Impermanent Loss Explained With Simple Numbers

Impermanent loss is the value difference between providing two tokens to an automated market maker and simply holding those same tokens when their relative price changes.
The basic example
Suppose ETH is worth $2,000 and you deposit 1 ETH plus $2,000 of USDC into a 50/50 pool. Your position is worth $4,000, split evenly between the two assets.
If ETH rises to $4,000 while the pool follows its constant-product rule, arbitrage traders rebalance the reserves. When you withdraw, you receive roughly 0.707 ETH and $2,828 USDC.
At the new ETH price, that withdrawal is worth about $5,657. If you had held the original 1 ETH and $2,000 USDC, your portfolio would be worth $6,000.
The impermanent loss is therefore about $343, or 5.72%, compared with holding. You did not lose 5.72% of your original deposit; you earned less than the passive alternative because the pool sold part of your ETH as its price increased.
The same effect occurs when ETH falls. Impermanent loss depends on the change in the price ratio, not on whether that change is upward or downward. For a full-range constant-product pool, the no-fee estimate is 2√r/(1+r)-1, where r is the new price divided by the starting price. At 2x, the result is approximately -5.72%.
How to evaluate a pool
The useful calculation starts before you deposit, with the actual pair, range and holding period.
- Record the amount of each token and their prices when you enter.
- Estimate the price ratio at the point where you may withdraw.
- Calculate what the pool would return at that ratio.
- Compare that value with holding the original token amounts.
- Add trading fees and incentives, then subtract gas, management costs and any cross-chain charges.
Fees can outweigh impermanent loss in a busy pool, but the advertised fee rate is not your personal return. Your share depends on the volume that reaches your position, the liquidity competing with you and how long your capital remains active.
The edge case: concentrated liquidity
Uniswap v3 and v4 let you choose a price range instead of supplying liquidity across every possible price. This improves capital efficiency, but it makes the position more directional.
While ETH remains inside your selected range, the position behaves like a concentrated version of the example above and can collect swap fees. If ETH moves above the range, the position becomes almost entirely USDC; if it moves below it, the position becomes almost entirely ETH. Once outside the range, it generally stops earning swap fees until the market returns or you reposition.
That means the familiar 5.72% table is not a reliable answer for a concentrated position. You must inspect the chosen range and ask what inventory you will hold after the price leaves it.
This is where older LP advice has become incomplete. Uniswap v4 went live in 2025 with hooks that can customize fees, swaps and liquidity management. In 2026, live hook designs include systems that source outside liquidity and strategies that move capital into concentrated ranges only when swaps arrive. A pool’s fee tier no longer tells the whole story: check its hook, fee rules and range-management policy. Custom logic may improve capital use, but it does not make price exposure disappear.
If the pool is on Manta Network, treat the journey there as a separate calculation: Symbiosis Finance and Celer Network can be part of cross-chain liquidity workflows, while their fees and slippage occur before the AMM deposit. The transfer step into Manta is handled through the Manta Bridge route.</p