Polygon (MATIC) has seen strong user adoption this year with major DeFi projects like Aave, Sushi, and Curve all deploying there. The time has come for Balancer to join the party. The polygon team has agreed to pledge $5M worth of MATIC as rewards over the course of 8 weeks to be allocated at our discretion. Balancer Labs has taken care of deployment logistics and with the approval of BAL voters we can jump start liquidity with BAL + MATIC incentives. This proposal would allow the Ballers to allocate BAL + MATIC incentives to chosen Balancer pools on polygon. To match polygon’s $5M worth of MATIC, I propose we begin with a 25,000 BAL per week allocation which can be adjusted in the future at the discretion of the Ballers but not reduced below 25,000 until the initial 8 week matching period has completed. These 25,000 BAL would come from the fixed 145,000 weekly issuance, reducing the total L1 incentive to 120,000 rather than increasing BAL inflation.
Motivation
Other polygon deployments have been wildly successful. Recently Sushi has done more volume on polygon than ETH mainnet (check this thread and this tweet). Balancer brings a unique primitive to polygon by being the only flexible AMM. Forthcoming features like the stable swap factory will give us additional competitive advantages on polygon compared to ETH mainnet where similar products exist in the market. Based on my analysis of Sushi’s metrics, I believe we can also expect to exceed fees earned on ETH mainnet after a few weeks go by on polygon.
The Ballers plan to focus on multi-asset pools with higher fees and potentially building partnerships with polygon native projects. Any community input about pool composition or projects worth connecting with is greatly appreciated (#governance in discord).
25,000 is a sizable commitment but the potential opportunity is significant. I strongly believe we can generate more fees per BAL spent on polygon than we can with our current strategy on ETH mainnet.
Specification
If approved, the Ballers will immediately begin designing pools and assigning the BAL + MATIC rewards. As mentioned, the MATIC rewards would last for 8 weeks ($5M total over 8 weeks) and a minimum of 25,000 BAL per week would last for 8 weeks. After that, the Ballers can decide to adjust that number up or down (with community input!).
The goal is to vote this weekend and begin incentives on the 28th - pending any community discussion of course.
Do we have an outline for the specific slots we migrate? I’d like to suggest:
Tier 2: 2x slots (10k BAL total)
Tier 3: 4x slots (10k BAL total)
Tier 4: 5x slots (5k BAL total)
So this would provide us with 11x incentivised Polygon pools. I’ve got some ideas for how we could design from there, something like:
Tier 2: USDC/USDT/DAI stable pool, WBTC/USDC/WETH/WMATIC/BAL
Tier 3: LINK/AAVE/SNX/WETH/BAL/USDC, QI/WMATIC/USDC/miMATIC, BAL/WETH 66/34, YFI/MKR/COMP/UMA/WETH/USDC (?)
Tier 4: some experimental pools? e.g. an oracle pool, some 8-asset pools, etc.
I think we can look at bringing over some of the “defi middle” from ETH too even though there isn’t great volume on Polygon for it yet, it’s a bit of a chicken-and-egg situation. As Polygon gains users, there’ll be retail demand for these tokens, and people will do smaller trades generally on Polygon rather than ETH.
I reckon for Tier 2 & Tier 3 we set fees at slightly less than 0.3% to undercut sushi via the aggregators, then Tier 4 could be used for some higher fee experiments, although I think high fees might be likely to do better on ETH, I could be wrong. On Polygon without UNI-V3 we should be able to better compete for trade volume via the aggregators, whereas on ETH I think we may end up needing to optimise fees towards the higher side as we can’t capture that so easily given Uni’s concentrated liquidity (high fees definitely work better for LPs when your volume is primarily arbs).
The 66/34 BAL/WETH pool gives us somewhere to direct those that want to migrate from the pool on Ethereum, we could go 80/20 again but think it’s nice to have another option for people on Matic.
If we can deploy stable swap pools in time for launch, then I’m good with throwing up a usdc/usdt/dai but I might start it at Tier 3 because the performance of this pool is more of a question mark imo.
As far as including Qi & miMATIC, I like the idea but I think this needs to be coordinated with Qi team. See if they are open to also incentivizing with Qi. Most users that want to farm Qi start at their website, so if we’re incentivizing their coin we should be on their site so their users can find us imo.
As far as pool designs, I think the base should be WETH/USDC/BAL/WMATIC. Unless there’s a good reason I think every pool should include these assets. WMATIC/USDC is top traded pair on polygon, no reason we should not have maximum exposure to that. Adding WETH as another base pair also makes sense. Add BAL because may as well do ourselves a favor and force farmers to hold BAL to make a little more painful to dump it.
Building on that, I like the idea of WETH/USDC/BAL/WMATIC/QUICK/ADDY/AAVE/LINK swap WBTC in for LINK or w/e. This is assuming we get a deal with QUICK for them to list BAL/WETH and we get ADDY to integrate our pools on their site ASAP.
Problem with bringing liquidity for tokens not already on polygon is there’s no other liquidity to arb against. Not completely against it but not sure it makes sense to make a pool for this express purpose. As part of an index I think it could be OK.
If we had to decide today my proposal would be WETH/USDC/WMATIC/BAL 15k BAL rewards 1% fee, and WETH/USDC/WMATIC/BAL/QUICK/ADDY/AAVE/LINK 10k BAL rewards 1% fee. Re-assess the following week after we see what happens.
We shouldn’t cripple ourselves with BAL/WETH style pools as that will still generate low fees with no WMATIC and/or USDC pairing imo.
