Impermanent loss (IL) is the hidden tax on liquidity providers.

It happens because AMMs rebalance tokens whenever prices move. If one asset in the pool pumps, the protocol automatically sells part of it to buy the other side. But that means users end up with fewer of the winner and more of the loser. The gap between what you’d have if you just held vs what you actually have in the pool is IL.

Take ETH/USDC on Uniswap as an example. Suppose you deposited $5,000 worth of ETH and $5,000 in USDC at a 1:1 ratio when ETH was $4,500. If ETH doubles to $9,000, you don’t walk away with $10,000 in ETH and $5,000 in USDC. Instead, the pool rebalances your position to roughly 0.785674 ETH (~$7,071.07) and $7,071.07 USDC—about $14,142.14 total. If you had just held, you’d have $15,000. The missing ~$857.86 (~5.72%) is impermanent loss, before accounting for trading fees.

At its core, impermanent loss happens because most AMMs use constant product market making. The pool continuously rebalances as prices move, leaving LPs with fewer of the appreciating asset and more of the depreciating one. If ETH rallies, your ETH gets sold off into USDC. If ETH dumps, you end up holding more ETH. Either way, compared to simply holding both tokens, your position is worth less.

During periods like May 2021’s DeFi crash, fees spiked but price swings outpaced them, leaving many LPs net negative. Even in calmer conditions, if volumes are thin or users move to cheaper venues, fees may not cover IL. That’s why LP profitability is inconsistent, even on blue-chip pools.

Because AMMs rebalance inventory mechanically, your position drifts toward the underperforming asset as prices move. That “impermanent” loss can disappear if price reverts because it’s only realized when you withdraw or when you rebalance. With full-range liquidity (v2-style), you don’t need to rebalance, so IL can remain unrealized and may shrink if price mean-reverts. The dynamic changes with concentrated liquidity (v3-style): going out of range pushes LPs or vaults to rebalance to stay active, which locks in the loss (the core of the LVR debate). In short, IL is inherent to AMM inventory shifts, but rebalancing is what makes it stick, especially in concentrated ranges.

Many guides suggest swap fees “cover” IL, but outcomes depend heavily on timeframe and volatility and on whether you’re using concentrated liquidity (where LVR kicks in). With full-range positions, it’s essentially a race: fees only accrue upward, while IL can shrink if prices mean-revert. In concentrated ranges, frequent rebalancing to stay in range can lock in losses, so fees don’t reliably offset IL. Research on Uniswap v3 revealed that roughly half of LPs lost money after accounting for IL, even with fee income. The “risk-free yield” narrative is a myth: LPs are essentially volatility sellers, and like any option writer, their risk is real and measurable.

Concentrated liquidity makes this problem sharper. By choosing a narrow range, LPs can collect higher fees when volume stays within it. But if price leaves the range, they stop earning altogether while still being exposed to IL. That’s why many retail LPs underperform even in pools with strong volume.

Can you “hedge away” impermanent loss (IL)? Not directly. IL comes from the AMM’s rebalancing: when price moves, the pool sells the winner and buys the loser. That mechanic keeps running no matter what other positions you add.

Borrowing liquidity can offset parts of that risk, especially for full-range market making. Taking an opposing MM exposure dampens your sensitivity to IL: for e.g. being long and short the MM profile at the same time but you’ll pay borrow fees that can eat into, or exceed, your swap-fee income. With concentrated liquidity, hedging gets much harder: once price leaves your range, execution stops and “loss versus rebalancing” (LVR) dynamics kick in, so any borrowed-liquidity hedge must be range-aware and actively managed, with higher cost and complexity.

What you can hedge is price (delta). An ETH/USDC LP can short ETH on a perp venue to offset directional moves. The AMM will still rebalance, so IL can still occur, but overall PnL is smoother because the short counteracts the price swing.

Options help hedge variance. Buying convexity (for example, a long straddle/strangle) adds payoff in big up or down moves, countering the concave profile that produces IL. Protective puts or collars can cap downside, but they don’t specifically neutralize rebalancing; they just bound outcomes. In all cases, you trade mechanical IL for option premium (and basis/slippage).

