Impermanent loss is one of those DeFi concepts where people nod along in conversation and then quietly google the formula later. The term itself is slightly misleading — it implies the loss will eventually disappear, which is only true under specific conditions — and most explanations stop at "prices diverged, you lost value" without showing you the actual number.
This post covers the calculation: what the formula is, how to use it, and how to weigh it against fee income, which is the part that actually determines whether providing liquidity was worth it.
The mechanics of why impermanent loss occurs — the constant product formula, how rebalancing happens automatically — are a separate topic. This is specifically about calculating the magnitude.
When you deposit into a 50/50 AMM liquidity pool — the most common type — you deposit equal values of two assets. Say $1,000 of ETH and $1,000 of USDC.
If you'd held those assets in a wallet, their combined value would track market prices directly. Inside the pool it's different. The AMM automatically rebalances your position as prices change, always maintaining the 50/50 ratio. That rebalancing is what creates impermanent loss.
Impermanent loss is not "how much money you lost." It's the difference between what your LP position is worth versus what it would have been worth if you'd simply held the same assets and done nothing. It measures underperformance relative to a hold, not absolute loss.
For a standard 50/50 constant-product pool (Uniswap v2-style), the impermanent loss formula is:
IL = 2 × √(P_ratio) / (1 + P_ratio) − 1
Where P_ratio is the ratio of the final price to the original price for the more volatile asset.
Running the numbers:
The loss is symmetric. A 50% price drop causes the same IL as a 100% price rise. A halving (P_ratio = 0.5) produces the same ~5.7% IL as a doubling (P_ratio = 2.0). The direction of the move doesn't matter — only the magnitude.
The loss is called impermanent because it only crystallizes when you withdraw. If ETH doubles while you're in the pool, you're sitting on ~5.7% IL — but if ETH then falls back to the original price, that IL goes to zero. You'd exit at the same value you'd have gotten from holding.
In practice, though, the word "impermanent" implies a reliability of recovery that doesn't match real LP experience. For volatile/stablecoin pairs like ETH/USDC, the price ratio very rarely returns exactly to where it started. The loss is technically impermanent in that it fluctuates — but it isn't reliably temporary.
If you withdraw at any price other than your original entry ratio, the IL is real. Plan accordingly.
Impermanent loss doesn't happen in isolation. Every swap through the pool generates a trading fee, and liquidity providers earn a share proportional to their pool stake. The actual P&L of an LP position is:
Net return = Fee income − Impermanent loss
Whether providing liquidity made financial sense depends entirely on whether fees exceeded IL over the holding period.
Two reference points help set expectations:
Volatile/stablecoin pools (ETH/USDC, SOL/USDC): High IL risk because prices diverge significantly over any meaningful holding period. These pools need substantial trading volume to generate enough fee income to compensate. High-activity pools with 0.3% fee tiers can sometimes cover IL — but it isn't guaranteed and requires monitoring.
Stablecoin/stablecoin pools (USDC/DAI, USDC/USDT): Near-zero IL because the price ratio barely moves. Even modest fee income produces net positive returns. This is why Curve-style stablecoin pools historically had reliable positive LP returns — the IL denominator is effectively zero, so any fee income compounds without being offset.
You can calculate this with actual on-chain data using tools like Uniswap Analytics, APY.vision, or DeBank's LP tracker. They show fee earnings vs. IL over any time window — more useful than a theoretical estimate.
The formula above applies to standard 50/50 constant-product pools. Two common variations change the math:
Uniswap v3 (concentrated liquidity): You select a price range instead of providing liquidity across the entire curve. When price stays in range, fee earnings are amplified significantly. But IL is also amplified if price moves sharply through your range. A narrow range set around the current price will suffer more IL than a v2 equivalent if price moves 3x outside that band.
Non-50/50 pools (Balancer, etc.): An 80/20 pool has lower IL than a 50/50 equivalent for the same price move, because less rebalancing occurs. The formula is pool-specific.
If you're not in a standard 50/50 v2-style pool, use a dedicated calculator for that pool type rather than applying the formula above directly.
The shift toward concentrated liquidity — Uniswap v3, Curve v2, Ambient Finance — has changed the IL calculus meaningfully. Passive LP in v2-style pools was straightforward: full-range positions, predictable IL, fee income that compounded over time. Concentrated LP in v3 is more active. Ranges need to be monitored and adjusted as prices move.
A position set outside a narrow price band earns zero fees while still carrying full price exposure. At that point, you'd have been better off just holding.
Automated range management protocols — Gamma, Arrakis, Bunni — handle rebalancing in exchange for a fee. They reduce the active management requirement, but add protocol risk and manager performance risk in return.
Fee income tracking showing net-positive returns over full calendar months in specific pools. Stablecoin/stablecoin pools continuing to demonstrate near-zero IL under real on-chain data. Automated LP management protocols showing consistent net-positive performance across full market cycles, not just bull periods.
A stablecoin depeg — USDC or DAI losing parity — would produce significant IL in stablecoin pools, breaking the assumption of near-zero price movement. Apparent positive returns masked by high fee income that drops when pool volume falls, leaving crystallized IL behind.
Now: The 50/50 constant-product formula is the right place to build intuition. Run the numbers before entering any new LP position. Fee income vs. IL is the first question to ask.
Next: Concentrated liquidity is increasingly the standard at major DEXs. Understanding range mechanics matters if you're providing liquidity in v3-style pools.
Later: Automated LP optimization tools are developing, but still early enough to treat with caution. The IL math doesn't change — the management layer is what's evolving.
This covers the impermanent loss calculation for standard 50/50 AMM pools and the framework for weighing it against fee income. It doesn't constitute advice about whether to provide liquidity in any specific pool, predict fee income, or assess protocol-level risk at any particular DEX.
The IL formula is a tool for setting expectations before entering a position — not a complete risk model. Smart contract risk, liquidity risk, and protocol governance are separate dimensions the formula says nothing about.




