Contributors: @Marcus and @mendesfabio (BLabs); @danielmk , @naly (Beethoven-X); @ZenDragon ,@Mike_B , @Tritium and @Xeonus (Maxis)

Summary

The launch of Balancer v3 presents the DAO with an opportunity to optimize its fee model and core operational framework, positioning it to dominate the yield-bearing market and enhance long-term revenue generation and redistribution for LPs, veBAL holders, and the DAO.

Upon approval, this BIP will implement the following changes:

Balancer v3

• Initial Launch: Mainnet and Gnosis chain
• Launch Strategy: Primary focus on boosted pools technology without base pairing provisioning (e.g. Aave boosted waUSDC:waUSDT instead of standard USDC:USDT)
• Timeline:
  • Tentative launch end of Q4 2024
  • Activation of veBAL gauges and core pools in Q1 2025
  • Expansion to more chains thereafter
• Simplified Fee Model:
  • Universal fees: 10% on yield, 50% on swaps (enforced via vault¹)
  • Core pools require 50% yield-bearing tokens and a minimum $100k TVL²
  • Non-core pools fee split: 82.5% to veBAL holders, 17.5% to DAO
  • Core pools fee split: 70% as voting incentives, 12.5% to veBAL, 17.5% to DAO

Balancer v2

• Maintaining current fee structure: 50% on yield, 50% on swaps
• Core pool framework [2] adjustments:
  • Discontinuing non-core pool fee redirect to core pools (as introduced per [BIP-457])³
  • Same fee split as v3: 70% as voting incentives, 12.5% to veBAL, 17.5% to DAO
  • No other change to the existing framework
• Non-core pool framework adjustments:
  • Same fee split as v3: 82.5% to veBAL, 17.5% to DAO⁴

Introduction

The launch of Balancer v3 represents a major milestone for the DeFi ecosystem. v3 is designed to propel on-chain adoption forward by significantly simplifying custom pool creation, enhancing the customizability and extension of proven pool types, and introducing the first version of boosted pools that channel all underlying liquidity into lending markets for optimized LP yield accrual.

Over the past year, Balancer Labs, in collaboration with numerous DAO contributors, has devoted extensive efforts to delivering one of the most flexible and innovative DeFi products to date. This journey began with a complete UI and UX overhaul via the ZEN interface, reached a significant milestone with the launch of CoWAMM—a novel solution for LVR mitigation—and now culminates in Balancer v3, a comprehensive reengineering of the platform’s underlying technology.

As outlined by Fernando, in his post on the protocol's future, Balancer v3 strategically focuses on key priorities such as fungible liquidity, stablecoins, long-tail, and yield-bearing liquidity, and introduces new customizable options for extending existing pool types through hooks.

Alongside these technical advancements, Balancer v3 provides a unique opportunity to refine the protocol’s core mechanics, particularly its fee structure and pool framework—critical elements for long-term DAO sustainability. The following temperature check proposes adjustments to the fee model and core pool framework, set to take effect with the release of Balancer v3.

Simplifying the Protocol Fee Model

Fee Model on Balancer v3

The development of custom pools provides the DAO with a unique opportunity to integrate innovative pool types and advanced mechanisms for fee accrual and redistribution.

Currently, a significant portion of Balancer v2's revenue is derived from a specialized fee-capture system on liquidity pools hosting yield-bearing assets like Liquid Staked Tokens (LSTs) and Liquid Reward Tokens (LRTs). With the launch of 100% Boosted Pools in v3, this established yield-capture advantage will be extended, allowing revenue generation from vanilla, non-yield-bearing tokens by deploying them into yield-generating markets. This approach enables any token with an external yield market to be composed and “transformed” into a yield-bearing asset, unlocking new revenue streams for the DAO - a system that no other DEX in DeFi has.

