Concentrated liquidity explained: Uniswap v3/v4 and how LP positions changed

For most of decentralized finance’s first decade, providing liquidity to an automated market maker was a passive act. You deposited tokens into a pool, received an LP token, and watched your capital spread evenly across every possible price from zero to infinity. Uniswap v2 made this model famous — and also made its flaw famous: the overwhelming majority of that capital sat at prices that never traded, earning nothing while bearing full exposure. Estimates at the time suggested that only a small fraction of pool liquidity was ever doing meaningful work. It was, in effect, a factory where most of the machines were switched off.

Concentrated liquidity — introduced with Uniswap v3 in 2021 and extended architecturally in v4 — was the industry’s answer to that inefficiency. Instead of spreading capital across all prices, liquidity providers allocate it to a chosen price range, and the same capital now works harder wherever it is actually needed. The change transformed what it means to be an LP: returns multiplied for skilled managers, and losses multiplied for those who treated the new system like the old one. This article walks through the mechanics, the economics, and the practical decisions that concentrated liquidity forces on anyone providing or analyzing liquidity today.

The problem concentrated liquidity solved

The v2 model carried an economic contradiction at its core. Capital efficiency and passive management could not coexist at scale: a passive LP had to provision for every conceivable price, which meant provisioning mostly for prices that never occurred. The consequences compounded across the ecosystem:

Dead capital. For a stablecoin pair trading tightly around its peg, nearly all of the pooled capital was positioned at prices the pair would never visit. That capital earned zero fees while still being available — and being exposed — to any catastrophic deviation.
Slippage that was worse than necessary. Because spread-out liquidity was shallow at the trading price itself, large swaps moved prices more than they needed to. Traders paid for the capital inefficiency of the LPs.
A one-size-fits-all fee. Every pool in v2 had a single 0.30% fee tier, whether the pair was a volatile meme asset or two stablecoins. Neither side of that trade was priced correctly.

Uniswap v3 did not tweak these parameters; it replaced the underlying model. Liquidity became a resource that LPs place at specific prices, and the pool itself became a composite of many different ranges stacked on top of one another. The result behaves like an order book made of intervals rather than of individual orders — which is, in many ways, what the mechanism actually is.

How concentrated liquidity works: ticks, ranges and positions

The mechanics take a little unpacking, but the vocabulary is small. The essential concepts:

Price ranges. Each LP chooses a lower and upper bound — for example, providing ETH/USDC liquidity only between 2,800 and 3,200. The pool never holds that LP’s capital at prices outside the range. Below the lower bound, the position is entirely in ETH; above the upper bound, entirely in USDC.
Ticks. The price scale is discretized into ticks — price points separated by a fixed increment. Each position’s range sits across one or more ticks, and liquidity within each tick segment is pooled with everyone else who chose the same range.
Capital concentration. If the current price is inside your range, your capital is active and earns fees on every swap that passes through. The narrower the range around the price, the more fee revenue a given dollar of capital captures — a v3 position concentrated to a narrow band can, in theory, deploy the capital efficiency of thousands of times more v2 liquidity.
Active range dynamics. When price exits a position’s range, the position stops earning and its composition has flipped entirely to one asset. It must be re-ranged — withdrawn and re-provisioned at the new price — to resume earning.

The re-rangability point is where the model quietly shifts responsibility from protocol to provider. In v2, an LP could genuinely do nothing and remain invested. In v3, doing nothing is itself a decision — usually a poor one — because price drift inevitably moves out of any static range.

The new role of the LP: active manager or outsource

Concentrated liquidity turned liquidity provision from a deposit decision into a portfolio-management decision. The range choice, the rebalancing discipline and the fee-tier selection together determine returns at least as much as the choice of pool itself. The main approaches observed in practice:

Approach	How it works	Strengths	Weaknesses	Who it suits
Wide passive position	Very broad or full range, rarely touched	Minimal maintenance; LP-like behavior	Much lower fee yield than narrow ranges	Long-horizon holders who value simplicity
Narrow active range	Tight band near price, re-ranged as price moves	Highest capital efficiency and fee capture	Requires constant monitoring; high impermanent loss when out-of-range	Experienced LPs or automated managers
Stablecoin or correlated pairs	Tight ranges on low-volatility pairs	Small price risk; near-yield-like fee income	Range breaks can still happen in depegs	Conservative LPs seeking modest returns
Managed vaults	A vault or service rebalances positions on behalf of depositors	Passive access to active strategy; professional management	Management fees; trust in the vault’s logic and contracts	Users who want concentrated returns without managing
Single-sided range orders	Liquidity placed above or below the current price	Functions like a limit order that also earns fees while unfilled	Execution depends on price actually reaching the range	Traders accumulating or exiting at target prices

The table’s most consequential row is the managed vault. Concentrated liquidity is technically demanding enough that a large infrastructure ecosystem grew around it — automated range management, fee-compounding vaults, hedged LP strategies. For most retail users, outsourcing management to a vault is the realistic path into concentrated positions, with the caveat that the vault’s strategy and its fees now sit between the user and the pool.

