Buy Crypto
Markets
Spot
FuturesBTC
Earn
Event Center
Rewards Hub
More
100%+ PoR
Overview In a 50∶50 constant product pool, a doubling in the ETH price leaves a liquidity position worth 5.72% less than the same assets held directly, and that shortfall is impermanent loss. On a $10Overview In a 50∶50 constant product pool, a doubling in the ETH price leaves a liquidity position worth 5.72% less than the same assets held directly, and that shortfall is impermanent loss. On a $10

ETH USDC Impermanent Loss: How to Calculate Pool Returns vs. Holding

Overview

 
In a 50∶50 constant product pool, a doubling in the ETH price leaves a liquidity position worth 5.72% less than the same assets held directly, and that shortfall is impermanent loss. On a $10,000 deposit, the pooled assets are worth $14,142.14 at that point against $15,000 for holding, a difference of $857.86. Note that the pooled position is still up 41.42% on the original deposit, because impermanent loss measures the gap against holding rather than a loss of capital, and that distinction is the single most frequently inverted point in the subject.
 
The calculation takes one variable. Impermanent loss equals two times the square root of the price ratio, divided by one plus the price ratio, minus one. The price ratio is the current price divided by the entry price, so the percentage depends only on the size of the price move and not on where ETH trades in absolute terms.
 
 

Key Takeaways

 
A doubling and a halving produce exactly the same impermanent loss of 5.719%, because the formula returns an identical value for a price ratio and its reciprocal. Small moves produce small gaps, with a 10% move in either direction costing 0.139%, or $13.17 on a $10,000 position, while larger moves accelerate, reaching 13.397% at three times the entry price and 20.000% at four times. The mechanism can be stated precisely: as the price rose from $1,824.80 to $3,649.60, the pool sold 0.802535 ETH for $2,071.07, an average of $2,580.66, which is exactly the geometric mean of the start and end prices, against an arithmetic mean of $2,737.20, and the distance between those two averages is the entire source of the loss. Fees and gas sit at completely different magnitudes: Etherscan read gas at 0.175 gwei on November 16, putting a swap near $0.112, while the gap from a single 10% price move is 25.8 times the cost of an entire entry-and-exit gas cycle on the same basis. On concentrated liquidity there is an exact identity: the capital efficiency multiplier a narrower range delivers and the factor by which it amplifies impermanent loss are the same number, both 21.49 times for a range of $1,659 to $2,007.
 

The Model, the Inputs, and the Starting Position

 
This calculation holds only under one specific model, so the model and its inputs come first.
 
The model is a 50∶50 constant product pool, meaning the product of the two asset quantities stays constant, which is the base form of a Uniswap v2 style automated market maker. It does not apply to unequal weights such as the 80∶20 structures Balancer supports, nor to invariants designed for correlated assets such as the curve Curve uses for stablecoin pairs, which compresses the asset rebalancing that price divergence causes by design. The USDC here is issued by Circle and functions purely as the unit of account, with its reserve structure entering none of the steps below.
 
The 50∶50 refers to equal value rather than equal quantity, which is where starting positions are most often set up wrongly. Per Coinglass at 14:59 on November 17 (UTC+8), ETH was at $1,824.8, with a daily high of $1,866.2 and a low of $1,786.1. Splitting $10,000 equally by value gives 2.7400 ETH and $5,000 of USDC, each side worth $5,000.
 
Input
Value
Source
Date and basis
ETH price anchor
$1,824.8
Coinglass
Spot quote, 14:59 UTC+8 November 17
Principal
$10,000
Set for this worked example
Split 50∶50 by value
Gas price
0.175 gwei
Etherscan
Reading at 09:07 UTC November 16
Per-action gas cost
Swap $0.17, add liquidity $0.30
Datawallet
That site's own table, 0.5 gwei and ETH at $1,900
Pool size and turnover
TVL $36.83M, daily volume $85.06M
Trading Strategy
ETH-USDC 0.05% fee tier, single pool
Chain-wide size
TVL $800.52M, daily fees $290,522
DefiLlama
November 10, all v3 pools on Ethereum
 
Those two sets of pool figures come from different providers with different coverage, so every division below stays within one set and never crosses between them.
 