A “true” liquidity hedge that pays precisely when the pool sells the winner and buys the loser would need a payoff that mirrors the AMM’s rebalancing in reverse. A few research and experimental designs point that way, but there’s no widely adopted, liquid product today.

Not to mention that protocol-level “IL protection” hasn’t proven durable. Bancor’s program aimed to reimburse LPs, but during the 2022 bear market, volumes fell and withdrawals surged; payouts became unsustainable and the program was paused, underscoring how difficult direct IL insurance is in practice.

Borrowing liquidity can be an effective hedge for market making, but it doesn’t erase IL and it introduces financing costs. In practice, it becomes a race between those borrowing costs and the swap fees your position earns. Under efficient condition,s these tend to converge, so you’re effectively betting that fees will outpace the cost of the hedge.

There are two ways to think about it:

• LP hedging with borrowed liquidity. An LP can borrow a position that offsets their own market-making exposure (e.g., an opposing MM profile). This reduces IL sensitivity, but the borrow fees act like negative carry and can eat into (or outweigh) swap-fee income - much like being long and short on perps at the same time. Net results depend on volatility, fee flow, and borrow costs.

• Traders renting MM exposure (“impermanent gain”). A trader can borrow the MM exposure to take the other side of IL. They pay a recurring fee to hold that exposure, which is their “carry.” If volatility plays out in their favor, they can profit after fees; if not, the carry drags returns. Those fees flow to LPs, monetizing the risk LPs already carry.

Crucially, borrowing doesn’t change the AMM’s rebalancing. The pool will still sell the winner and buy the loser. That’s the mechanism that creates IL. Borrowing simply layers on an opposing payoff stream. In other words: you can’t cancel IL mechanically, but you can redistribute or neutralize parts of it at a cost. LPs may earn more by renting out exposure; borrowers gain a targeted volatility bet but pay carry to hold it.

Some LPs try to mitigate IL by actively managing ranges, especially in Uniswap v3. Concentrated liquidity means tighter ranges and higher fee capture, but it also means more frequent rebalancing and sharper IL when prices leave your range. The trade-off is bigger upside when right, bigger pain when wrong.

A few protocols, like Charm’s Alpha Vaults or Arrakis, built vault strategies that rebalance ranges on behalf of LPs. Others, like Gamma or DeFiEdge, offer automated management to retail users. But performance has varied widely. In volatile sideways markets, vaults can churn positions and rack up gas costs without outpacing IL. In trending markets, many strategies lag behind simple holding.

Active management reduces the “passive bleed” of IL but introduces new risks: timing errors, higher costs, dependence on the vault’s algorithm and loss vs rebalancing (LVR). For most users, it’s not the silver bullet they hoped for.

Systemic approaches like designing AMMs where LP risk can be shared, rented out, or priced differently are a new model. They don’t remove IL, but they reframe how it’s compensated.

For instance, Ammalgam allows LPs to rent out their exposure so traders who want volatility pay them for the privilege. The IL is still there but LPs collect more fees on top of swap revenue to make the trade-off more balanced. Think of it less as “hedging away IL” and more as “being paid more fairly for taking the risk.”

Other experimental designs look at derivative overlays (creating IL tokens to trade) or dynamic fee markets where fees increase in volatile periods to better match risk. None have solved IL outright, but they highlight an important shift: treating LP risk as something explicit, transferable, and compensable.

Hedging IL isn’t about “switching it off.” It’s about understanding what risk you’re exposed to (price, liquidity, or both) and using the right tool for the job. Price hedges (like perps) change exposure, range strategies chase fees, and new AMM models experiment with shifting who pays for volatility.

The challenge for DeFi is making that trade-off clearer and fairer. LPs should understand that they are volatility sellers, not passive depositors. Borrowers and traders should have tools to take the other side when it suits them. And protocols should design liquidation and fee models that reflect real liquidity conditions.

That’s the path to a healthier system: not eliminating impermanent loss, but making sure everyone knows what role they’re playing in the market.

