Summary
BAL is a governance token that has been successfully used to define important parameters of the Balancer ecosystem. The community believes that there is further room for improving the tokenomics of BAL and some sort of staking/locking should be implemented. Staking of BPT has already been proposed and approved by the community in the past[1]. The current proposal contains details of a concrete implementation of locking BPT.
To avoid reinventing the wheel and spending additional time with non-core development, Curve’s tokenomics[2] seems an obvious fit for Balancer Protocol as well. It has been battle-tested with billions of dollars and attracted a thriving ecosystem around it.
Core proposal
Curve’s ve (vote-escrowed) system will be used as an all-around solution for Balancer protocol tokenomics, solving at once a few outstanding problems the Balancer ecosystem currently has:
- BAL minting is not automated and the inflation schedule is not yet locked forever. This reduces predictability and trust in the overall system.
- BAL holders do not have any direct power over how BAL liquidity mining gets distributed.
- Lack of a mechanism for distribution of protocol level fees, which have recently been approved by governance[3].
Instead of pure BAL, BPT of the 80/20 BAL/ETH pool[4] will be locked into veBAL, similar to how CRV can be locked into veCRV. This has the big advantage of keeping BAL liquid as well as setting a precedent for other teams to do the same with their 80/20 Balancer pools. Alternatives to veBAL have been considered, like vebptBAL. The simplicity of veBAL — even though what is locked is not pure BAL — has however been the preferred option.
veBAL and governance power
All votes, onchain or on snapshot, will be done considering veBAL balances instead of BAL balances as happens today. This ensures long-term alignment as only users locking BPT will have a say in Balancer’s governance.
To get veBAL, anyone will be able to lock BPT of the 80/20 BAL/ETH pool for any amount of time between 1 week and 1 year. Notice this is a change from the max duration of 4 years initially proposed in [1]. This is proposed to allow for an eventual migration to a new governance system with a wait time of 1 year instead of 4. Of course this migration can only happen if the veBAL governance approves it.
After locking the 80/20 BAL/ETH BPT, the user will have a non-zero balance of veBAL and will be able to help govern Balancer protocol, also benefiting from other parts of the new tokenomics system as described in more detail below.
New proposed BAL inflation schedule
The current BAL supply today (as of Jan 27th) is 47,470,000 BAL. This includes BAL that is being vested by early stakeholders, the fundraising fund and the ecosystem fund. Please refer to the BAL launch medium post[5] for more details.
Since the launch of BAL, the inflation has been a constant 145,000 BAL per week. To change that and make the Balancer ecosystem more sustainable, a new inflation schedule is proposed. Every 4 years the inflation should be halved, with gradual steps every year starting one year after the launch of the new tokenomics system. This would mean a final BAL supply of about 94,000,000 BAL. A more detailed calculation can be seen on this public spreadsheet[6].
This new inflation schedule will be immutable:
Onchain Gauge System to determine Liquidity Mining distribution
Currently, liquidity mining is done via a tier system approved by the community[7]. The distribution is fixed for tier 1 slots (it can only be changed via governance vote) totaling 60k BAL. The remaining slots are allocated by the Liquidity Mining committee, which is part of the Partnerships subDAO[8]. BAL token holders have no direct say in which pools receive BAL from liquidity mining.
In the new system, however, all liquidity mining BAL minted will be distributed through the gauge system. Gauges are contracts that allow LPs (liquidity providers) to stake their BPT (Balancer Pool Token) and periodically claim BAL from liquidity mining.
The amount received by each LP will depend on:
- how much the pool they are providing liquidity to receives
- what share of the pool’s liquidity they have
- the boost applied to their LP share based on how much veBAL they hold[9]
Each pool receives a share of the total BAL minted every week. That share is defined by how much veBAL voted for that pool.
Gauge types
There are different gauge types that are used in the ve system, each type gets a fixed percentage of the BAL supply minted weekly. Within each gauge type there can be many gauges that share the amount of BAL sent to that gauge type. The following gauge types and weights are proposed:
- Liquidity Mining committee: 10% (14,500 BAL)
- veBAL: 10% (14,500 BAL)
- Ethereum mainnet pools: 56% (81,200 BAL)
- Polygon pools: 17% (24,650 BAL)
- Arbitrum pools: 7% (10,150 BAL)
New gauge types can be added and the weights/percentages of all gauge types can be changed by governance. These changes are however not expected to happen frequently.
Gauge type 1) is an allowance for the Liquidity Mining committee to grant LM to strategic partnerships for Balancer protocol with the expectation that long term these partners will accrue veBAL to ensure they get LM through their own gauge. The LM committee should not accumulate BAL, that is, it commits to sending any unused BAL for any given week to the DAO treasury (not later than 7 days from receiving it).
Gauge type 2) is BAL that will be distributed to veBAL holders, similarly to LM that is distributed today to LPs of the 80/20 BAL/ETH pool. It is meant to keep locking 80/20 BPT attractive as it helps veBAL holders avoid dilution and compensates for the impermanent loss risk of LPing.
