What Is Balancer? The Programmable AMM Behind Weighted Pools and V3 Hooks, Explained

How Balancer's programmable AMM works: weighted and boosted pools, V3 hooks, the BAL token, and the 2025 exploit that reshaped it in 2026.

By Web3Wagmi Team9 min read
Table of contents

Balancer is a programmable automated market maker (AMM) and decentralized exchange where a single liquidity pool can hold up to eight tokens in any weighting you choose — not just the fixed 50/50 pair that defines most AMMs. That one design decision turns Balancer into something broader than a swap venue: it is liquidity infrastructure that can behave like a self-rebalancing index fund, a custom-weighted treasury, or a yield-bearing stablecoin hub. This is the full picture in 2026 — how the mechanism works, what Balancer V3's hooks and boosted pools changed, the token reset that followed a brutal exploit, and where the protocol actually stands today.

What Balancer is and who built it

Balancer started as a research project at the software firm BlockScience in 2018, led by Fernando Martinelli and Mike McDonald. The core insight was mathematical: Uniswap's constant-product formula is just a special case of a more general equation. Generalize it, and you can build a pool that holds many tokens at arbitrary weights while still guaranteeing a tradeable price for every pair inside it. Balancer Labs spun out, raised $3 million in 2020, and launched on Ethereum mainnet that year. The BAL governance token was distributed to liquidity providers starting June 1, 2020 — one of the first big "liquidity mining" programs in DeFi.

Since then Balancer has shipped three major generations. V1 proved the multi-token pool concept. V2 (2021) introduced the single-Vault architecture that most of DeFi later imitated. V3 (December 2024) rebuilt the whole system for programmability, which is where the modern story lives. Throughout, Balancer's identity has been the same: not the flashiest DEX, but the one whose pool math and Vault design keep getting copied by everyone else.

How Balancer works: weighted pools and the Vault

The foundational product is the weighted pool. Instead of forcing a 50/50 split, each token in a Balancer pool carries a weight — its target share of the pool's total value. An 80/20 BAL/WETH pool keeps roughly 80% of its value in BAL and 20% in WETH; an equal-weight pool might spread value across eight assets at 12.5% each. As traders swap and prices move, arbitrageurs are incentivized to trade the pool back toward its target weights, so the pool continuously self-rebalances without a manager. That is why Balancer is often described as an on-chain index or portfolio primitive: an 80/20 pool is effectively a rebalancing fund that also collects swap fees.

A concrete way to see the difference: on Uniswap you can only make an ETH/USDC pair. On Balancer you can build a single pool that holds ETH, wBTC, and three stablecoins at whatever weights you want, and it will price every one of the ten possible pairs inside it. That is why teams have used Balancer pools as on-chain treasuries and index products, and why liquidity-bootstrapping pools (LBPs) — weighted pools whose weights shift over time to launch a token at a falling price — became a popular fair-launch mechanism built directly on Balancer's math.

Balancer offers several pool types beyond weighted pools — stable pools (tuned for assets that should trade near parity, like stablecoin baskets or staked-ETH derivatives) and, in V3, boosted pools (below). All of them plug into one shared contract: the Vault.

The Vault is Balancer's architectural signature. Rather than every pool holding its own token balances, a single Vault holds the assets for all pools and handles the accounting, while individual pools only supply the math that prices a trade. This separation means less code lives in each pool (so pools are cheaper and safer to build), and it unlocks batch swaps that route a single transaction through several pools while only moving the net token amounts in and out at the end — a major gas saving. V2 pioneered this; V3 took it much further.

What Balancer V3 changed: hooks, boosted pools, and the Vault rewrite

Balancer V3 went live on Ethereum and Gnosis in December 2024, then expanded to Arbitrum and Base in early 2025 and further EVM chains. It is best understood as a developer-experience overhaul: V3 pushes even more shared logic down into the Vault, so building a new pool type is dramatically simpler than before. Three features matter most.

Transient accounting (EIP-1153). V3 uses Ethereum's transient storage opcodes to track balances within a single transaction cheaply, which is what makes secure Vault reentrancy — and therefore hooks — possible without the gas and risk that used to come with it.

Hooks. A hook is custom code that runs at a defined point in a pool's lifecycle — before or after a swap, or when liquidity is added or removed. This is the "programmable" in programmable AMM: developers can attach dynamic fees, custom rebalancing, or protective logic to an otherwise standard pool. The launch example is the StableSurge hook, which raises fees automatically when a stablecoin or pegged asset starts drifting off peg, defending the peg while paying LPs more during exactly the volatile moments that usually hurt them. Hooks turn Balancer from a fixed set of pool types into an open design space.

100% boosted pools. In partnership with Aave, V3 lets a pool keep up to 100% of its liquidity in yield-bearing form. Idle liquidity that would otherwise sit unused is supplied to Aave's lending market, so LPs earn lending interest on top of swap fees. Small "liquidity buffers" hold a little base-asset inventory to make routine swaps gas-efficient without unwinding the lending position on every trade. The result is capital efficiency that a plain AMM can't match: your liquidity works two jobs at once.

