Balancer: Programmable Liquidity and the V3 Architecture Explained

What Is Balancer?

Balancer stands out as one of the original programmable automated market makers (AMMs) in DeFi. Launched on Ethereum in 2020, it disrupted the market by moving beyond the standard two-asset 50/50 pools used by constant-product AMMs like Uniswap V2.

Instead, Balancer introduced flexible liquidity pools capable of holding up to eight different tokens at custom weight ratios, such as 80/20 or 20/20/20/20/20. This design transformed Balancer from a simple trading venue into a powerful platform for portfolio index funds, stablecoin pools, boosted yield pools, and highly customised liquidity strategies.

In 2025, Balancer successfully completed its transition to the V3 architecture. This ground-up rewrite centralises pool logic inside a singleton Vault, introduces programmable hooks, adds first-class support for yield-bearing tokens, and standardises 100% boosted pools.

In this guide, we explore how Balancer V3 works, explain the mechanics of veBAL governance and BAL token economics, review the V2 security incident from late 2025, and highlight the essential risks every liquidity provider (LP) must understand before engaging with the protocol.

The Singleton Vault Architecture

The singleton Vault is the most distinctive feature of Balancer. It originated in V2 and received further refinement in V3.

Rather than deploying a separate smart contract for every liquidity pool, Balancer routes all swaps, deposits, and withdrawals through one central Vault contract. This Vault holds the assets for every pool across the protocol.

Individual pools register with the Vault and provide their own custom math, the specific pricing curve that defines how swaps execute. The Vault itself manages token accounting, fee collection, and flash loans.

This centralised architecture delivers several key benefits. First, it significantly improves gas efficiency for multi-hop trades, because the protocol rebalances tokens inside the Vault without transferring them between separate pool contracts.

Secondly, it simplifies protocol audits by concentrating all asset-holding logic in a single location. It also enables sophisticated trading features like batch swaps, flash loans, and internal balances. This allows users to hold tokens "inside" the Vault to save gas on frequent trades.

Building on this foundation, V3 introduces transient accounting using EIP-1153 transient storage. This upgrade allows multiple operations to settle within a single transaction without writing intermediate state to permanent storage.

As a result, the V3 Vault is cheaper to use, easier to integrate, and vastly more flexible for DeFi developers and pool designers.

Programmable Hooks and Custom Pools

A major innovation in the V3 architecture is the programmable hooks framework. Every Balancer V3 pool can now attach a custom hook contract that executes at crucial moments during the pool's lifecycle.

These moments include before and after swaps, before and after liquidity adjustments, or whenever the pool calculates a dynamic swap fee. Therefore, pool builders can implement custom logic without rewriting the core pool code.

Developers use hooks to build highly practical DeFi solutions. For example, dynamic fee curves automatically raise fees during volatile market periods and lower them when markets stabilise.

MEV-protection hooks actively discourage sandwich attacks. Oracle hooks stream accurate pool prices directly into external smart contracts. Lottery or rebate hooks redirect a share of fees back to participants.

Additionally, TWAMM (time-weighted-average-market-maker) hooks execute large orders over many blocks to minimise slippage. Ultimately, this permissionless extensibility mirrors the functionality of Uniswap V4 hooks, transforming Balancer V3 into a robust framework for experimenting with advanced AMM designs.

Standardised Pool Types in Balancer V3

Balancer V3 supports a variety of standardised pool types designed to maximise capital efficiency and yield generation.

Weighted Pools are the classic Balancer format. They hold two to eight tokens at configurable weights, functioning as self-rebalancing index funds.

Stable Pools use StableSwap-style math pioneered by Curve. They optimise trading for assets that maintain parity, such as stablecoins or liquid staking tokens (LSTs).

Boosted Pools automatically route idle liquidity to external lending markets like Aave or Morpho. As a result, LPs earn swap fees and lending yield simultaneously. Notably, V3 makes these pools 100% boosted by default, which is a significant upgrade from V2's partial boosting mechanism.

Moreover, V3 introduces re-CLAMM pools, a concentrated-liquidity implementation that focuses liquidity within defined price ranges, similar to Uniswap V3 but utilising Balancer's unique mechanics.

V3 also integrates Gyroscope's CLP (concentrated liquidity pool) math. These concentrated-liquidity models play a central role in Balancer's strategy to compete with leading DEXs on capital efficiency, as outlined in the DAO's 2026 roadmap.

BAL Token and veBAL Governance

BAL is Balancer's governance and incentive token, with a hard cap of 94 million tokens. The emission schedule halves every four years, similar to Bitcoin and primarily directs rewards to liquidity mining on whitelisted pools.

Balancer governance relies on the vote-escrowed model, adapted from Curve's veCRV system. To gain governance power, users lock an 80/20 BAL/WETH Balancer Pool Token for up to one year to receive veBAL.

The protocol scales the amount of veBAL based on both the deposit size and the lock duration. A full one-year lock grants maximum voting weight, which then decays linearly until expiration. Importantly, users cannot transfer or sell veBAL as a binding commitment to the protocol's long-term success.

veBAL holders wield two primary forms of power. First, they vote on governance proposals via Snapshot to authorise DAO changes, such as modifying pool gauges or approving budgets.

