Two liquidity providers each deposit $100,000 into the same ETH/USDC pool. The full-range LP — capital spread from zero to infinity — earns $180 in fees this month. The concentrated LP, same capital but placed inside a $2,800–$3,200 range, earns $2,400 in fees. Same pool, same tokens, same price action. The gap is not luck and not alpha. It is one design choice introduced by Uniswap v3 in May 2021: concentrated liquidity.
This piece covers how concentrated liquidity actually works, the math behind the capital-efficiency boost (and the impermanent loss it amplifies), what Uniswap v4 hooks change for LPs in 2026, and a practical framework for when concentrated positions win versus when they quietly bleed.
TL;DR. Concentrated liquidity lets LPs allocate capital inside a specific price range instead of from zero to infinity, producing up to 4,000× greater capital efficiency versus full-range positions in stablecoin pairs. The tradeoff is amplified impermanent loss — if price exits your range, your position stops earning fees and holds 100% of the less-valuable token. Uniswap v4 (live January 31, 2025) extends concentrated liquidity with hooks — modular on-chain plugins that modify pool behavior at swap, deposit, and withdrawal points. Hooks enable dynamic fees, on-chain limit orders, TWAMM, and custom oracles, but add a new smart-contract risk surface.
How concentrated liquidity actually works
A standard automated market maker — the original Uniswap v2 design — uses the constant-product formula x × y = k. Liquidity is spread from price zero to infinity, which is elegant but wasteful. Consider a USDC/USDT pool: the price almost never leaves the 0.99–1.01 band, yet 99.5% of the LP capital sits permanently idle, earning nothing because no trades ever reach those prices. Every dollar parked at price 0.50 or price 2.00 is dead weight.
Uniswap v3, launched May 2021, fixes this. A liquidity provider selects a lower tick and an upper tick, and their capital is only used to make trades when the market price sits inside that range. The pool still uses the familiar curve, but only inside the band you chose. Outside it, your position is effectively inactive.
A tick is a price step of one basis point (0.01%). The mapping is price(i) = 1.0001^i, and the full theoretical range runs from tick −887,272 (price effectively zero) to tick +887,272 (price effectively infinite). Inside that, you can be as narrow as a single basis point or as wide as the entire range — your choice determines your capital efficiency.
The capital-efficiency gain is the headline number. If you provide liquidity in a $1.00 ±1% range on a stablecoin pair, then $10,000 of capital provides the same trading depth as $1,000,000 in a full-range position. The Uniswap v3 whitepaper cites up to 4,000× efficiency for narrow stablecoin ranges. That is the source of the 13× fee gap in the opening example.
Worked example: $100K in ETH/USDC at $3,000
- Full-range position: capital spread from $0 to ∞. Depth at $3,000 is thin; you capture every trade but earn roughly $180/month in fees on $100K (annualized ~2.2%).
- Concentrated ±10% ($2,700–$3,300): the full range is ~$6,000 wide in price terms; your band is ~$600 — roughly a 10× boost in capital density at the current price. Same $100K earns ~$1,800–$2,400/month while ETH stays inside the band.
- Catch: if ETH moves outside $2,700–$3,300, you earn zero fees and hold one side of the pair. Active management is not optional.
The math behind the boost — and the impermanent loss
The boost itself is simple. It equals the width of the full price range divided by the width of your chosen range, in price terms. A stablecoin pair where trades happen between 0.99 and 1.01 (a 2% range) gives roughly a 50× boost. ETH/USDC at ±5% (a 10% range) gives about a 7× boost. ETH/USDC at ±20% gives about a 2× boost. The narrower you go, the higher the multiplier — and the higher the risk.
That risk is the catch. Impermanent loss is amplified by the same multiplier that boosts your fees. When price moves inside your range, your position's value diverges from a balanced hold faster than a full-range position would. The deeper math — how the curve reshapes your holdings as price crosses ticks — is covered in our impermanent loss explainer, but the rule is: the boost that gave you 10× fees will also give you 10× the pain when the market moves against you.
Out-of-range risk is even blunter. If the market price exits your band entirely, your position stops earning fees and ends up holding 100% of whichever token is less valuable at that moment. You are now out of the trade, holding the wrong asset, and paying gas to rebalance back in. Many new LPs learn this on the first 15% move.
The honest framing: concentrated liquidity is an active strategy, not a set-and-forget yield. You need to set price alerts, monitor time-in-range, and rebalance — weekly is a reasonable cadence for medium-volatility pairs, daily for narrow stablecoin ranges that drift on depeg events. If you are not willing to manage the position, you should not be running it.
Uniswap v4 hooks — what changes for LPs in 2026
Uniswap v4 went live on January 31, 2025, per Uniswap Labs' announcement. The headline architectural change is the singleton: instead of one contract per pool (v3's model), all pools live inside a single PoolManager contract. For LPs the practical effect is lower gas on multi-hop swaps and simpler cross-pool routing, since routing no longer crosses contract boundaries.