The BAL token and the 2026 tokenomics reset

BAL is Balancer's governance token. Under the model that ran from 2020 through 2025, holders locked an 80/20 BAL/WETH pool token to mint veBAL (vote-escrowed BAL) for up to a year, which granted governance votes, a share of protocol fees, and the power to direct BAL emissions toward specific pools — the same "vote-escrow" design Curve popularized. It worked, but it also created constant inflationary sell pressure and the familiar "bribe" economy around emission votes.

In 2026 that model was torn up. Following the exploit and the wind-down of Balancer Labs, the DAO moved on a sweeping reset proposed by co-founder Martinelli in March 2026:

  • All BAL emissions halted — immediately, with no transition period, ending inflationary issuance.
  • veBAL wound down — the vote-escrow system is being sunset, with a stablecoin compensation campaign for locked holders.
  • 100% of protocol fees to the DAO treasury, and the V3 protocol fee take reduced from 50% to 25% to make LPing more attractive.
  • A BAL buyback program funded from treasury holdings, intended partly to provide exit liquidity for holders who want out.

The strategic message is blunt: Balancer is pivoting from token-incentivized growth to fee-funded survival. BAL still exists and trades, but its value now rests on real protocol fees and treasury policy rather than on a stream of new tokens. If you're weighing BAL as an asset, that is the single most important change to understand.

The numbers: TVL, scale, and market position

Balancer's scale in 2026 is a fraction of its peak. Total value locked fell roughly 95% from a 2021 high near $3.5 billion to around $150–190 million by 2026 (it bottomed near $157 million after the exploit and has since partly recovered), with legacy V2 holding a chunk of what remains and V3 carrying the newer liquidity. That collapse reflects the broader DeFi drawdown, the 2025 exploit, and the loss of BAL emissions that once subsidized deposits.

But TVL undersells Balancer's actual footprint. Its Vault design, weighted-pool math, and boosted-pool concept have been forked and imitated across DeFi — dozens of V2 forks exist, and Balancer's ideas show up in protocols that never carried its name. For years Balancer sat comfortably among the top decentralized exchanges by liquidity alongside Uniswap and Curve; today it is smaller and more specialized, leaning on V3's programmability and yield-bearing pools rather than raw size. It is a blue-chip primitive operating at a diminished scale — influential, still live, but no longer a giant.

The ecosystem and integrations

Balancer's most important relationships are structural. The Aave integration is what powers V3 boosted pools. CoW Protocol runs AMMs and solver-based routing built on top of Balancer liquidity, and Balancer liquidity is routed by most major DEX aggregators. Ecosystem partners like Gyroscope deploy specialized pool types (E-CLPs) on Balancer's infrastructure, and Beets (formerly Beethoven X) is a long-running Balancer-based deployment on chains like Sonic and Optimism — effectively a sister protocol running the same engine.

This is the pattern to notice: Balancer's value to the ecosystem is often as the liquidity layer other products build on, not just as a consumer-facing exchange. When you swap on an aggregator or use a staked-ETH strategy, Balancer pools may be underneath even if you never visit the app.

How to actually use Balancer

If you want to use Balancer directly, the flow is straightforward — but do it carefully given the protocol's current state.

  1. Go to the official app at balancer.fi and connect a wallet (MetaMask, Rainbow, WalletConnect). Always reach it via your own bookmark, never a search ad or DM link.
  2. Pick a network. Balancer runs on Ethereum, Arbitrum, Base, Gnosis and other EVM chains; fees are far lower on L2s than on Ethereum mainnet.
  3. To swap, use the swap interface like any DEX — it routes your trade through the best available pools. Check the price impact on large or illiquid trades.
  4. To provide liquidity, browse pools and prefer V3 pools (unaffected by the 2025 exploit) over legacy V2. Boosted pools earn lending yield plus swap fees; understand that you take on the risk of every token in the pool, not just one.
  5. Understand impermanent loss. Weighted pools rebalance automatically, which means you are systematically selling winners and buying losers relative to holding — the classic AMM trade-off. An 80/20 pool reduces this exposure versus a 50/50 pool, which is one reason the 80/20 structure is popular.

Risks and what to watch

Be clear-eyed. Balancer in 2026 is a functioning but higher-risk protocol.