Second, and more critically, they vote on gauge weightings every week to direct BAL emissions to their preferred pools. Because BAL rewards actively attract liquidity, protocols fiercely compete for these emissions.

This competition has fuelled the rise of vote-bribe markets like Hidden Hand and Votemarket, where projects pay veBAL holders for their votes. Furthermore, the protocol distributes 75 percent of its fees directly to veBAL holders, providing a lucrative revenue share alongside governance control.

Multi-Chain Deployments

Balancer is deployed across a broad set of EVM networks, including Ethereum mainnet, Arbitrum, Optimism, Polygon, Base, Avalanche, Gnosis Chain, and others.

Each deployment runs its own Vault and set of pools, but all veBAL voting happens on Ethereum mainnet, with the gauge results cross-chain-relayed to other networks. This allows Balancer to capture volume across many chains while keeping governance coordinated in one place.

The November 2025 V2 Exploit

On 3 November 2025, Balancer V2 suffered one of the largest DeFi exploits of the year, with losses estimated between roughly 100 and 128 million US dollars across Ethereum mainnet and several Layer 2 networks.

The vulnerability was a precision-loss rounding error in the integer fixed-point arithmetic used by Balancer V2's Composable Stable Pools. The attacker carefully executed a sequenced series of small swaps under low-liquidity conditions, compounding the rounding error into significant cumulative profits.

Because Balancer V3 uses a different pool architecture and different math, the V3 Vault and V3 pools were not affected by the bug.

However, the incident had direct consequences for the protocol: the Composable Stable Pool V6 factory was disabled. The gauges for affected pools were killed to halt BAL emissions to compromised liquidity. Lastly, the affected chains, including Berachain and Sonic, took emergency action to contain contagion.

Recovery has been partial. StakeWise DAO's emergency multisig whitehat-recovered a large share of the stolen osETH and osGNO. The Balancer DAO has proposed a framework (BIP-892) to redistribute rescued assets to affected liquidity providers.

A further 21,000 ETH sits dormant in exploit-linked wallets, and the DAO has offered recovery-assistance bounties (BIP-908) to encourage its return.

The exploit is a clear reminder that even a protocol with years of audits and a large bug bounty can have subtle accounting bugs. Also, the legacy contract versions (like V2) can remain a material source of risk for long after a protocol has moved on to its next release.

Competitors and Market Position

Balancer competes most directly with Curve Finance on stable-asset swaps, with Uniswap V3 and V4 on concentrated-liquidity volatile-pair trading, and with newer programmable-liquidity platforms (including Uniswap V4 and various intent-based DEXs) on hook-driven customisation.

Balancer's historical advantage lies in its weighted multi-asset pools and composable integrations. Aave's boosted-pool partnership, index-fund products, and protocol treasury portfolios are classic Balancer use cases that are harder to replicate on simpler AMMs.

Beyond trading, Balancer also plays a critical infrastructure role in DeFi. Many protocols rely on Balancer V2 and V3 for pool creation. For example, the 80/20 BAL/WETH BPT that underlies veBAL itself, or the pools used by liquid-staking protocols and stablecoin issuers for secondary-market liquidity.

Risks and Considerations

While Balancer V3 offers powerful yield opportunities, users must carefully evaluate the associated risks before committing capital.

Smart-contract risk is the most material concern. The November 2025 exploit of Balancer V2 demonstrated that even heavily audited code can harbour subtle mathematical flaws. Although the new V3 codebase avoided the V2 rounding bug, it remains younger and less battle-tested than its predecessor.

Impermanent loss is a permanent feature of weighted AMM pools. LPs in weighted or stable pools implicitly run a rebalancing strategy that sells outperforming assets and buys underperforming ones. During sustained directional market moves, this can result in meaningful losses compared to simply holding the assets.

Boosted-pool risk compounds protocol exposure. Because boosted pools route idle liquidity into external lending markets, LPs face layered risks, not only Balancer's own smart-contract risk, but also the external lending protocol's vulnerabilities, interest-rate models, and overall solvency.

Governance and emission risk matters for BAL holders. Since veBAL voting controls BAL emissions and 75 percent of fee distributions, governance capture by a small group of large lockers could redirect value away from smaller participants. Bribe markets, while economically rational, can also shift incentive structures in ways that may not align with the protocol's long-term health.

Oracle and external-dependency risk applies to any pool relying on price feeds or external lending markets. Users should always thoroughly read a pool's documentation, including details on attached hooks before supplying liquidity.

Exploring Balancer via Portals.fi

Portals.fi is a DeFi aggregation platform that makes it easier to interact with protocols like Balancer through a unified interface. Users can route swaps through Balancer pools, manage liquidity positions, and compare opportunities across protocols from a single access point. For more information about how Portals.fi works, visit portals.fi.

This article is for informational purposes only and does not constitute financial advice. DeFi protocols carry inherent risks including smart contract vulnerabilities, impermanent loss, oracle failures, governance risk, and market volatility. Balancer V2 was exploited for approximately 100–128 million US dollars on 3 November 2025; while V3 was not affected, recovery and remediation for V2 users remain ongoing. Always read the latest protocol documentation, post-mortems, and your own risk tolerance before supplying liquidity. For our full disclaimer, please visit here.