Skip to content
OneKitly

Impermanent Loss Explained: What Providing Liquidity Really Costs

Published 5/13/2026 · 6 min read · Finance calculators

Camille Laurent

Camille LaurentFinance writer at OneKitly

Tax · Personal finance

Checked against 2 sources

View profile
In short

In a constant-product pool holding two assets in equal value, impermanent loss depends only on the ratio r between the two tokens' price changes: IL = 2 × √r ÷ (1 + r) − 1. If one token doubles against the other, r = 2 and the loss is 5.72 percent; if it quadruples, r = 4 and the loss is exactly 20 percent; at r = 5 it is 25.46 percent. The loss is symmetric — a token that halves gives the same 5.72 percent as one that doubles — and it is measured against holding, not against your deposit. Put $10,000 into an ETH/USDC pool as 2 ETH at $2,500 plus $5,000, and if ETH doubles you withdraw $14,142 while a holder has $15,000: an $858 gap. Nothing makes that gap come back once you withdraw, so trading fees earned while the position was open have to exceed it for providing liquidity to have been worth doing.

Impermanent loss is not a fee and it is not temporary: it is the gap between your liquidity position and simply having held the two tokens. Here is the formula, a table of price change against loss, and the fee income you would need to come out ahead.

The formula, and why the loss is symmetric

A constant-product pool keeps the product of its two reserves fixed, which forces it to sell whichever token is rising and buy whichever is falling. That is the entire mechanism: the pool rebalances against you, automatically and continuously. Because the rebalancing depends only on how far the price ratio has moved, the loss depends only on r — the price of token A divided by the price of token B, expressed relative to where the ratio stood on the day you deposited. IL = 2 × √r ÷ (1 + r) − 1 comes straight out of that constraint, with no assumption about volatility, trading volume or time.

The symmetry surprises people. Feed r = 2 into the formula and you get −5.72 percent; feed r = 0.5 and you get −5.72 percent again. A token that doubles and a token that halves cost the liquidity provider exactly the same, because 2 and 0.5 are the same distance from parity in ratio terms. That is also why the loss grows so slowly at first and then bites: a 25 percent move costs 0.62 percent, a 50 percent move 2.02 percent, a doubling 5.72 percent, a five-fold move 25.46 percent. Small drifts are almost free; a token that runs away from its pair is where the money goes.

Nothing about it is impermanent

The name is the most misleading word in decentralised finance. It survives because the loss does unwind if the price ratio returns exactly to where it stood when you deposited — and that is true. But it is a conditional promise about the future, not a description of your position. The moment you withdraw, the pool hands you whatever token amounts the current ratio implies, and the gap against holding is realised in full. Nothing about it is more temporary than any other loss you choose to take.

It is also not a fee. Nobody deducts it, no line item shows it, and the pool interface will happily report a position that has grown in dollar terms while it has fallen behind holding. In the worked example the position went from $10,000 to $14,142 — a 41 percent gain that a dashboard will display in green — and still cost $858 against doing nothing. Impermanent loss is an opportunity cost you only see if you compute the counterfactual yourself, which is exactly why so many providers never notice it.

Fees are the only reason the position can pay

Liquidity provision is a trade: you accept the rebalancing loss in exchange for a share of the swap fees. So the only question that matters is whether the fees beat the gap. At a 20 percent annualised fee return — a healthy figure for a major pair — it takes about 104 days of fee income to cover the 5.72 percent gap that a doubling creates, and about 209 days at 10 percent. If the pair moves that far in a fortnight, the fees never had time.

The trap is that the two sides move together. Volatility is what generates swap volume, and volatility is what pushes r away from 1 — so the pools advertising the highest fee returns are usually the ones where the ratio is moving fastest. Concentrated liquidity sharpens both edges: inside your range you earn far more fees per unit of capital, and outside it you hold 100 percent of the losing token with the loss no longer capped by the gentle curve. Run the numbers on the pair you are actually considering, over the period you actually intend to stay in, before you accept the headline yield as the return.