The impermanent loss percentage depends only on the price ratio and not on the absolute price level, so substituting a different price source changes no percentage in the table below and only changes the dollar shortfall. The quote on the live ETH price page can go straight into the formula as a starting point, and a price read on MEXC or any other quote source works through the identical arithmetic.
 
The recent price backdrop needs stating. ETH went through a deep drawdown in November, and per Fortune's daily price record it stood at $1,953.96 at 8:30 a.m. ET on November 13 against $2,037.25 the previous day, while BitcoinWorld's November 17 compilation put the weekly decline near 14.7% through November 16 with a weekend low around $1,712. That range means moves of more than 10% in either direction are ordinary in current conditions rather than extreme cases.
 

What Happens Inside the Pool After the Price Moves

 
A constant product pool does not hold fixed quantities. It adjusts both sides passively on every trade, so after a price move the pool holds different amounts of ETH and USDC than were deposited.
 
Entering with 2.7400 ETH and $5,000 of USDC gives a product of 13,700.13, and that product stays constant while nobody adds or removes liquidity. After the price becomes r times its original level, the pool's ETH equals the square root of the product divided by the new price, and the pool's USDC equals the square root of the product multiplied by the new price.
 
ETH move
ETH price
Pool ETH
Pool USDC
Pool value
Holding value
Impermanent loss
-50%
912.40
3.874982
3,535.53
7,071.07
7,500.00
5.719%
-30%
1,277.36
3.274958
4,183.30
8,366.60
8,500.00
1.569%
-20%
1,459.84
3.063443
4,472.14
8,944.27
9,000.00
0.619%
-10%
1,642.32
2.888241
4,743.42
9,486.83
9,500.00
0.139%
Unchanged
1,824.80
2.740026
5,000.00
10,000.00
10,000.00
0.000%
+10%
2,007.28
2.612513
5,244.04
10,488.09
10,500.00
0.113%
+20%
2,189.76
2.501290
5,477.23
10,954.45
11,000.00
0.414%
+50%
2,737.20
2.237222
6,123.72
12,247.45
12,500.00
2.020%
+100%
3,649.60
1.937491
7,071.07
14,142.14
15,000.00
5.719%
+200%
5,474.40
1.581955
8,660.25
17,320.51
20,000.00
13.397%
+300%
7,299.20
1.370013
10,000.00
20,000.00
25,000.00
20.000%
 
Working each row out from the asset quantities gives the same percentage the formula returns, on every row, and that set of percentages also matches the seven data points Uniswap's own documentation lists in its note on understanding returns, namely 0.6% at 1.25x, 2.0% at 1.5x, 3.8% at 1.75x, 5.7% at 2x, 13.4% at 3x, 20.0% at 4x and 25.5% at 5x. Other onchain market makers use the same formula, and the derivation in Raydium's documentation is identical, which indicates this is not one protocol's particular convention.
 

Symmetric, But Not Because Anything Cancels

 
One feature of the table invites misreading. A 50% fall and a 100% rise both produce 5.719%, three times and one third both produce 13.397%, and four times and one quarter both produce 20.000%. Nothing is cancelling out. The formula simply returns the same value for a price ratio r and for 1/r, which can be verified algebraically.
 
The practical consequence is that both directions have to be calculated. On the way down the pool's ETH rises to 3.874982 and its USDC falls to $3,535.53, meaning the pool bought ETH continuously as the price declined. On the way up the ETH falls to 1.937491 and the USDC rises to $7,071.07, meaning it sold continuously. In either direction the provider ends up holding more of whichever asset performed worse.
 