  • The November 2025 exploit. An attacker drained roughly $128 million from V2 pools across multiple chains in under 30 minutes, exploiting a rounding error in the pool invariant math combined with crafted batch swaps. V2 was vulnerable; V3 was not. This is the strongest argument for using V3 pools and treating remaining V2 liquidity with caution.
  • Corporate wind-down. Balancer Labs shut down as a company in 2026. The protocol continues under DAO control with a lean team, but there is no longer a well-funded corporate entity driving development, marketing, and incident response the way there was.
  • Smart-contract risk is never zero. V3 is heavily audited and partly formally verified, and the exploit didn't touch it — but DeFi history is a long list of "audited and still exploited." Size your exposure accordingly.
  • Token and treasury risk. With emissions gone and veBAL winding down, BAL's trajectory depends entirely on fees and treasury decisions. That's healthier long-term but removes the incentive flywheel that once attracted liquidity.
  • Phishing. A shrinking, restructuring protocol is a prime target for fake "migration" and "claim" scams. Verify every link and contract address through official channels.

Bottom line

Balancer's lasting contribution is architectural: it proved that an AMM pool doesn't have to be a 50/50 pair, and its Vault, weighted-pool math, and boosted-pool design became reference patterns that the rest of DeFi copied. V3's hooks and 100% boosted pools are genuinely advanced, and the fact that V3 sailed through the 2025 exploit that gutted V2 says something real about the newer code. But the honest 2026 picture is a blue-chip primitive at a diminished scale — TVL down ~95%, corporate backer gone, token model reset toward survival. If you use it, prefer V3, understand the pool risk you're taking, and treat it as important infrastructure operating in a leaner, more fragile phase than its reputation suggests.

For related reading: what is DeFi, what is an AMM, and how to earn yield on stablecoins.

Not financial advice. BAL is volatile, the protocol is in a restructuring phase, and smart-contract risk is real — always verify details on official Balancer channels.

Frequently asked questions

What is Balancer in simple terms?

Balancer is a decentralized exchange and automated market maker (AMM) on Ethereum and other EVM chains. Where a typical AMM pool holds two tokens split 50/50, a Balancer pool can hold up to eight tokens in any weighting you define — for example 80% BAL and 20% WETH, or an evenly-weighted basket of stablecoins. That flexibility makes it as much a liquidity infrastructure layer and on-chain index tool as a place to swap tokens.

How is Balancer different from Uniswap?

Uniswap pools are fixed 50/50 pairs. Balancer generalizes that formula so a single pool can hold multiple tokens with arbitrary weights, which lets it act like a self-rebalancing index fund that also earns swap fees. Balancer V3 goes further with a single-Vault architecture, a hooks framework for custom pool logic, and boosted pools that push idle liquidity into lending markets. In practice the two protocols overlap on simple swaps but diverge sharply on advanced pool design.

What are Balancer weighted pools?

Weighted pools are Balancer's signature product: liquidity pools where each token has an independent weight representing its share of the pool's value. A pool might be 80/20, 60/20/20, or an equal split across eight assets. The pool automatically maintains those ratios as prices move, so it behaves like a rebalancing portfolio. This is why Balancer is often described as "programmable liquidity" rather than just a swap venue.

What are boosted pools and hooks in Balancer V3?

Boosted pools let liquidity providers keep up to 100% of their position in yield-bearing form — idle liquidity is routed into a lending market like Aave, so LPs earn lending interest on top of swap fees, with liquidity buffers handling gas-efficient swaps. Hooks are a V3 framework that lets developers run custom logic at key points in a pool's lifecycle (before or after a swap, add, or remove), enabling things like dynamic fees and the StableSurge hook that defends stablecoin pegs during volatility.

What happened in the Balancer exploit?

In early November 2025 an attacker drained roughly $128 million from Balancer V2 pools across multiple blockchains in under 30 minutes. The root cause was a rounding-error bug in the pool invariant math (in the token up-scaling logic) that, combined with crafted batch swaps, let the attacker distort pool token prices and extract value. The older V2 code was vulnerable; the newer V3 architecture was not. Dozens of V2 forks, including Beets, were also affected.

Does Balancer still work after the 2026 shutdown?

Yes. What shut down in March 2026 was Balancer Labs, the corporate entity, which co-founder Fernando Martinelli said had become a legal and financial liability after the exploit. The protocol itself continues online under DAO governance with a leaner operating team. The smart contracts, the app, and V3 pools keep functioning — governance simply moved fully on-chain.

Is the BAL token still worth anything after the tokenomics reset?

BAL still exists and trades, but its role changed fundamentally in 2026. The DAO voted to halt all BAL emissions, wind down the veBAL vote-escrow system, route 100% of protocol fees to the treasury, and run a BAL buyback program funded from treasury holdings. The goal is to end inflationary sell pressure and let fees — not new token issuance — sustain the protocol. It is a bet on survival over growth, and BAL's value now hinges on protocol fees and treasury policy rather than emissions.

Is Balancer safe to use in 2026?

Balancer V3 was not affected by the 2025 exploit and has been heavily audited and formally verified, but the protocol carries real, elevated risk: the corporate backer wound down, TVL fell roughly 95% from its peak, and smart-contract risk is never zero in DeFi. Treat it as a still-functional but higher-risk venue, prefer V3 over legacy V2 pools, verify every URL, and don't over-allocate.

Sources & further reading