Ratio r
Impermanent loss by price ratio, from IL = 2√r ÷ (1 + r) − 1, with the gap in money per $10,000 you would have held
Price change of one token vs the otherRatio rImpermanent lossGap per $10,000 held
−50 % (halves)0.5−5.72 %−$572
−25 %0.75−1.03 %−$103
−10 %0.9−0.14 %−$14
0 % (unchanged)10.00 %$0
+25 %1.25−0.62 %−$62
+50 %1.5−2.02 %−$202
+100 % (doubles)2−5.72 %−$572
+200 %3−13.40 %−$1,340
+300 %4−20.00 %−$2,000
+400 %5−25.46 %−$2,546
+900 %10−42.50 %−$4,250

Worked with our own calculator

Impermanent loss calculator

Given

Deposit value
$2,000.00
Token price at deposit
$4,000.00
Current price
$6,000.00

Result

Price change
50%
Impermanent loss
-2.02%
Value if held (HODL)
$2,500.00
Value in the pool
$2,449.49
Loss vs holding
-$50.51
Fee yield to break even
2.02%

These figures are produced by the calculator below, not typed in by hand — they are recomputed whenever the tool changes.

Run it on your own figures

Frequently asked questions

Is impermanent loss deducted from my position?
No. It is not a charge and it does not appear on any statement. Your position can be worth more than you deposited and still carry a large impermanent loss, because the comparison is against holding the same two tokens untouched, not against your entry value. The only way to see it is to compute both numbers side by side.
Do stablecoin pairs have impermanent loss?
Yes, but a negligible one while both pegs hold: a 10 percent divergence gives r = 0.9 and a loss of 0.14 percent. The risk is not the day-to-day drift, it is the depeg. A stablecoin that loses half its value takes the pair to r = 0.5 and the same 5.72 percent as any doubling, on top of the fact that you now hold mostly the broken token.
Can I avoid impermanent loss entirely?
Not in a constant-product pool holding two assets whose prices can diverge — the loss is the mechanism, not a defect. You can shrink it by pairing assets that move together, which keeps r near 1, and some protocols offer single-sided positions where a counterparty absorbs the divergence for a price. The only genuinely zero-loss case is a ratio that ends exactly where it started, and you cannot arrange that in advance.

Articles you may find interesting

All guides
GuideYield Farming: What an Advertised APY Actually PaysThe number on the farm's front page is a gross figure before every subtraction. Here is a 120 percent APY walked down, one layer at a time, to the 33.6 percent that actually landed — plus why APR and APY are not the same number.ExplainerWhat Is Slippage in Crypto? Price Impact, Tolerance and What It CostsPrice impact is arithmetic: on a constant-product pool it equals your trade size divided by the reserve plus your trade. Here is the formula, a table of trade size against impact, and why the setting called slippage tolerance changes none of it.ComparisonCrypto Market Cap vs Trading Volume: What Each Number Can and Cannot Tell YouMarket cap is price times circulating supply — an arithmetic product, not money invested. Volume is what actually changed hands. Here is what each measures, how their ratio exposes a thin market, and where fully diluted valuation fits.ExplainerHow Staking Rewards Actually Work: Nominal Rate, Compounding, and What Eats ItAn advertised 8 percent becomes 8.33 percent once daily rewards compound — and then 7.46 percent after a 10 percent validator commission, and less again after unbonding time. Here is each step, with the arithmetic laid out.ExplainerVolatility Is Not Risk, and the Square Root of Time Is a ChoiceAnnualised volatility = period standard deviation × √(periods per year), and that √t scaling assumes independent increments. It is a model, not arithmetic: at a daily autocorrelation of 0.1 a 60 percent annualised figure should read 66.3. The payload is volatility drag — the arithmetic mean exceeds the geometric by about σ²/2, so at 8 percent average return and 40 percent volatility the compound outcome is zero.ExplainerThe Sharpe Ratio, and What It Quietly AssumesSharpe = (return − risk-free) ÷ standard deviation, so it prices return per unit of volatility — and volatility is symmetric. Two funds can share a Sharpe of 0.4939 while their Sortino ratios are 8.59 and 0.74. Annualising by √12 assumes independent returns: at an autocorrelation of 0.2 the published figure is 20 percent too high.

Related tools

This article explains how a calculation works. It is not investment advice. Leveraged trading and mining can lose more than you put in, and past results say nothing about future ones.

Sources

Spotted a mistake in this article?