Impermanent Loss Explained: AMM Liquidity Pools and Fee Break-Even
Impermanent loss describes the difference in value between providing a token pair to an automated market maker (AMM) and simply holding the starting tokens as their relative prices change, excluding fees and other income. In a basic equal-value, constant-product pool, the relative loss can be expressed as 2√r/(1+r) − 1, where r is the change in the price ratio. Trading fees may offset part of this difference, but do not automatically guarantee positive total returns.
Providing liquidity is not the same as placing a passive cash deposit. A pool continuously changes a provider’s token balance as traders swap one asset for another. This creates an economic exposure to relative prices and volatility. This article explains x×y=k, a complete fourfold-price-change example, fee break-even and why concentrated liquidity changes the calculations.
Key takeaways
- Impermanent loss is measured relative to holding the initial token amounts, not simply the pool balance being lower than initial capital.
- Constant-product pools rebalance token amounts through trading.
- For the full-range, equal-weight x×y=k case, the loss function depends on relative price movement.
- Fees can compensate for some loss but are uncertain and may be outweighed by adverse selection.
- Concentrated liquidity, leverage and smart-contract risk make practical outcomes more complex than the simple formula.
The economic role of an AMM
An automated market maker supplies prices according to a mathematical trading rule and an inventory of tokens in a liquidity pool, rather than requiring an order book with a direct counterparty for each trade. Liquidity providers deposit eligible assets and receive a claim on a share of pool assets or its accounting representation. Traders pay fees and shift the pool’s token balance. The provider becomes exposed to inventory changes as the relative prices of the deposited tokens move.
How constant-product pricing works
A conventional two-token constant-product model uses x×y=k, where x and y are the quantities held in the pool and k is constant in the fee-free simplified state. When a trader buys one token from the pool, they supply more of the other to maintain the invariant. The pool’s marginal price follows the ratio of reserves. Arbitrageurs trade against external markets to bring the pool’s effective price closer to prevailing prices, subject to fees and transaction costs.
Why the provider’s inventory changes
Suppose the external price of token X rises relative to token Y. Arbitrageurs can buy X from the pool while supplying Y until prices align, leaving the pool with fewer units of X and more units of Y. The provider thereby sells some of the outperforming token into its rally. If X later falls, the process can reverse. This automatic inventory adjustment is the mechanism behind impermanent loss relative to simply holding the original tokens.
The basic impermanent-loss formula
For a simple equal-value, full-range constant-product position with no fees, the ratio of LP value to HODL value after a relative price change r is 2√r/(1+r). Subtracting one expresses relative loss. When r=1, the value is zero loss; when the relative price doubles, the LP-to-HODL ratio is about 94.28%, a loss of 5.72% against holding. The relation is symmetric under a reciprocal price change, though absolute wealth outcomes differ. Different AMM designs require different formulas.
A concrete initial pool position
Imagine a hypothetical provider deposits 1 ETH worth $100 and 100 USDC, total $200, in a fee-free equal-value constant-product pool. The initial token inventory is 1 ETH and 100 USDC, giving the invariant k=100 under these simplified units. The provider’s economic benchmark is holding 1 ETH and 100 USDC outside the pool. The comparison assumes a stable USDC reference price and excludes gas, taxes, contract losses and reward token emissions.
What happens when ETH quadruples
If ETH rises from $100 to $400, arbitrage changes the pool’s reserves toward 0.5 ETH and 200 USDC, because 0.5×200=100 and the reserve-implied price is 200/0.5=$400. The LP assets are worth 0.5×$400+$200=$400. But the original 1 ETH plus 100 USDC would now be worth $500. The difference is $100, or a 20% shortfall relative to holding. Importantly, the LP made a dollar gain from $200 to $400 even though it underperformed the HODL alternative.
Fees and the break-even question
Swap fees are earned based on applicable pool activity and fee accounting, but high trading volume does not guarantee high net profit. To offset a $100 underperformance against HODL in the illustration, cumulative net LP fee or incentive value would need to be $100, assuming no other changes and the same valuation moment. Fees accrue in assets whose own value changes. Gas to enter, adjust or exit the pool and any protocol charges also matter.
Why high volatility may be difficult for LPs
A volatile asset pair may generate substantial trading fees but also large relative price movements, rebalancing and adverse selection. Volume generated by arbitrage during sharp price changes can occur exactly when a passive liquidity position underperforms a directional holding. A market maker should evaluate fees earned per unit of inventory risk rather than view headline annualised APR as an unconditional yield. Future volume and price variance are uncertain.