Reasoning for 1% vs sub 0.3% is because there is more trading activity due to the low gas fees and volatility for polygon native tokens is quite high. We should take advantage of that with 1% fee. Only reason to go 0.25% or lower is to generate high volume numbers, which could be good for PR purposes. I suggest we go with 1% and try to educate the masses as to why earning fees is better than having high volume.
Thanks a lot for the proposal and I generally agree with the post and maybe avoiding pools like BAL/WETH could be a good idea.
Regarding the 1% fee, I would be in favor of something lower. Is the only alternative sub 0.3% or something in the middle is possible? Although is true that there is high volatility 1% seems a high value.
anything is possible. I have made probably 50+ pools on balancer, mostly 8 tokens. I experimented with 0.5%, 1%, and 2% fees. 1% always performed the best. It’s that sweet spot between still getting tagged relatively often with arbitrage but actually earning enough to make it count. You do not meaningfully increase your number of trades with 0.5% to make up for the lost revenue compared to 1%. This is anecdotal and based on my experience of course.
Snatching revenue from arb bots is a huge benefit. Remember that in Polygon GAS fees are essentially zero, so arbitrage will be a big factor there - even bigger than in Mainnet.
Mainly because I don’t see any other good options. They are similar but I think the 4 token one appeals to more people as its simpler with assets users are familiar with. The index is a bit more exotic but that should translate to higher fee earnings. Though I expect both should do quite well.
I’m open to other considerations but I believe we have to include USDC and WMATIC for fee generation. and I don’t see a good argument to not include BAL across the board.
I expect to iterate as time goes on. As we reach out to more projects, get deals done, we can build out more pools. Though they should all be some variation of WETH/USDC/BAL/WMATIC imo. Subject to change as I see more data of course
It just seems kinda limited to me. Consider that you’re a prospective user of a protocol and this is their pitch: “We offer flexible baskets of up to 8 tokens! But each basket must contain precisely these 4 tokens at a minimum.” Doesn’t that design space feel arbitrarily constrained?
If you’re implying that there is some data to show that there’s only one good pool design on Polygon, then I’d say just incentivize the one pool (rather than duplicating it). But I don’t fundamentally believe that there is only one good design.
If I was familiar with more polygon projects I could suggest more tokens. Limiting factor is I don’t know much about these projects yet.
With the understanding that we will be iterating in the future and adding more pools, as an introduction to balancer I think the two pools I’ve suggested make sense.
It might work to split them up somehow to create a few pools but probably doesn’t completely solve your criticisms.
This may be the case, but that was on ethereum where gas was higher, and by going with a 1% fee you’re basically giving up on being a competitive exchange altogether and only serving LPs who want to index (not traders at all).
I also agree with Rab that we should have some variety and minimise incentives for subsets. As Polygon builds up there’ll be liquidity for other tokens building up there, not sure why we wouldn’t want to be involved in getting in early on that. There’s no reason that major defi tokens won’t live on both Eth & Polygon.
my motivation is to generate maximum fees. when we deploy to polygon, we aren’t going to be a destination for traders. any prospect for that happening is far into the future. by making that our primary focus, we’re crippling our fee generation potential.
Balancer is built for LP’s, not traders imo. Balancer being a trading destination is a meme at this point. Maybe that changes, a long time from now
I’m talking about aggregators more than traders direct to Balancer, as an example you can see what Quickswap quickly achieved on Polygon with a straight Uni V2 clone. I’m optimistic that without Uniswap V3 on Polygon we can do some decent trade volume via the aggregators there.
I could be wrong but I don’t think Quickswap got as big as it did by porting over blue chip defi tokens. I think tier list of fee generating tokens goes USDC → WMATIC → polygon natives → WETH → WBTC & defi bluechips
In general, I think this is a good idea. However, I have reservations that I would like to share:
We all need to realize that Balancer is late to the party. Quickswap became so successful because they benefited from the first-mover advantage. SUSHI, for example, does not produce as much volume as QUICK.
I strongly believe that Balancer didn’t get the attention it probably deserved, because it focused all the attention on creating pools that compete directly with protocols that already control much of the market (this is happening now with pools incentives on L1). Ok, you can gnaw a few% of the market here and there but certainly not become a major player. On polygon I repeat, Balancer is late. The market is now saturated. I would gladly suggest other options (deployment on Arbitrum or Fanthom --which does not have any major AMM) rather than pushing for MATIC. Also, please consider that Curve already offers “IL” exposed pools which would mean competing directly with them.
IMO Balancer should offer indexed pools on Polygon (Oracles; top DeFi; AMM (UNI; SUSHI; QUICK; BAL; BNT) 80/20 BTC / ETH and vice versa - these are just examples) and so on but not stable pools or confusing ones as those proposed above. In this way, Balancer would propose an alternative product and gain some market share.
Regarding BAL allocation, I would probably go for 10% of weekly supply (14,500 BAL). This would allow future scenarios including deployments on other L2s if the opportunity arises.
You can find the assets by TVL here - QUICK is a biggie with $500M and SX, DFYN, LINK $50M each. These are tokens that have been bridged from L1, not the Polygon native issuance like Qi, miMATIC or Addy.
In addition to ETH DeFi bluechips, you can also try new BSC blue-chip tokens like bZx, BIFI, AUTO, BANANA (apeswap), FOR (fortube) as another tier 4 pool. Lot of users will be using this with BSC retail now coming onto Polygon. $20M per day via bridges.
