We’re now heading into the third week of liquidity mining on V2. The early pools are growing well, but as debates have been taking place among the Ballers Committee plus wider governance, a recurring challenge has been that there are pools we’d like to experiment with incentivising, where the Tier 3 BAL allocation (2,500 BAL per week) is too great of an allocation or risk.
This is slowing some of the progress in migrating LM rewards from V1 to V2.
PROPOSED CHANGE
This proposal is a simple update to the current Tier system which was previously approved by governance, and was as follows:
Tier 1 (T1): 15,000 BAL (4 slots, total of 60k BAL per week) - Pools where strong liquidity is paramount, e.g. for acting as hubs by connecting liquidity in order routing.
Tier 2 (T2) : 5,000 BAL (10 slots, total of 50k BAL per week) - Strategic pools, e.g. having strong potential for high volume given high liquidity.
Tier 3 (T3) : 2,500 BAL (14 slots, total of 35k BAL per week) - Promising pools, e.g. containing newly launched tokens or small cap projects (unproven potential for high volume).
This proposal passing would replace the previous tiers with the following Tier system:
There will be 4 different liquidity mining tiers for eligible pools:
Tier 1 (T1): 15,000 BAL (4 slots, total of 60k BAL per week) - Pools where strong liquidity is paramount, e.g. for acting as hubs by connecting liquidity in order routing.
Tier 2 (T2) : 5,000 BAL (8 slots, total of 40k BAL per week) - Strategic pools, e.g. having strong potential for high volume given high liquidity.
Tier 3 (T3) : 2,500 BAL (10 slots, total of 25k BAL per week) - Promising pools, e.g. mid caps, and experimental pools that have proven significant traction in Tier 4.
Tier 4 (T4) : 1,000 BAL (20 slots, total of 20k BAL per week) - Experimental pools, e.g. many-token pools, newly launched/small cap projects, early stage partnerships.
In practice this means that 2x Tier 2 slots and 4x Tier 3 slots will be used to create a new Tier with 20 slots. The committee will then typically add new pools into Tier 4, measure success and then bump those which are successful up, and similarly move pools down from Tier 3 should it make strategic sense.
BENEFITS
There are a number of significant benefits members of the committee can envisage from this additional tier:
The committee can run more liquidity mining experiments with lower risk. As it stands, the minimum we can allocate is equivalent to ~$75,000 USD per week, which is a large amount for a small project or pool structure which may not for some reason attract sufficient liquidity & trade volume.
More pools can be incentivised overall. This should help us to migrate more pools from L1, and mean we upset fewer L1 LPs who suddenly lose all their rewards.
The committee can more easily setup LM partnerships with other smaller cap projects, where dual LM incentives are offered.
Tier 3 pools which for some reason no longer warrant the 2,500 BAL/week can be downgraded, rather than completely excluded.
Please discuss any thoughts/ideas/concerns here on the forum, we’ll try to merge any thoughts from the Governance channel on Discord here too. All other aspects of the Liquidity Mining program would remain unchanged. If there seems to be general agreement that the new tier makes sense, we’ll then look to push it to a Snapshot vote within the next ~week.
Do you agree that an additional Tier 4 makes sense, as described?
Thank you for writing this up. Fully support this as it allows us to partner with small projects early in their token’s lifecycle and experiment with some more creative pool configurations.
Sorry about the resolution; reposted with larger font and resolution. I did include a few indexes, but I guess I need to better understand how many more indexes we think will gain traction. I’m wondering if we have too many indexes whether many people are going to want to own all the assets in the index as opposed to pairs.
all depends on your opinions and assumptions about LP behavior I suppose. I’m of the opinion that an index + meta pool will generate higher fees than a rando 50/50 eth/defi coin. I also believe BAL rewards APY on a defi index will be approximately equal to BAL rewards APY on rando 50/50 eth/defi pool. Large defi coins are so highly correlated that I think it all comes down to what you can earn the best yield on.
Thus my view that instead of going down the list of adding link/weth, snx/weth, comp/weth etc we just skip straight to the index meta pools. Because once we add those 50/50 pools, then do indexes and realize I’m right, we’ll be faced with the choice of continuing to waste BAL every week or removing it from 50/50 pool and giving to more indexes.
But it is mostly theory and if you say you disagree with one of my assumptions you’d be totally justified.
I think the way forward is to do a mixture and prove the theory out properly. I have concerns about whether people will go through the friction of acquiring all the tokens to deposit (or eat a lot of slippage), and also think that people are more sensitive to which tokens they’re exposed to than you expect, but I could be wrong.
Worst case is we do a mixture and then one or the other gets reduced over time with the other pool type(s) winning out. If the 3-8 asset pools are wildly out-performing the 2-asset pools it’ll be easy to see and there won’t then be any real objection to demoting the 50/50s in favour of more of them IMO.
Being able to route single-asset deposits through the SOR would be game changing I think.
Anyway, this is probably a little off topic for this thread, does anyone have any concerns on the revised slots / tier structure generally, or are we good to move this to a Snapshot vote asap?