Although the technology has demonstrated its product-market fit for yield-bearing liquidity, the existing 50% yield-capture fee has faced resistance from LST protocols and liquidity providers, limiting the DAO’s ability to capture this market segment fully. Given the critical role of yield capture in revenue generation, the DAO must reassess the current fee model to remain competitive and dominate this market.

Thus, our strategic focus for v3 centers on accelerating the adoption of boosted pools and solidifying the DAO’s position within the LST and LRT market. Additionally, we aim to drive continuous product innovation to continue to unlock unique revenue streams through hooks and custom pool logic.

To support this, we propose implementing two key changes:

1. Reducing yield fees to 10% to increase adoption of yield-bearing liquidity on Balancer v3
2. Strengthening veBAL holder benefits with a simplified fee model⁵

The veBAL system and BAL token emissions, being immutably established in Balancer v2, will continue utilizing the core pool framework and gauge system. This ensures alignment between token emissions and performance while maintaining veBAL's long-term relevance. Importantly, veBAL holders will benefit from concurrent revenue streams from both v2 and v3.

Fee Structure Overview:

• 10% fees on yield-bearing asset yields
• 50% on collected swap fees

Pool Classifications:

• Core pools: as per new revised core pool framework
• Non-core pools: All other pools

Fee Distribution:

• Core pools: 70% voting incentives⁶, 12.5% veBAL, 17.5% DAO
• Non-core pools: 82.5% veBAL, 17.5% DAO

This model achieves three key objectives:

1. Streamline veBAL holder benefits through an 82.5% total fee share
2. Aligns BAL emissions with high-performing pools
3. Maintains sustainable DAO SP (service provider) funding and operations at 17.5% fee share.

Modified Fee Model on Balancer v2

Balancer v2's success as a governance token, and LST / Yield Bearing liquidity hub in 2023-2024 informs our approach to fee model optimization. While preserving the core framework's integrity, we're proposing targeted adjustments to align with v3's launch:

• Maintaining current fee rates: 50% yield fees, 50% swap fees
• Updating non-core pool fee distribution:
  • Previous: Redirected 50% of fees to voting incentives
  • New: 82.5% to veBAL, 17.5% to DAO
• Updating core pool fee distribution to be in line with v3:
  • 70% as voting incentives, 12.5% to veBAL, 17.5% to DAO

This modification serves to:

• Sustain momentum in successful core pools
• Direct swap volume success to veBAL holders
• Encourage v3 migration through lower yield fees (10% on v3 vs 50% on v2)

Revised Core Pool Framework

For a pool to be eligible for core pool status, certain criteria must be met. We propose to make it more clear what will be a core pool in the future:

Composition requirements

• Minimum 50% yield-bearing or boosted tokens for pool types such as weighted or composable stable pools
• 80/20 weighted pools that utilize Balancer as their primarily liquidity hub (e.g. RDNT:WETH 80/20 pool on Arbitrum). Protocols need to apply to core pool status for such pool types⁷
• Must maintain a minimum $100k TVL

Technical Requirements

• No yield fee exemption
• Fee settings delegated to Balancer governance
• If no protocol fee settings are present or managed by another entity, the pool can only become core pool if an alternative protocol fee setting is setup in place⁸

Token Requirements

• All tokens must have verified smart contracts
• No tokens with transfer restrictions or rebasing mechanics

Core Pool list maintenance

• Core pool status is evaluated on a bi-weekly basis before fee sweeps by automated checks managed by Balancer Maxis [5]
• New pool types or compositions not described within this framework need governance approval to be added to the core pool list
• The DAO shall review this framework and propose adjustments through DAO governance

Impact of Fee Change

Based on the proposed fee changes, we have conducted backtesting on v2 data and ran simulations to assess the potential impact of v3 adoption.

