Impermanent Loss: The Hidden Cost of Providing Liquidity

A market stall run by maths

Imagine opening a small stall with a friend. One of you brings apples, the other brings oranges, and together you let customers swap one fruit for the other. For every trade, you collect a small fee. If lots of customers suddenly want apples, your stall ends up with fewer apples and more oranges than when you started. You are still in business, but your shelf now looks different.

Liquidity providers on decentralized exchanges do something similar, except the stall is a smart contract and the price is set by a formula instead of haggling. These systems are called automated market makers, or AMMs. They allow people to trade tokens without waiting for another person to take the other side of the trade. Instead, traders swap against a shared pool of two assets. If a trader wants to buy one token, they add more of the other token to the pool. The classic formula behind many pools is simple: the amount of token A multiplied by the amount of token B must stay the same. That keeps the pool balanced and prices moving with supply and demand.

For providing this inventory, liquidity providers earn a slice of every swap. It sounds like passive income, and sometimes it is. But there is an invisible trade-off hiding in the formula. It is called impermanent loss.

The quiet gap: what impermanent loss really measures

Impermanent loss is not a fee that gets taken out of your account. It is an opportunity cost. It compares two possible worlds: one in which you left your tokens safely in your wallet, and another in which you put them to work in a liquidity pool. If the pool version is worth less, that gap is your impermanent loss.

Here is how it happens. Suppose you deposit one ETH and 2,000 USDC at a time when ETH trades for $2,000. Your deposit is worth $4,000. Later, ETH doubles to $4,000. If you had simply held your tokens, you would have $6,000. But your pool position does not stay at one ETH and 2,000 USDC. Arbitrage traders, spotting that ETH is cheaper in the pool than on other exchanges, buy ETH from the pool until the pool price catches up with the wider market. By then, your position has been rebalanced to roughly 0.707 ETH and 2,828 USDC. At the new price, that is about $5,656.

You have made money against your starting point, but you are about $344 behind the person who just held the tokens. That gap is impermanent loss. It happens because the pool automatically leaves you with more of the asset that did not rise as much, and less of the one that did. Direction does not matter, either. A sharp fall in one token can create a similar gap compared with holding. With the classic full-range formula, a doubling or a halving of one asset relative to the other leaves roughly the same kind of gap: about 5.7% behind a simple hold strategy.

Why the word “impermanent” deserves caution

The loss is called impermanent because it can shrink if prices return to their original ratio before you withdraw. If the two tokens converge again, most of the gap may disappear and you keep the fees you earned along the way. But if you pull your liquidity while prices are far apart, that gap becomes real and lasting.

In practice, the word “impermanent” should be read as “maybe temporary.” Prices do not always return. If you withdraw at the wrong moment, the difference between your pool position and a simple hold is locked in permanently. Many experienced participants prefer calling it divergence loss, because it grows out of the divergence between the two assets, not out of some mysterious pool error.

Concentrated liquidity: precision has a price

Older AMMs spread your capital across every imaginable price, from zero to infinity. That sounds safe, but it is also wasteful. In a stablecoin pair like DAI/USDC, where both tokens usually trade near one dollar, only a tiny slice of the deposited capital ever sees real trading. The rest sits idle.

Uniswap v3 and similar designs introduced concentrated liquidity. As a liquidity provider, you choose a price range. Your capital works only inside that range. If the price stays within it, a much smaller deposit can earn fees comparable to a much larger, full-range position. You can set a tight band around the current price, or spread several positions across different prices. This is genuinely powerful, and it made trading on many pools cheaper and more efficient.

But concentrated liquidity is a sharper double-edged sword. Inside your chosen range, your effective exposure is larger than your actual deposit, so a given price move can have a bigger effect on your position than it would in a diluted pool. If the price leaves your range entirely, your liquidity goes inactive. You stop earning fees. And because the pool has rebalanced along the way, your position may now consist of just one token, the one that became less valuable relative to the pair. Your liquidity stays in this single-asset state until the price re-enters your range, or until you actively move your range to follow the market.

A narrow position placed above or below the current price can even behave like a limit order, converting entirely from one token to another as the price crosses your range. If the price later returns, the position can convert back automatically unless you withdraw it. That means someone has to watch the position, or delegate that watching to an automated service. The old “set it and forget it” style of liquidity provision becomes much harder.

Fees, volume, and the comparison that matters

Impermanent loss is not the whole story. Fees are the other half. A high-volume pool can generate enough swap fees to outweigh a painful-looking gap, while a quiet low-volume pool may not. This is why simple loss calculations can be misleading if they ignore fees.

Even the direction of arbitrage is more interesting than it first appears. Arbitrage traders cause the rebalancing that leads to impermanent loss, but every arbitrage trade also pays a fee. Dynamic simulations of these markets suggest that a healthy, easy-to-arbitrage environment can actually help liquidity providers, because the fee income from arbitrage activity outweighs the rebalancing effect in moderate conditions. Significant, sudden, and lasting price moves are where the pain concentrates.

The ingredients that matter are price volatility, asset correlation, pool depth, trading volume, and fee level. Stablecoin pairs and closely correlated assets tend to create very little divergence loss. Pairs with one volatile asset, like ETH against a stablecoin, create more. A higher fee tier will not rescue you if no one trades there, and a low fee tier can still pay well if trading volume is enormous. There are online calculators that compare the value of holding versus providing liquidity, but remember to add fee income before drawing conclusions.

A final word before you provide liquidity

Impermanent loss is not a bug in automated market makers; it is a built-in feature of the design. It is the rent you pay, in missed upside, for the right to earn swap fees. That trade-off can be worthwhile, especially in active pools, but it is never a free lunch.

Keep your early experiments small enough to learn from. Choose asset pairs whose behavior you understand. If you use concentrated liquidity, set a range you can monitor and know what happens when the price leaves it. And since the rules and tax treatment around liquidity positions vary by country, and are still evolving, a trusted tax adviser is a better guide than a quick online search.

In the end, the goal is not to avoid impermanent loss entirely. It is to understand it well enough that the fees you earn are an honest reward for the risk you are actually taking.