I believe LP’s are advanced sophisticated investors that have conviction in a few projects and do not have much interest in diversifying out of those crypto assets into a larger basket of crypto assets. I did include three targeted basket pools in my spreadsheet including 33DPI33WBTC33WETH, 50DPI50WETH, and 13UMA13YFI13UNI13LINK13AAVE13MKR13SNX13WETH, so maybe the truth as always is something in between.
Thanks for this proposal, @bakamoto20 . I fully agree Balancer could use a new, lower stakes tier for experimentation.
One concern I have is that increasing the number of slots so much so fast might hurt Ballers ability to properly assess each decision. The proposed structure increases the number of slots at the discretion of Ballers from 24 to 38. I think it’s safe to assume pools in T2 would rotate less frequently, so maybe we’re really talking about an increase from 14 to 30. Personally, I think there’s a high chance that the increased overhead when at this early stage would lead to either voter apathy or decision paralysis. So, at least for now, I would suggest the following distribution:
T1: 4 slots x 15,000 BAL = 60k BAL
T2: 10 slots x 5,000 BAL = 50k BAL
T3: 12 slots x 2,500 BAL = 30k BAL
T4: 5 slots x 1,000 BAL = 5k BAL
In addition to that, since we’re looking into deploying to layer 2, I would like to suggest we take this opportunity to expand the scope of this proposal and include explicit permission from BAL holders for Ballers to allocate LM slots to pools on any of the official deployments. This will allow us to move fast and incentivize liquidity on layer 2 soon after deployment, without the need for an additional governance vote. Thoughts?
I think one rationale for large number of T4 slots is so we can find out what works without a large commitment then promote the top performers. but perhaps not needed if our guiding philosophy is moving over the top v1 pools by assets
Thanks for the feedback markus. In coming up with the larger T4, the idea was that we can move faster by starting pools in T4, assessing traction on a pool-by-pool basis then upgrade based on the data/performance.
Part of the challenge we’ve had with the Tier 3 minimum has been reaching agreement on pools, whereas I think if we had the “low stakes” tier available, it’d be easier to get approvals agreed among the Ballers for adding to T4, and then the data proves itself on what gets bumped up. That’s my thinking, anyway!
I can see that perhaps once the tiers are close to full management could become a bit more difficult, but I suspect that will be the case regardless of how many slots we have. If anything I’d suspect having the larger number of T4 slots will make it easier to make decisions (more ideas get included, even if not as heavily incentivised as people would perhaps like for their specifi cidea), but I may be wrong.
Nice, I like this direction.
Agree on both fronts:
Ballers should have the autonomy to allocate across both L1 and L2.
T4 is a good idea, but a reduced number of slots seems better to manage (not too much overhead on analyzing candidate pools), as those slots could/should rotate faster than the ones from T2 and T3.
5 slots for T4 instead of 20 seems to be a good compromise.
Thanks for your thoughts followthechain, I disagree though. I think the 5 slots on T4 isn’t enough: the idea is we can do less analysis for a T4 slot and experiment more, generating our own data to prove out what works/doesn’t.
Approving a T4 slot shouldn’t be overly contentious if we have 20 of them, with 5 I fear we’ll have the same issues we’ve been having to date finding agreement on what should be included, and lots of assets will go without any incentives what-so-ever, which costs us longer tail opportunies.
My sense is that Balancer’s success will come from providing good liquidity between non-traditional pairs, due to the vault structure and its impact on gas for multi-hop. Making that a success means being able to have liquidity for a lot of tokens, and I think the current 3-tier system is too concentrated in a few pools to bootstrap that (and that the 5 additional T4 slots doesn’t go far enough to help).
This is especially the case when we’re aiming to maintain V1 80/20 BAL APYs (going to mean multiple T2 slots to be able to achieve that for BAL), and matching V1 rewards for the AAVE safety module (I personally believe this is too much BAL for an 80/20 pool, but I do understand the importance of the partnership).
As a final comment, I’m concerned if we don’t have enough pools with incentives, we’ll create a lot of bad faith with V1 LPs who see their rewards dwindle over time and have no where to migrate. That would be likely to on some level lead to BAL dumping & LPs migrating liquidity to other dex’s that’d be harder to win back.
As summary for anyone not following the discussions in the governance channel on Discord, there seems to be enough traction to move forward with this proposal with a couple of changes in the number of slots:
T1: 4 slots x 15,000 BAL = 60k BAL
T2: 9 slots x 5,000 BAL = 45k BAL
T3: 12 slots x 2,500 BAL = 30k BAL
T4: 10 slots x 1,000 BAL = 10k BAL
Hi everyone, as markus stated we’ve moved forward with proposing the 4 tiers as outlined, please vote before Tuesday at 00:00 UTC here: Snapshot Vote Page
Excited to get this moving forward and (hopefully!) work on some interesting T4 pool experiments over the coming weeks should the proposal be approved by governance.