Where the Shortfall Comes From

 
The source of the shortfall can be computed exactly rather than described qualitatively.
 
As the price moved from $1,824.80 to $3,649.60, the pool sold 0.802535 ETH in total and received $2,071.07, an average of $2,580.66. That figure is exactly the geometric mean of $1,824.80 and $3,649.60, while the arithmetic mean of the two is $2,737.20. A constant product pool moving in one direction always realises an average execution price equal to the geometric mean, and the geometric mean is never above the arithmetic mean, so that gap is the whole source of impermanent loss.
 
The downward direction works identically. Moving from $1,824.80 to $912.40, the pool bought 1.134956 ETH for $1,464.47, an average of $1,290.33, which is again the geometric mean of those two prices. The average purchase price sits above the ending price, so the same shortfall appears. For a more compact statement of the mechanism, the impermanent loss entry sets out the definition.
 

Fees and Gas Sit Three Orders of Magnitude Apart

 
Once fee income and gas costs enter the calculation, the two affect the conclusion to very different degrees, so they need computing separately rather than grouping together as costs.
 
Gas is barely a constraint at current levels. Etherscan's onchain gas reading at 09:07 UTC on November 16 was 0.175 gwei, with low, average and high at 0.175, 0.175 and 0.185 gwei respectively and an estimated swap cost of $0.112, with ETH at $1,798.08 at the time. Datawallet's cost table uses a basis of 0.5 gwei and ETH at $1,900 and gives $0.04 for an approval, $0.30 for adding liquidity and $0.17 for a swap, summing to $0.51, which is 0.0051% of a $10,000 position. That basis uses a gas price 2.86 times Etherscan's reading for the day, so $0.51 can be treated as a conservative ceiling.
 
Setting that ceiling against the table above shows the difference in magnitude. A 10% price move produces a $13.17 shortfall, 25.8 times the entire gas cycle; a 20% move produces $55.73, 109.3 times; a doubling produces $857.86, 1,682.1 times. Another reference point: the same 688,000 units of gas converts to $37.11 at 30 gwei holding the ETH price constant, or 0.371% of the position, against $0.22 and 0.00216% at 0.175 gwei. The old guidance that gas makes small positions uneconomic has therefore had its binding constraint move to impermanent loss and range management. On converting gwei into an actual fee, how to read gwei works through the arithmetic.
 
Fee income sits at the same order of magnitude as impermanent loss and needs computing carefully. Per Trading Strategy's ETH-USDC 0.05% fee tier page, that pool's TVL was $36.83M against daily volume of $85.06M, a daily turnover of 2.31 times, with a contract address of 0x88e6a0c2ddd26feeb64f039a2c41296fcb3f5640. At that tier's 0.05% rate, fees for the day come to roughly $42,530, or 0.1155% of TVL. For comparison, DefiLlama recorded all v3 pools on Ethereum at $800.52M of TVL and $290,522 of daily fees on November 10, which is 0.0363% of TVL at a daily turnover of 0.51 times. The two differ by nearly three times, because a single tight pair turns over far faster than the chain-wide average, so a chain-wide ratio cannot be used to estimate a single pool's return.
 
Working from that single pool's daily ratio, and conditional on volume holding and the position remaining in an active range, offsetting the 0.139% from a 10% move takes roughly 1.2 days, the 0.619% from a 20% move roughly 5.4 days, and the 5.719% from a doubling roughly 49.5 days. That is division on a single day's ratio rather than an annualised yield, and either falling volume or a position leaving its range lengthens the actual count.
 

Concentrated Liquidity Amplifies the Same Multiple on Both Sides

 
Concentrated liquidity confines capital to a price range so that the same principal supplies deeper quotes, which is its design purpose. Its effect on impermanent loss has a result that can be derived exactly.
 