Impermanent loss, amplified

Concentrated liquidity does not eliminate impermanent loss — it concentrates it. Within the chosen range, a v3 position behaves like a leveraged version of a v2 position: the same price move produces a proportionally larger divergence loss, because the position’s effective exposure is multiplied by concentration. The trade-off is stark and direct:

In-range, fees flow generously. A narrow position inside its range collects fees at a rate that can exceed a v2 position’s by a large multiple, especially in high-volume pairs.
Out-of-range, the position earns nothing while holding 100% of one asset. If price runs away from a range, the LP watches the trend from the sideline, having converted a chunk of the appreciated asset into the other one at the range boundary — the worst moment for that conversion in a trending market.

Empirical analyses of v3 pools have repeatedly shown a bifurcated outcome: a minority of well-managed, often sophisticated or informed positions collect disproportionate profits, while the average passive LP frequently underperforms simply holding the two assets. That finding is the honest summary of the concentrated liquidity era: the mechanism redistributed returns from passive providers to active ones, from the pool’s average participant to its best participants.

What Uniswap v4 changed: the architecture catches up

Uniswap v4, deployed in 2025, did not change the core concentrated liquidity math — it reorganized where and how it lives. The centerpiece is the singleton architecture: instead of one contract per pool, all pools live inside a single contract, with a hook system that lets pools run custom logic. The practical changes that matter to LPs and integrators:

Singleton and flash accounting. A single contract hosts every pool, and complex multi-pool operations settle in one transaction with intermediate balances tracked internally. Routing across pools became dramatically cheaper, which matters for every position that rebalances or compounds fees.
Hooks. Each pool can attach custom contracts that execute at defined points in a swap or position’s lifecycle — modifying fees, adding dynamic curves, or imposing custom rules. In effect, every pool can become a bespoke venue: pools with fees that vary with volatility, pools that add Oracle data, pools with asymmetric ranges.
Native ETH support and cheaper approval flows. Positions can hold and transact in native assets directly, trimming gas costs around common LP operations.
Dynamic fees as a first-class feature. Fee tiers can adjust based on conditions, letting a pool’s pricing respond to volatility regimes rather than being fixed at deployment.

The ecosystem consequence of hooks deserves emphasis: it blurred the boundary between the base protocol and the applications built on it. Yield vaults, range-order tools and custom AMM variants now build as hooks on shared infrastructure rather than as parallel forks — a consolidation that keeps Uniswap’s liquidity unified while allowing experimentation at the edges.

Fees, tick tiers and choosing a pool

Fee-tier selection interacts with concentration in ways that are not always intuitive. The tiers — from 0.01% for extremely stable pairs up through 0.05%, 0.30% and 1.00% and beyond — exist to match the expected volatility of each pair. The decision framework is straightforward in principle, trickier in practice:

High-volume, high-volatility pairs reward narrow ranges in higher fee tiers, but demand frequent management; the fees are rich precisely because the risk of being out-of-range is constant.
Stable pairs reward very tight ranges in the lowest tiers; the return per unit of risk is attractive, but it is competed for aggressively, and returns compress as more capital arrives.
Long-tail tokens may offer the highest headline yields, which usually reflect the realities of a thin market: large price swings, unpredictable out-of-range periods and exit costs that can exceed a season of collected fees.

One persistent source of confusion deserves a plain statement: the fee yield quoted by a pool at any moment is annualized from recent activity, not a promise. Volatility concentrates fees — the same trading activity that makes narrow ranges profitable also makes them fragile. Any yield that looks stable in a volatile market should be treated with suspicion, because in concentrated liquidity, stability is not the state the system is built to reward.

Practical guidance for prospective LPs

The decisions that actually determine outcomes can be reduced to a manageable checklist — one that differs sharply from the v2-era habit of picking a pool and forgetting it:

Choose the pair by your edge, not by its yield. Provide liquidity in pairs whose price behavior you understand and can track; in concentrated liquidity, ignorance about the underlying asset converts directly into losses.
Match the range width to your management capacity. A range you cannot re-range when price escapes is a liability; it is better to run a wider range honestly than a narrow one dishonestly.
Account for rebalancing costs. Re-rangings are transactions, and in volatile weeks they can be frequent; the fee edge of narrow positioning must clear those costs, or it is not an edge.
Decide whether you are the manager. If the honest answer is no, use a managed vault — and read its strategy, fee structure and rebalancing rules before depositing.
Plan the exit. Concentrated positions are exit-fee-sensitive: withdrawing when the pool holds the asset you no longer want, or when price has just left your range, can crystallize exactly the loss the strategy was meant to avoid.

None of these steps require sophistication beyond patience and record-keeping; what they require is accepting that the v3/v4 era made LPing a strategy rather than a savings account.

Conclusión

Concentrated liquidity is the change that turned automated market making from a passive yield product into an active portfolio discipline. Uniswap v3 introduced the core idea — capital placed at chosen price ranges, ticks as the price grid, fees as the reward for staying in range — and redistributed returns in the process: toward active and informed LPs, away from the passive average. Uniswap v4 then industrialized the model with a singleton architecture, hooks and dynamic fees, opening each pool to customization while keeping liquidity unified.

The practical takeaway for anyone considering LP positions today is to start from the responsibility question, not the yield question. A narrow range, a high fee tier and a volatile pair produce the biggest numbers — for the LP who manages them continuously. For everyone else, the honest options are wider ranges, correlated or stable pairs, or managed vaults that trade some of the return for professional management. The mechanism rewards exactly what it asks for: attention, timing and a view on where price is going. Since it started asking for those, being an LP stopped being passive — and that single shift, more than any protocol detail, defines how the positions changed.