Gauges 3, 4 and 5 will be LM that is distributed to mainnet, Polygon and Arbitrum according to the veBAL voting power each pool gauge receives from veBAL holders.
Initial gauges
In order to avoid the governance overhead of voting on adding dozens of initial gauges, all pools that will be receiving liquidity mining at the moment of system activation will be allowlisted. This however does NOT mean that they will have an initial percentage of LM. This will be defined by veBAL voting power.
Protocol revenue distribution
75% of protocol revenues collected by the protocol fee collector[10] will be distributed to veBAL holders. The other 25% of the fees will be kept by the DAO treasury as a reserve. This has been suggested by many thought leaders like Hasu, on his New Mental Model for DeFi Treasuries[11].
Given Balancer protocol is permissionless and meant to have thousands of pools, fees are going to be collected in potentially thousands of different tokens. Some type of consolidation will be necessary for the fees to be claimable by veBAL holders in a meaningful, gas-efficient way. The fee consolidation prior to distribution will initially be handled by the treasury subDAO, but should be replaced in the near future by a fully decentralized smart contract implementation, akin to Curve’s burner contracts[12].
Upgrading the system
Like with Curve, the system will not be upgradeable. In case the veBAL governance decides for a new system, the way to implement it would be to abandon the current system and direct minted BAL to the new system. This can only be done with a 1-year time delay, which ensures that anyone locking BPT for the longest duration possible will have the opportunity to withdraw their BPT before the minter points elsewhere.
The inflation schedule however will NOT be upgradeable and will be fixed forever.
Benefits
Beyond increasing the overall predictability of Balancer Protocol and the BAL governance token as mentioned above, there are many other benefits worth mentioning:
- Long-term alignment. By locking BPT, token holders will be encouraged to support Balancer over the long-term instead of speculating short-term.
- Plug&play compatibility with Curve’s ecosystem. Curve has achieved tremendous success with a thriving ecosystem built around their ve tokenomics. By using the same system we make it trivial for teams like Yearn, Element, Convex and so many others to become part of Balancer’s ecosystem too.
- Developer time/effort saved for core Balancer work. Tokenomics is super important but it is not what makes Balancer (or any other protocol for that matter) unique: precious development time should be focused on pool primitives and making Balancer easier to integrate with. Only this will help Balancer achieve the vision of becoming the leading liquidity platform for others to build on top.
- Motivation for other DAOs to take a position in Balancer beyond just token-swaps. It has become common for DAOs and protocols to purchase CRV in the open market to make sure they have a say in the Curve Wars. By having protocols aligned with Balancer longer term we expect utility and usage of Balancer to increase as well.
Alternatives considered
- Launching our own staking contracts (which are ready and audited but don’t solve all the problems that this proposal does).
- Using mStable rewards distributor or a variation of it.
- Keeping status-quo.
Risk assessment/mitigation
- Immutability of the system will likely be a problem at some point in the distant future: It’s impossible to be at the same time predictable and upgradeable, so the trade-off chosen was that the system itself cannot be changed, but veBAL governance can choose to direct BAL inflation to a new system with a minimum 1-year delay. This is not expected to happen any time soon, but as things change hyper fast in this space, this system will likely not be the best option in say 5 years from now.
- Platforms built on top (like Convex on Curve) can take over control with central points of failure. We expect veBAL governance to consider single point of failures for projects that apply to have their smart contracts be allowlisted for locking BPT. Ideally only projects which let token holders vote directly with their voting power (instead of delegating all their power to a multisig) will be approved.
Implementation
The main goal of the implementation is to change as little as possible from the battle-tested original Curve implementation. There could be some minor necessary technical changes that are necessary to adapt it to Balancer. The essence of this proposal or the parameters laid out however should not be changed.
Voting
This proposal will be posted on the forum for 1 week prior to voting. Voting will then be open on Snapshot for three days. This vote will be a single choice vote. You may vote on the Proposal by selecting “Yes, let’s do it” or “No, this is not the way”.
Thanks to everyone who contributed with feedback and suggestions, also thank you to Curve for having created this great tokenomics system and also to the Yearn community[13] for inspiration for this proposal!
I’m curious to hear everyone’s thoughts and if you think anything proposed can be improved.
References:
- Staking of BPT for Economic and Governance Benefits
- Curve’s veCRV documentation
- Activate the Protocol Fee
- 80/20 BAL/ETH pool
- BAL is live!
- Proposed supply inflation schedule
- Balancer V2 Liquidity Mining Program
- Partnerships subDAO and the Liquidity Mining committee
- Curve’s LP boost calculation
- Protocol fee collector
- New Mental Model for DeFi Treasuries, by Hasu
- Curve burner contracts and fees documentation
- YIP-65: Evolving YFI Tokenomics
at some point you may want to ask yourself “what amount of liquidity is too much liquidity?”, but because the pool is 80/20, it should be ok.