week or something? don’t think I’ve heard of such an agreement before.

otherwise, seems good to me.


Thank you for taking the time to dive into the numbers, Andrea.

I would suggest waiting a few weeks before the community makes any hard commitments involving shifting Tier 1 allocations, as we’ll soon need to transition WETH/DAI and staBAL3 allocations to boosted pools.


Love the proposal and furthering our partnership with Fei.

I can see the argument about concentration of voting power, this should however not threaten our protocol given how much other stakeholders hold in comparison.

If ever any contentious vote that benefits Fei and not Balancer takes place I’m confident all the currently apathetic BAL holders (who mainly don’t vote because they agree with the few who do) will step up to protect Balancer.

@markus how long would we have to wait in your opinion? Can’t we already discuss the future allocations when Boosted Pools are live? We should also consider the period for transitioning LM from staBAL3 to Boosted pools. So my main point is what are the advantages of waiting versus the opportunity cost of not doing this already if governance agrees it’s a good idea?


With respect to reducing Tier 1 allocations, I would suggest drawing all 6,000 BAL from the BAL-WETH pool and leaving the WETH-stable pool alone. The latter is probably our most valuable liquidity, and BAL-WETH is arguably overpaid right now.

I also think 50/50 fee sharing is unnecessary. The protocol already has a mechanism for extracting a percentage of swap fees, and if DAO members feel that it is too low to be profitable then they should consider raising it. Off-chain swap fee sharing feels unsustainable to me, and the 50/50 split seems to shift this proposal far in favor of Balancer and not offer much for Fei. I’d like to see both parties benefit from any agreement.


BAL/WETH has several T2 slots that can be given to FEI/WETH, so there is really no governance vote required if the entire allocation is drawn from there. (I agree that is the best course of action)


There is a case to be made around the idea of making FEI the biggest USDpegged<>WETH pool on Balancer.

Cooperation also means pushing a partner’s product (FEI) vs some other’s (DAI) hence the reduction on rewards for the DAI pool.

BAL is our Governance token and a robust APR incentivises token holding and reinvestment under current conditions (no utility apart from Governance voting). Further APR reduction could trigger selling pressure on our token.

I remain in favour of making DAI the best candidate for this.

Please consider that these are preliminary conversations and nothing has been decided on this front.
Nevertheless, I would encourage you to think long term and on a way to assure the survival of the protocol. How can any pool be sustainable (for us) with a .5M$ expenditure per month? I would be interested on having your view on this.

No split would do the opposite in my view: the FEI<>WETH on Uniswap is getting zero incentives in UNI tokens. So for them makes total sense moving the liquidity over. We are happy to have them with us and to build an incentive program around the pool. But I would see it fair for us to reduce the delta at an absolute minimum if possible.


WETH/DAI liquidity is far more valuable in terms of revenue generation and usefulness compared to BAL/WETH. The protocol’s earnings would go up significantly if BAL/WETH incentives were diverted to WETH/DAI, which is bullish for BAL token price. I also believe most BAL/WETH liquidity is very sticky at this point - taking the yield to 10% from 20% wouldn’t have a material impact on TVL imo.

It is true that no pool is sustainable for us today - you have to take the long view of Fernando where other projects build on top of us and increase our trading volume to the point where these pools will turn profitable. Of course if they agree to send us 50% of the fees we’d be silly to say no, but to me it’s pretty awkward to do that and they would be totally justified to push back.

Getting $300M TVL of weth/fei for 6k BAL is already a great score for us and you deserve credit for putting that deal together.


First of all, I agree with @DavisRamsey: Bringing that liquidity over is a huge win regardless of any revenue sharing. And if this proposal were to boil down to a 6,000 BAL weekly expenditure to retain $300M in FEI-WETH liquidity, I think that’s well worth it.

