Providing liquidity on Uniswap in 2027 may generate fees, but it can also underperform simply holding the same assets. Liquidity providers accept that trading changes the composition of their position. Fees received should therefore never be assessed in isolation.
The official documentation consulted on November 9, 2026 distinguishes between several pool versions and mechanisms. This guide explains the underlying principles without promising any rates for 2027. The simulations are fictional and intentionally simplified: they are designed to clarify inventories, ranges and economic comparisons, not to reproduce a real quote.
In brief
Our guide to trading fees and gas helps distinguish fees collected from costs paid. Liquidity providers must add another factor: the value of the assets they will actually hold when they exit.
- A position earns fees based on trading activity and active liquidity, not through a guaranteed coupon.
- A narrow range concentrates exposure and requires appropriate monitoring.
- Performance must be compared with holding the initial assets after all costs.
Understanding the economics of liquidity provision
A pool allows users to trade assets against the available reserves. Liquidity providers supply capital to this mechanism and may receive a share of the applicable fees. They do not earn interest because a borrower personally owes them money: they participate in a market function that is different from lending.
This activity changes the provider’s inventory. When trading removes more of one asset, the position accumulates proportionally more of the other. Providers do not choose each transaction individually; they accept the rules of the curve and their selected range. Their results therefore also depend on the price path during the period.
The protocol, its interface and the UNI token must be kept separate. Buying UNI does not automatically create a liquidity position in a pair. Conversely, supplying two assets to a pool is not the same as directly speculating on UNI. Each decision requires its own reasoning and monitoring.
Before entering, the position details should specify the network, version, asset addresses, pool and parameters. Names and logos are not enough to identify a token. A malicious asset or an illiquid representation can undermine the result even if the protocol’s underlying contracts operate as intended.
A price range is not capital protection
The documentation on concentrated liquidity explains that v3 and v4 positions use selected intervals. When the price moves outside that interval, the position’s liquidity becomes inactive and stops earning fees from trading until the price returns to the range.
Consider a fictional range between 1,800 and 2,200 quotation units for an asset. This selection does not mean the user has bought insurance against a drop below 1,800. It defines where their capital participates in trading. Outside the range, the position’s composition may end up entirely in one of the two assets.
A narrow range can increase the relative share of capital active around a given price, but it can also be exited more quickly. A wider range reflects a different exposure and distributes potential fees differently. No choice guarantees both high returns, low maintenance and no relative loss.
Re-centering the range is a new economic decision. Closing a position, converting part of the assets and opening another may generate costs and change exposure. Repeating these operations to follow every move does not ensure a better result than a less active strategy or simply holding the assets outside a pool.
Calculating relative loss against an explicit benchmark
The term impermanent loss refers to underperformance caused by changes in inventory compared with a holding benchmark. The word “impermanent” does not promise that the loss will disappear. Readers must define what they would have held, on which date and in which unit before interpreting any percentage.
Let us build a theoretical example without fees or concentrated liquidity, using a constant-product model. A position contains one unit of an asset worth 1,000 monetary units and 1,000 units of stablecoin, for an initial total value of 2,000. The product of the quantities is therefore 1,000.
If the asset’s external price doubles and the model adjusts to that price, the theoretical reserves become approximately 0.7071 units of the asset and 1,414.21 stablecoins. Their total value approaches 2,828.43. Holding the initial quantities would have represented 3,000: the gap reaches approximately 171.57, or 5.72% of the benchmark’s value.
The pool position nevertheless shows a gain compared with the initial 2,000. There is no contradiction: absolute gain and relative underperformance measure two different things. This distinction prevents positive fees from being presented as proof that providing liquidity outperformed simply holding the assets.
These figures should not be applied mechanically to a v3 or v4 position limited to an interval. Uniswap position calculations account for the position’s parameters. A real simulation must also include fees, the price path, rebalancing and costs.
Measuring the fees actually allocated to a position
The official Uniswap fee documentation distinguishes trading fees, protocol charges and, depending on the case, hook-related fees. Their distribution depends on the version and the pool. The percentage paid by the trader therefore does not necessarily correspond to an amount distributed entirely to liquidity providers.
An estimate can start with eligible volume and the share of active liquidity, but these figures change. For illustration only, assume 1 million units of volume, 0.2% in net fees allocated to providers and a constant 1% active share. The theoretically attributable revenue would reach 20 units, before the position’s own costs.
This estimate becomes misleading if the liquidity share, fees or activity within the range change. Nor does it measure the quality of trading activity. A concentration of volume over a short period does not prove that a comparable pace will continue for a year; annualizing one day can create the impression of sustainable income without any basis.
The holding period must remain consistent with fixed costs. A small position may collect a few units while paying more to enter, claim, adjust and exit. The amount shown as “fees earned” in an interface is therefore not enough to calculate the final profit in euros or another currency.
Checking pool- and asset-specific risks
The documentation on Uniswap v4 hooks describes optional contracts that customize pool behavior. Their presence requires a specific analysis. Familiarity with a v3 pair does not automatically validate a new v4 pair using additional logic or different permissions.
Liquidity provision risks include contracts, assets and position management. A stablecoin pair does not eliminate these risks. If one asset loses its peg, the pool may accumulate precisely the asset that other participants are trying to sell.
Signature security also matters. The hardware wallet guide points out that a well-protected key does not turn an unknown contract into a safe one. An approval must be understood and linked to verified addresses, without sharing a recovery phrase with someone claiming to be support.
Using borrowed funds adds debt and constraints that are not reflected in the pool’s yield alone. Our guide to crypto lending and liquidations helps distinguish these layers. Variable liquidity fees should not be assumed to systematically cover borrowing interest.
Keeping records that enable a meaningful comparison
Tracking begins with the initial quantities and their value. It then records deposits, withdrawals, fees collected, network costs and conversions. The holding benchmark must use the same capital flows and dates; without this discipline, a new injection can be mistaken for strong performance.
A separate record should describe the range, the reasons for moving it and the expected costs. Changing the strategy afterward to improve the performance narrative prevents an honest comparison. A useful record must be able to show that a position generated fees while meeting the initial objective less effectively than a simpler solution.
The quotation direction also deserves verification. A range expressed in units of B per unit of A does not use the same numbers as its inverse presentation. Copying two bounds without checking this order can create exposure different from what was intended. Before signing, users should therefore read the quantities actually committed and the price direction rather than relying on a chart’s resemblance to an old screenshot.
The guide to a €500 crypto portfolio also helps limit concentration. Several pools may share the same fragile asset or bridge. In 2027, the decisive criterion will not be the most spectacular yield, but the ability to explain the final inventory, net fees and gap versus a comparable holding.