Take a range running from the entry price divided by n to the entry price multiplied by n, a structure that happens to produce an exact 50∶50 value split at entry. Under it, the capital efficiency multiple equals one divided by one minus n to the power of negative one half, while the amplification of impermanent loss relative to a full range equals the square root of n divided by the square root of n minus one. Those two expressions are algebraically identical, and checking them numerically shows agreement to twelve decimal places. The capital efficiency a narrower range buys and the impermanent loss it amplifies are therefore the same multiple, and no configuration improves one side alone.
 
Price range
Corresponding move
Capital efficiency
Loss amplification
Impermanent loss at ETH +5%
1,658.91 to 2,007.28
-9.1% to +10.0%
21.49x
21.49x
0.639%
1,459.84 to 2,281.00
-20.0% to +25.0%
9.47x
9.47x
0.282%
1,216.53 to 2,737.20
-33.3% to +50.0%
5.45x
5.45x
0.162%
912.40 to 3,649.60
-50.0% to +100.0%
3.41x
3.41x
0.102%
 
A full-range position in the same scenario loses 0.030%, and dividing each row above by that figure returns precisely that row's amplification factor. One further point deserves attention: within the range this factor is independent of the size of the price move, so a range with n equal to 2 amplifies by 3.4142 times whether the price moves 1% or 5%.
 

Two Things Happen at Once Outside the Range

 
Behaviour outside the range differs completely from behaviour inside it, and this is the genuine discontinuity in concentrated liquidity calculations.
 
Once the price crosses the upper bound the position becomes entirely USDC, and below the lower bound entirely ETH. Further movement in the same direction no longer changes the position's value while the comparison portfolio keeps changing, so the gap continues widening. Taking the $1,459.84 to $2,281.00 range as the example, a doubling to $3,649.60 produces an impermanent loss of 29.399%, far above the 5.719% a full range produces in the same scenario.
 
More importantly, a position outside its range earns no fees at all. The offsetting calculation in the previous section requires an active range, and the scenarios with the largest impermanent loss are precisely the scenarios in which the position has already left it. Fees and impermanent loss are therefore not two items that hedge each other under the same conditions but two that take effect under opposite ones, which is why applying full-range fee experience to a narrow range produces wrong conclusions.
 

Five Common Misconceptions

 

Impermanent Loss Means Losing Money

 
Impermanent loss measures the gap against holding, not a gain or loss on principal. When ETH doubles, a $10,000 position is worth $14,142.14 in the pool, up 41.42% on the deposit, and simultaneously $857.86 below holding. Both are true at once, and only the second is impermanent loss. The reverse holds too: when ETH halves the pooled position is worth $7,071.07, down 29.29% on the deposit, of which only $428.93 is attributable to the comparison with holding and the rest is ETH's own decline.
 

The Loss Fully Disappears If the Price Returns

 
Impermanent loss does go to zero when the price returns to the entry level, which is what the word impermanent refers to. One item does not come back with it: the fees a concentrated position did not earn while outside its range. The longer a position sits outside, the larger that permanent gap, and because it does not appear in the impermanent loss formula it has to be accounted for separately.
 

Fees Will Always Outweigh Impermanent Loss

 
Whether they do depends on the number of days spent in an active range and on volume over that period, neither of which is known in advance. Working from the single pool daily ratio above, offsetting a 20% move takes roughly 5.4 days and a doubling roughly 49.5 days, throughout which the position must stay active and volume must hold. Narrowing the range raises fee density per unit of time, raises impermanent loss by the identical multiple, and shortens the time the position stays in range.
 

Gas Is the Main Obstacle for Small Positions

 
Not at current gas levels. On Datawallet's own basis, an approval, an add-liquidity transaction and an exit swap total $0.51, or 0.0051% of a $10,000 position, while the shortfall from a single 10% price move is 25.8 times that figure. The guidance held at 30 gwei, where the same gas usage converted to $37.11, but at the 0.175 gwei Etherscan read on November 16 it is no longer a deciding cost. What needs controlling is the range judgment behind rebalancing frequency rather than the fee on any individual transaction.
 