With that being said, are we 100% committed to FEI-WETH? I think a more powerful architecture might be a stable coin metapool: FEI-3pool. Benefits are:

  • Deep liquidity between FEI and any other major stable coin
  • Cheap multihop to WETH or any other token, basically a FEI-WETH pool for “free”
  • No IL risk for LPs (barring FEI peg risk)
  • Our 3pool absorbs a bunch of new liquidity because any FEI-3pool LP would have to own 3pool tokens as a prerequisite

I think the benefits above could be summarized into a general principle for pool deisgn: always plug a new token in at the most price-efficient juncture. To put it another way, the most size-insensitive invariant (e.g., StableSwap) that can be used, should be used. This, combined with cheap multihop, reduces price impact for large orders on multiple token paths. For example, consider hypothetical WETH->FEI and DAI->FEI trades in each scenario:

Architecture: FEI-WETH (plus WETH-3pool)

Trade: WETH->FEI (single hop)

  • WETH->FEI hop is size-sensitive (Weighted invariant)

Trade: DAI->3pool->WETH->FEI (multihop)

  • DAI->3pool hop is size-insensitive (Stable invariant)
  • 3pool->WETH hop is size-sensitive (Weighted invariant)
  • WETH->FEI hop is size-sensitive (Weighted invariant)


  • WETH->FEI is size-sensitive, as it must be because there is no price peg between those tokens
  • DAI->FEI is doubly size-sensitive despite the price peg between the tokens: there is price impact at both the 3pool->WETH hop and also the WETH->FEI hop

Architecture: FEI-3pool (plus WETH-3pool)

Trade: WETH->3pool->FEI (multihop)

  • WETH->3pool hop is size-sensitive (Weighted invariant)
  • 3pool->FEI hop is size-insensitive (Stable invariant)

Trade: DAI->3pool->FEI (multihop)

  • DAI->3pool hop is size-insensitive (Stable invariant)
  • 3pool->FEI hop is size-insensitive (Stable invariant)


  • WETH->FEI is size-sensitive, as it must be because there is no price peg between those tokens
  • DAI->FEI is size-insensitive because there is a price peg between those tokens

Of course, this architecture would have to wait at least about a month for our new 3pool release.


Well, in this case let’s make a quick pool:

From which of the existing pools shall we get liquidity from?

  • DAI<>WETH & BAL<>WETH (Andrea’s proposal)
  • BAL<>WETH (Rab’s proposal)

0 voters

Thank you for sharing all this @rabmarut, much appreciated.

I believe we should move with little steps towards products that fit the narrative.
While I understand your proposal (and I think we should work towards that), at present time lets KISS (Keep It Stupid and Simple).

Let’s offer them what they asked for. Times for deeper conversations could be established in a later stage (happy to support you).


I’m perfectly happy with that outcome if it helps to solidify the deal in the near term - as I mentioned, the liquidity is a big win independent of form. But I think we should be mindful about architectural optimizations on a long time horizon and try pitching them to teams who might be especially receptive. Fei might be receptive, or they may very well have their own reasons for preferring FEI-WETH and that’s fine too.


I think the nature of the Fei<>Balancer partnership is a longer term one. I sincerely believe that there’s a lot that these two protocols can achieve together and with other partners like Rari and Ondo (as examples).

As such, I don’t think it’s unreasonable to take a moment and consider @rabmarut’s point. Migrations suck and 1 month is not a long time when one considers the longer term approach of this partnership. Plus, I think 1 month is a conservative estimate.

Also, curious to know, what the Tribe and @joey think about a FEI/3pool and how that would fit into the Fei ecosystem given that it’ll help with bot FEI/WETH and FEI/DAI/USDC/USDT liquidity while reducing IL for LPs.


I have mixed feelings about the choice between DAI/WETH and BAL/WETH when it comes to raising the 6k incentives. Yes, DAI/WETH is one of the larger generating pools on mainnet, however I do think some people will be very annoyed that the APR for BAL/WETH is dropping. A lot of the liquidity is likely sticky and the APR would only be going from 22% ->18% so not a massive change, but it is currently the only real option to get paid for squirreling your BAL away. It seemed like some mixture with little impact to both was a good path to take.

That being said, it sounds like there are a lot of upcoming changes required for boosted pools and the LM program needs an overhauling, so maybe it makes sense to wait if FEI is onboard with that. It will also allow for a discussion about the optimal pool construction.


Thank you Andrea for putting the work into this proposal. This looks really valuable for Balancer.