The bigger conceptual shift is hooks. A hook is a smart contract the pool calls before or after a defined action — before swap, after swap, before liquidity change, after liquidity change. That gives developers a place to inject custom logic at exactly the moments that matter for pool behavior.
| Hook type | What it does | LP impact | Maturity |
|---|---|---|---|
| Dynamic fee | Raises fee when toxic flow detected (anti-MEV) | Higher capture in volatile periods | Live |
| On-chain limit order | Fills at a specific price, not a range | Converts LP into familiar order type | Live |
| TWAMM | Slices large orders over time | Smoother flow, less toxic sweep | Early |
| Custom oracle | Builds TWAP from pool data | Reduces dependency on external oracles | Early |
A dynamic-fee hook in practice looks like this: the hook reads incoming swap size and price impact per block, and if a swap looks adversarial — large size, sandwich pattern, or oracle deviation — the pool fee spikes from 0.05% to 1% for that transaction only. LPs capture the spike as extra yield, and the attacker's economics break. The tradeoff is that benign large trades also pay the elevated fee, so hook design is a constant balancing act between MEV protection and overcharging legitimate flow.
On-chain limit-order hooks are conceptually simpler but operationally interesting: instead of providing liquidity across a range, an LP deposits at a single tick and the hook fills the order linearly as price crosses it, converting the AMM position into something that behaves like a CEX limit order. For LPs used to thinking in ranges, this is a different mental model — you are picking a price, not a band, and your fill rate depends on how long price spends at that level.
The honest caveat: hooks are smart contracts with their own attack surface. A pool with a buggy or unaudited hook can be drained regardless of how well the core PoolManager was audited. Not every hook in production has been reviewed by a reputable firm, and the difference between a clever hook and a clever-looking bug is often invisible until exploit. Treat hook-enabled pools as a separate risk category — check the audit, the bug bounty, and the deployer reputation before depositing meaningful capital.
When concentrated liquidity wins — and when it loses
Concentrated liquidity wins where price genuinely stays in a narrow band most of the time. That describes stablecoin pairs (USDC/USDT, USDC/DAI) where 99%+ of volume happens inside 0.99–1.01, and correlated assets like stETH/ETH and wstETH/ETH where the spread is structural rather than incidental. In both cases the boost is large and time-in-range is high — the strategy earns its keep.
It loses on volatile pairs. ETH/SOL with a narrow range gets hit by impermanent loss on every swing; with a wide range the boost collapses to the point where you are essentially running a full-range position with extra gas. Long-tail tokens with low volume are even worse: the fee revenue is too thin to cover the management cost, and the occasional 30% move wipes out months of earnings.
A useful rule: boost × probability of price staying in range = expected return. Below 5× in stablecoin pairs, concentrated positions are usually worth running; below 2× in volatile pairs, they usually are not. Multiply by your confidence in the band holding, then subtract realistic rebalancing gas, then compare to a passive full-range position. The arithmetic rarely lies.
If the management work is the blocker, auto-compounding vaults handle it for you. Beefy and Yearn both run concentrated-liquidity vaults that monitor ranges, rebalance when price approaches edges, and compound fees — for a performance fee that's still cheaper than doing it badly yourself. See our reviews at /protocols/beefy and /protocols/yearn.
Sources and further reading
- Uniswap v3 whitepaper — official source for tick math, capital efficiency, and the 4,000× claim.
- Uniswap v4 documentation — singleton PoolManager, hooks architecture, and lifecycle callbacks.
- ethereum.org — DeFi developer overview — broader context on AMM designs and the v3/v4 lineage.
Frequently asked questions
What is concentrated liquidity in Uniswap v3?
Concentrated liquidity is the Uniswap v3 design where an LP deposits capital into a chosen price range (a lower and upper tick) instead of from zero to infinity. The pool only uses your liquidity when the market price sits inside that range, which concentrates your capital around the prices where trades actually happen and can produce up to 4,000× greater capital efficiency than a full-range position in stablecoin pairs.
Is concentrated liquidity more profitable than full-range LP?
Often yes, but only while price stays inside your range. A concentrated position can earn many times more fees per dollar than a full-range position in the same pool, because the same depth is provided with less capital. The tradeoff is that impermanent loss is amplified by the same multiplier, and if price exits your range you stop earning fees entirely and hold 100% of the less-valuable token. Net profitability depends on boost, time in range, and rebalancing cost.
What is a Uniswap v4 hook?
A Uniswap v4 hook is a smart contract that the pool calls before or after key actions — swap, liquidity deposit, liquidity withdrawal — to modify pool behavior. Hooks enable dynamic fees that spike when toxic flow is detected, on-chain limit orders, TWAMM (large orders sliced over time), and custom oracles. Hooks are modular plugins live since v4 launched on January 31, 2025, but each hook carries its own smart-contract risk and not all are audited.
When should I not use concentrated liquidity?
Avoid concentrated liquidity on highly volatile pairs like ETH/SOL where narrow ranges get hit by impermanent loss on every swing and wide ranges eat the boost, and on long-tail tokens where trading volume is too low to cover the cost of active management. A simple rule: boost multiplied by the probability of price staying in your range equals expected return. Below roughly 5× in stablecoin pairs it is usually worth it; below 2× in volatile pairs it is usually not.
How is concentrated liquidity different from concentrated market making on CEXs?
Centralized exchanges run off-chain order books where a market maker posts bids and offers at discrete prices and can move them instantly with no on-chain cost. Concentrated liquidity on Uniswap is on-chain and uses a continuous curve inside your tick range; rebalancing requires a transaction and pays gas, and any unused range earns nothing. The capital-efficiency math is similar, but the operational model and risk surface are very different.
Grade your concentrated LP position
The DifiCalc yield-risk grader scores your tick range's impermanent-loss exposure and out-of-range probability before you commit capital.
Open the Yield-Risk GraderRelated reading: Impermanent Loss, Explained, Uniswap vs PancakeSwap, Curve vs Balancer, Beefy Protocol Review.