The Stablecoin Side Carries No Risk

 
The USDC quantity changes with the price, which is itself where the exposure sits. When ETH rises 100%, the pool's USDC goes from $5,000 to $7,071.07, meaning the position sold 0.802535 ETH into the rally at an average of $2,580.66. Nothing was lost in dollar terms, but the holder was passively reduced in the asset that appreciated. Supplying a liquidity pool means letting that set of passive trades execute automatically by rule, and that set always trades against the price trend.
 

Exclusive View from the MEXC Editorial Team

 
The result worth retaining from this calculation is not the figure 5.72% but the mechanism that produces it. A constant product pool moving in one direction always realises an average execution price equal to the geometric mean of the start and end prices, and the geometric mean is never above the arithmetic mean, with the gap widening as the divergence grows. Over the move from $1,824.80 to $3,649.60, the pool's realised average sale price was $2,580.66 against an arithmetic mean of $2,737.20. Impermanent loss is not a friction cost or a protocol take. It is the distance between those two averages, which is why no change of protocol, parameter or timing removes it and only a smaller price divergence reduces it.
 
The concentrated liquidity figures settle a widely repeated judgment. The capital efficiency multiple and the loss amplification factor are the same algebraic expression, both 21.49 times for the $1,659 to $2,007 range and both 9.47 times for $1,460 to $2,281. Narrowing a range is not trading a little extra risk for a lot of extra return but scaling both sides by one identical proportion. Beyond that, the multiple is independent of the size of the price move while the position stays in range, and once the price crosses a bound the position stops earning fees while the gap keeps widening. Fees and impermanent loss are consequently not two items that hedge under the same conditions but two that take effect under opposite ones.
 
Gas deserves separate comment because its role changed materially over the past year. At the 0.175 gwei Etherscan read on November 16, a complete cycle of entering, collecting and exiting comes to $0.51 even on a conservative basis, or 0.0051% of a $10,000 position, where the same gas usage cost $37.11 in a 30 gwei environment. Carrying the old guidance forward points attention at the wrong variable: once per-transaction cost falls from four tenths of a percent of the position to two hundred-thousandths, what determines the outcome is entirely the range setting and the holding period rather than the number of transactions.
 
A final point on basis. The ratio of daily fees to TVL is 0.1155% on a single tight pair and 0.0363% across all v3 pools on the chain, a difference of nearly three times, driven by daily turnover of 2.31 times against 0.51 times. Using the chain-wide ratio to estimate a single pool understates it and using the single pool ratio to estimate the chain overstates it, while multiplying either by 365 embeds two assumptions, that volume and range validity both hold for a full year. The figures given here are a single day's ratio and the day counts derived from it, and both conditions have to be carried along when substituting fresh inputs.
 

FAQ

 

What is impermanent loss

 
Impermanent loss is the gap between the value of a liquidity position in a 50∶50 constant product pool and the value of the same assets held directly. It arises from the price diverging from the entry level and returns to zero when the price comes back, which is what the word impermanent denotes. It measures opportunity cost relative to holding and is not equivalent to a loss of principal.
 

How do you calculate impermanent loss for an ETH USDC pool

 
Apply the formula: impermanent loss equals two times the square root of the price ratio, divided by one plus the price ratio, minus one. The price ratio is the current ETH price divided by the entry price. At a price ratio of 2, the result is two times 1.41421 divided by three, minus one, giving negative 5.719%. The percentage depends only on the price ratio and is independent of the absolute price level and of the size of the principal.
 

How much is impermanent loss when ETH doubles

 
5.719%. Entering at $1,824.8 with $10,000, a rise to $3,649.60 leaves 1.937491 ETH and $7,071.07 of USDC in the pool, totalling $14,142.14, against $15,000 for holding, a shortfall of $857.86. Note at the same time that the pooled position remains up 41.42% on the original deposit.
 