Impermanent loss (IL) is the hidden tax on liquidity providers.

It happens because AMMs rebalance tokens whenever prices move. If one asset in the pool pumps, the protocol automatically sells part of it to buy the other side. But that means users end up with fewer of the winner and more of the loser. The gap between what you’d have if you just held vs what you actually have in the pool is IL.

Take ETH/USDC on Uniswap as an example. Suppose you deposited $5,000 worth of ETH and $5,000 in USDC at a 1:1 ratio when ETH was $4,500. If ETH doubles to $9,000, you don’t walk away with $10,000 in ETH and $5,000 in USDC. Instead, the pool rebalances your position to roughly 0.785674 ETH (~$7,071.07) and $7,071.07 USDC—about $14,142.14 total. If you had just held, you’d have $15,000. The missing ~$857.86 (~5.72%) is impermanent loss, before accounting for trading fees.

At its core, impermanent loss happens because most AMMs use constant product market making. The pool continuously rebalances as prices move, leaving LPs with fewer of the appreciating asset and more of the depreciating one. If ETH rallies, your ETH gets sold off into USDC. If ETH dumps, you end up holding more ETH. Either way, compared to simply holding both tokens, your position is worth less.

During periods like May 2021’s DeFi crash, fees spiked but price swings outpaced them, leaving many LPs net negative. Even in calmer conditions, if volumes are thin or users move to cheaper venues, fees may not cover IL. That’s why LP profitability is inconsistent, even on blue-chip pools.

Because AMMs rebalance inventory mechanically, your position drifts toward the underperforming asset as prices move. That “impermanent” loss can disappear if price reverts because it’s only realized when you withdraw or when you rebalance. With full-range liquidity (v2-style), you don’t need to rebalance, so IL can remain unrealized and may shrink if price mean-reverts. The dynamic changes with concentrated liquidity (v3-style): going out of range pushes LPs or vaults to rebalance to stay active, which locks in the loss (the core of the LVR debate). In short, IL is inherent to AMM inventory shifts, but rebalancing is what makes it stick, especially in concentrated ranges.

Many guides suggest swap fees “cover” IL, but outcomes depend heavily on timeframe and volatility and on whether you’re using concentrated liquidity (where LVR kicks in). With full-range positions, it’s essentially a race: fees only accrue upward, while IL can shrink if prices mean-revert. In concentrated ranges, frequent rebalancing to stay in range can lock in losses, so fees don’t reliably offset IL. Research on Uniswap v3 revealed that roughly half of LPs lost money after accounting for IL, even with fee income. The “risk-free yield” narrative is a myth: LPs are essentially volatility sellers, and like any option writer, their risk is real and measurable.

Concentrated liquidity makes this problem sharper. By choosing a narrow range, LPs can collect higher fees when volume stays within it. But if price leaves the range, they stop earning altogether while still being exposed to IL. That’s why many retail LPs underperform even in pools with strong volume.

Can you “hedge away” impermanent loss (IL)? Not directly. IL comes from the AMM’s rebalancing: when price moves, the pool sells the winner and buys the loser. That mechanic keeps running no matter what other positions you add.

Borrowing liquidity can offset parts of that risk, especially for full-range market making. Taking an opposing MM exposure dampens your sensitivity to IL: for e.g. being long and short the MM profile at the same time but you’ll pay borrow fees that can eat into, or exceed, your swap-fee income. With concentrated liquidity, hedging gets much harder: once price leaves your range, execution stops and “loss versus rebalancing” (LVR) dynamics kick in, so any borrowed-liquidity hedge must be range-aware and actively managed, with higher cost and complexity.

What you can hedge is price (delta). An ETH/USDC LP can short ETH on a perp venue to offset directional moves. The AMM will still rebalance, so IL can still occur, but overall PnL is smoother because the short counteracts the price swing.

Options help hedge variance. Buying convexity (for example, a long straddle/strangle) adds payoff in big up or down moves, countering the concave profile that produces IL. Protective puts or collars can cap downside, but they don’t specifically neutralize rebalancing; they just bound outcomes. In all cases, you trade mechanical IL for option premium (and basis/slippage).