I propose a further allocation of 1000BAL to a new pool TRIBE<>WETH 80/20, the deal is based on a two month renewable allocation

This would be done via the flexible BAL allocation?

This has to be discussed in more depth and will not be activated immediately .

Can this clarification be made simultaneously to the pool architecture? Based on the rabs comment it seems like having a FEI-3pool is more efficient.

I feel like the time to clarify these topics with FEI is in line with waiting for the new 3pool release.

Where to get the 6k BAL from though seems like a tough choice. I lean more towards getting the BAL from the BAL-WETH pool.


Hi everyone, I’m a core dev @ Fei Protocol, author of the matching thread on the Tribe forums.

Just wanted to comment on a few things, also feel free to @ me if you have any questions about Fei.

I don’t think it’s a problem for decentralization if our DAO owns a large amount of BAL and captures emissions from your LM program. Sure, in your snapshot/DAO votes it will look like a large holder decides, but behind the scenes, it is all TRIBE holders that proxy-vote through our protocol-owned BAL. So in fact, I’d argue that this is good for decentralization, because you get exposed to the voting of a whole new population (that may not hold BAL in the first place).

We are doing meta-governance with INDEX already :slight_smile:

What you say about trade routing and having price-efficient junctures is very interesting, definitely something we should consider to design a strategy of liquidity deployment in the Balancer ecosystem.

We also understand liquidity against [DAI, USDC, USDT] is important, in fact, we heavily incentivize our Curve FEI metapool, it has over 230M$ liquidity.

That being said, we feel pretty strongly about keeping 0 exposure to custodial assets as part of the PCV, so the chance that our protocol deposits assets in the staBAL3 or [staBAL3, FEI] pool is very low, because of USDC and USDT exposure.

WETH/FEI sounds like a better option for now.

I’m very excited about boosted pools.

If you can make a [bbaUSD, bbaFEI] pool that provide liquidity between FEI and [DAI, USDC, USDT], while also deploying these 4 assets on Aave, that would be very cool, and we should incentivize it. Our protocol can provide liquidity in the [WETH, FEI] pool, and we can rent the [FEI, custodial stables] part of the liquidity by co-incentivizing this boosted pool with TRIBE.

If it were possible to write custom plugins for boosted pools, to deposit in Fuse for instance instead of Aave, I’m sure we’d be interested in implementing it, too.

The potential of what we can do with Balancer is pretty exciting, but let’s keep this proposal simple for a first partnership. We can start with the WETH/FEI and WETH/TRIBE pools, and build up from there :slight_smile:

I can’t speak on behalf of the whole Tribe DAO, but 6,000 BAL for WETH/FEI long-term and 1,000 for WETH/TRIBE sounds good to me. I’ll continue to draft the proposal in our forums.


Thanks, @Eswak!

That being said, we feel pretty strongly about keeping 0 exposure to custodial assets as part of the PCV, so the chance that our protocol deposits assets in the staBAL3 or [staBAL3, FEI] pool is very low, because of USDC and USDT exposure.

Perfect example of “having their own reasons for preferring FEI-WETH” - no problem.

Our protocol can provide liquidity in the [WETH, FEI] pool, and we can rent the [FEI, custodial stables] part of the liquidity by co-incentivizing this boosted pool with TRIBE.

This sounds like a great compromise between the two options.

If it were possible to write custom plugins for boosted pools, to deposit in Fuse for instance instead of Aave, I’m sure we’d be interested in implementing it, too.

This is definitely possible and definitely something we are interested in doing, but in due time. I think we’d like to see how the Aave Boosted Pools play out as a proof of concept first. Most likely, I would say let’s wait a month or two after launch to gather data and firm up integrations before expanding the boosted pool offering.


I really like this idea! Unfortunately we won’t have capacity for it in Q1 but we can support your dev efforts if you have capacity for it. Alternatively, I’m curious if the Balancer Grants sees this as something they could put up a RFP for. If this does become an RFP, would you guys be interested in a joint grant @Eswak?

1 Like

Sure :slight_smile: @DioDionysos should have contacted @zekraken to set this up.


Hi @Eswak, indeed they reached out. We have an ongoing chat on the subject. Let’s see where that takes us :muscle:

1 Like