Does impermanent loss disappear

 
It disappears entirely when the price returns to the entry level. One item does not recover with it, namely the fees a concentrated position failed to earn while outside its price range, which is a permanent gap, does not appear in the impermanent loss formula, and has to be accounted for separately.
 

Is providing liquidity always worse than holding

 
Before fees, a pooled position sits below direct holding at any degree of price divergence and equals it only at zero divergence. After fees the answer depends on days in range and volume over the period. Working from a single tight pair's November daily ratio, offsetting a 20% move takes roughly 5.4 days of fees and a doubling roughly 49.5 days, throughout which the position must remain in an active range.
 

Can fee income cover impermanent loss

 
It depends on the time the position spends in an active range and on volume over that period, neither known in advance. More importantly, the scenarios with the largest impermanent loss are frequently the ones where the position has already left its range and earns no fees at all, so the two do not offset under the same conditions.
 

Will gas fees eat the returns on a small position

 
The effect is minor at current gas levels. Etherscan read 0.175 gwei on November 16 with a swap near $0.112, and on Datawallet's conservative 0.5 gwei basis an approval, an add-liquidity transaction and an exit swap total $0.51, or 0.0051% of a $10,000 position, roughly one twenty-sixth of the shortfall from a single 10% price move. The older guidance held at 30 gwei, where the same gas usage converted to $37.11.
 

Does narrowing the price range increase returns

 
Narrowing a range raises capital efficiency and impermanent loss by one identical multiple, the two being the same algebraic expression. For the $1,659 to $2,007 range both are 21.49 times, and for $912 to $3,650 both are 3.41 times. A narrow range is also crossed more easily, and once crossed the position stops earning fees while the gap keeps widening.
 

Disclaimer

 
This article is for general information only and does not constitute investment, financial, legal or tax advice, nor any recommendation to provide liquidity or to buy or sell any asset. Figures are current as of November 17, 2026, and the impermanent loss percentages, capital efficiency multiples, gas cost shares and fee offset day counts are all calculated from public data and published formulas, on a model assumption of a 50∶50 constant product pool that does not apply to unequally weighted pools or to invariants designed for correlated assets, and that excludes protocol risk, contract risk and oracle deviation. Crypto assets, onchain liquidity positions and leveraged products are highly volatile and can produce losses exceeding the initial outlay. Readers should verify onchain data and protocol documentation and judge independently, consulting a qualified professional where appropriate. MEXC operates in compliance with applicable laws and regulations and does not offer services to residents of restricted jurisdictions.
 

About the Author

 
MEXC Editorial Team provides the latest platform updates, product launches, trading feature guides, and industry insights. With a focus on developments in the global crypto market, it delivers transparent information, professional data, and forward-looking analysis. Both beginners and experienced traders can find firsthand updates, in-depth observations, and official announcements related to the MEXC ecosystem.
 

Research References

 
 
Want the fastest access to MEXC's latest updates? Join our official Telegram group now!
Join MEXC Community: X (Twitter) | Telegram | Discord
Account Verification: Understand KYC | How to Complete KYC
External Content Platforms: Substack | Medium | Paragraph | LinkedIn | X(News)
Market Opportunity
Ethereum Logo
Ethereum Price(ETH)
$1,811.92
$1,811.92$1,811.92
USD

The articles shared on this page are sourced from public platforms and are provided for reference only. They do not represent the position or views of MEXC. All rights belong to MEXC Editorial Team. If you believe any content infringes upon the rights of a third party, please contact [email protected] for prompt removal. MEXC does not guarantee the accuracy, completeness, or timeliness of any content and is not responsible for any actions taken based on the information provided. The content does not constitute financial, legal, or other professional advice, nor should it be interpreted as a recommendation or endorsement by MEXC. For expert insights and in-depth analysis, visit MEXC Learn.