A “true” liquidity hedge that pays precisely when the pool sells the winner and buys the loser would need a payoff that mirrors the AMM’s rebalancing in reverse. A few research and experimental designs point that way, but there’s no widely adopted, liquid product today.

Not to mention that protocol-level “IL protection” hasn’t proven durable. Bancor’s program aimed to reimburse LPs, but during the 2022 bear market, volumes fell and withdrawals surged; payouts became unsustainable and the program was paused, underscoring how difficult direct IL insurance is in practice.

Borrowing liquidity can be an effective hedge for market making, but it doesn’t erase IL and it introduces financing costs. In practice, it becomes a race between those borrowing costs and the swap fees your position earns. Under efficient condition,s these tend to converge, so you’re effectively betting that fees will outpace the cost of the hedge.

There are two ways to think about it:

• LP hedging with borrowed liquidity. An LP can borrow a position that offsets their own market-making exposure (e.g., an opposing MM profile). This reduces IL sensitivity, but the borrow fees act like negative carry and can eat into (or outweigh) swap-fee income - much like being long and short on perps at the same time. Net results depend on volatility, fee flow, and borrow costs.

• Traders renting MM exposure (“impermanent gain”). A trader can borrow the MM exposure to take the other side of IL. They pay a recurring fee to hold that exposure, which is their “carry.” If volatility plays out in their favor, they can profit after fees; if not, the carry drags returns. Those fees flow to LPs, monetizing the risk LPs already carry.

Crucially, borrowing doesn’t change the AMM’s rebalancing. The pool will still sell the winner and buy the loser. That’s the mechanism that creates IL. Borrowing simply layers on an opposing payoff stream. In other words: you can’t cancel IL mechanically, but you can redistribute or neutralize parts of it at a cost. LPs may earn more by renting out exposure; borrowers gain a targeted volatility bet but pay carry to hold it.

Some LPs try to mitigate IL by actively managing ranges, especially in Uniswap v3. Concentrated liquidity means tighter ranges and higher fee capture, but it also means more frequent rebalancing and sharper IL when prices leave your range. The trade-off is bigger upside when right, bigger pain when wrong.

A few protocols, like Charm’s Alpha Vaults or Arrakis, built vault strategies that rebalance ranges on behalf of LPs. Others, like Gamma or DeFiEdge, offer automated management to retail users. But performance has varied widely. In volatile sideways markets, vaults can churn positions and rack up gas costs without outpacing IL. In trending markets, many strategies lag behind simple holding.

Active management reduces the “passive bleed” of IL but introduces new risks: timing errors, higher costs, dependence on the vault’s algorithm and loss vs rebalancing (LVR). For most users, it’s not the silver bullet they hoped for.

Systemic approaches like designing AMMs where LP risk can be shared, rented out, or priced differently are a new model. They don’t remove IL, but they reframe how it’s compensated.

For instance, Ammalgam allows LPs to rent out their exposure so traders who want volatility pay them for the privilege. The IL is still there but LPs collect more fees on top of swap revenue to make the trade-off more balanced. Think of it less as “hedging away IL” and more as “being paid more fairly for taking the risk.”

Other experimental designs look at derivative overlays (creating IL tokens to trade) or dynamic fee markets where fees increase in volatile periods to better match risk. None have solved IL outright, but they highlight an important shift: treating LP risk as something explicit, transferable, and compensable.

Hedging IL isn’t about “switching it off.” It’s about understanding what risk you’re exposed to (price, liquidity, or both) and using the right tool for the job. Price hedges (like perps) change exposure, range strategies chase fees, and new AMM models experiment with shifting who pays for volatility.

The challenge for DeFi is making that trade-off clearer and fairer. LPs should understand that they are volatility sellers, not passive depositors. Borrowers and traders should have tools to take the other side when it suits them. And protocols should design liquidation and fee models that reflect real liquidity conditions.

That’s the path to a healthier system: not eliminating impermanent loss, but making sure everyone knows what role they’re playing in the market.