Balancer v2

The revised fee model adjustments to Balancer v2 would result in the following change in fee distribution to key stakeholders:

• veBAL: increase of 46.7% of revenue share compared to the current model
• Voting incentives (to veBAL and vlAURA holders): 30.3% decrease
• DAO: no change

Monthly fee distribution on v2: actual vs proposed (average of August, September and October 2024)

Balancer v3

While the reduction of yield fees from 50% to 10% in v3 represents a significant change, our analysis [1] demonstrates how optimized yield fee utilization will maintain robust protocol revenue. We conducted detailed simulations using 5 out of our top 10 pools by TVL, focusing on two key pool categories⁹:

1. Stable USD vs boosted stable pools on v3
2. Composable stable LST/LRT vs Aave or Morpho boosted pools

Monthly income to liquidity providers comparing Balancer v2 vs v3

Monthly yield fees to the fee collector

Key Findings:

• LPs will experience a 407% increase in income from boosted yield and reduced yield fees
• Protocol yield fee revenue maintains 56% of current levels even with reduced yield fees
• Enhanced yield capture from new boosted stable pools effectively compensates for most of the yield fee reduction to 10% assuming the same TVL levels

This analysis shows that while yield fees are decreasing, the protocol's ability to capture and optimize yield fees through new pool types ensures sustainable revenue generation. The 407% increase in LP yield income demonstrates the effectiveness of this strategic shift toward yield optimization.

Launch timeline

The proposed Balancer v3 deployment schedule:

• Q4 2024: Initial launch on Mainnet and Gnosis chain with strategic pool selection
• Q1 2025: veBAL gauge system activation with new fee model implementation
• Q1-Q2 2025: Multi-chain expansion phase

Technical Specification

A detailed technical specification execution plan will be posted as a separate BIP outlining the modified fee collection mechanism on v2 and the implementation for v3 in early 2025

Conclusion

The launch of Balancer v3 marks a strategic evolution in the protocol’s development. It brings significant enhancements to custom pool creation and extends the flexibility of proven pool types while preserving the strong foundation and market leadership established with v2.

By refining the yield fee capture structure and streamlining core functionality, v3 is poised to deepen yield-bearing liquidity, boost returns for liquidity providers, increase benefits for veBAL holders, and foster sustainable long-term growth for the protocol.

This proposal strikes a balance between innovation and stability, setting the stage for Balancer’s continued leadership in DeFi infrastructure. We invite the community to support these enhancements as we embark on this exciting new chapter in Balancer’s journey.

References

[1] Fee distribution modeling
[2] Balancer v2 core pool framework
[3] Historical core pool dashboard
[4] Protocol Fee Dune Dashboard
[5] Balancer v2: Automatic core pool list

Footnotes

1. As with v2 yield fee can be set to be exempt. For stable pools no rate provider = no yield fees ↩

2. Same globally for Balancer v2 and v3, see [BIP-457] and the automatically bi-weekly updated core-pool list ↩

3. A core pool's incentives for a given round will be capped at 70% fees it earned over that round. This effectively removes the payment of non-core-pool swap fees to core pools as specified for L2s in BIP-19 and for Mainnet in BIP-457) ↩

4. As opposed to a non-core pool redirect where 50% of non-core pool fees were recycled as core pool voting incentives to boost BAL emissions to core pools. ↩

5. Based on Balancer v2 backtesting, voting incentives on v3 might first decrease but that assumes the same TVL numbers. The biggest fee earners on Balancer v3 are LPs. By decreasing yield-fees, we are incentivizing LPing greatly with boosted pools which then results in more core pool voting incentives based on higher TVL targets. ↩

6. Core pool incentives [3], [4] are only placed if the minimum incentive thresholds are met. Currently, these are $500 for the veBAL market and $800 for vlAura market to hit minimum voting thresholds. These values are set by the Maxis in coordination with Aura contributors in the fee allocator constants as per [BIP-457]. ↩

7. Check the ALCX/ETH proposal as a good example ↩

8. CoWAMM v1 pools can’t become core pools as there is no protocol fee setting present. Gyroscope v1 pools hosted on Balancer v2 can only become core pools if the Balancer DAO protocol fee collector is set as target ↩

9. Comparison data between v2 and v3 can be accessed here ↩