Concentrated liquidity requires different analysis
In concentrated-liquidity AMMs, a provider selects one or several price ranges where its capital is active. A narrow range can earn a higher share of fees while the market price remains within it but may leave the position concentrated almost entirely in one token outside the range. The standard full-range equal-weight formula does not describe those outcomes without adjustment. Rebalancing the range also creates fees, gas costs and potentially realised directional trades.
Stablecoin pools are not risk-free
A pair of tokens intended to trade near one dollar can still experience depegging. If one loses its peg, an AMM provider can end up holding disproportionately more of the weaker token as traders exchange it for the stronger asset. The idea that both assets are ‘stable’ does not eliminate smart-contract, issuer reserve, bridge or banking risks. Some specialised stable-swap invariants behave differently from basic x×y=k and require their own risk model.
The effect of token incentives
Some protocols distribute governance or other reward tokens to liquidity providers. The quoted APY may depend heavily on emissions and the market price of incentives. If reward tokens fall sharply, realised economic return can be much lower than the advertised estimate. A liquidity pool can attract fresh deposits after a promotional increase in rewards, diluting each provider’s share of fees. High nominal APR alone does not establish an attractive risk-adjusted outcome.
What a full LP performance report needs
Measure deposited assets at entry, actual token quantities at exit, fees in each currency, incentive token distributions, gas and any changes in protocol obligations. Compare the outcome with holding the initial asset quantities over the same period. Also consider a relevant cash benchmark if the investor’s objective is preserving spending power. Multiple changes in the relative price and range can generate a path-dependent outcome in practical concentrated-liquidity systems.
Worked example
Using the 1 ETH + 100 USDC case, ETH price moving from $100 to $400 changes an otherwise fee-free pool position from 1 ETH and 100 USDC to 0.5 ETH and 200 USDC, worth $400. HODL would be $500. Relative impermanent loss = ($400/$500)−1 = −20%. The formula gives 2√4/(1+4)−1 = 4/5−1 = −20%, confirming the example.
For a price doubling instead, r=2 gives 2√2/3−1 ≈ −5.72%. If a provider’s HODL benchmark value at the later point is $300, the simplified LP value before fees is approximately $282.84. The difference of $17.16 would need to be offset by net fees or rewards for the LP to match the holding result. The actual token balances and fee accrual must be calculated for the specific pool.
Professional comparison table
| Scenario | HODL value, illustrative | LP value excluding fees | LP vs HODL |
|---|---|---|---|
| Initial | $200 | $200 | 0% |
| ETH price doubles | $300 | ≈$282.84 | ≈−5.72% |
| ETH price quadruples | $500 | $400 | −20% |
| Price unchanged | $200 | $200 | 0% |
| Fee income | Accrues only through other strategy | May offset IL | Not guaranteed |
Research and risk checklist
Identify the exact pool invariant, fee tier, token decimals and whether the liquidity range is full or concentrated. Model at least two large opposite price scenarios, stablecoin depeg and reduced trading volume. Compare net performance with holding the initial token portfolio rather than just looking at a positive account P&L.
Include gas, reward token price changes, contract upgrade risk and withdrawability. Review whether fees are auto-compounded or claimable separately. If the position uses leverage, test liquidation thresholds as well as LP inventory outcomes; the standard impermanent-loss formula is not a leverage stress test.
Frequently asked questions
Does impermanent loss mean the LP position lost money?
No. It measures underperformance relative to simply holding the original assets and can occur during positive absolute returns.
How is impermanent loss calculated for a basic AMM?
For a simple 50/50 constant-product pool, use 2√r/(1+r) − 1 for relative price change r, excluding fees.
Can swap fees offset impermanent loss?
They may, but amount and future volume are uncertain.
Is impermanent loss always temporary?
No. Price divergence can persist or change further, and exiting realises the relative outcome.
Do stablecoin pools have impermanent loss risk?
Yes, especially if a token depegs or the pool’s rules create asymmetric inventory exposure.
Sources and further reading
- Uniswap Labs: What Is Impermanent Loss?
- Uniswap Developer Documentation
- Ethereum.org: Decentralised Finance
Further Traders MBA education: DeFi liquidity pools and lending introduction and DeFi borrowing and liquidation risk.
Education only: neither yield, protocol security nor asset returns are guaranteed. Smart-contract, validator, custody and volatile-market risks can cause partial or total loss.