What Is AMM Rebalance?

AMM rebalancing is the automatic token-ratio shift your liquidity pool executes on every trade — and it is the direct cause of impermanent loss.

AMM rebalance is the automatic adjustment of token ratios inside a liquidity pool. It happens on every trade, enforced by the pool’s invariant formula, with no input from the liquidity provider.

Most people assume the term refers to something the pool does periodically — a scheduled tune-up, like a monthly portfolio review. That is not how it works. AMM rebalancing is continuous, involuntary, and happens at the transaction level. Every swap shifts the pool’s token composition, and that shift has direct consequences for anyone who has deposited funds.

The sections below cover the full mechanism: how the math works, what it costs you as a liquidity provider (LP), why a metric called loss versus rebalancing (LVR) measures that cost more accurately than impermanent loss does, and what changes when you move from classic AMMs to concentrated liquidity pools like Uniswap v3 and v4.

Key Takeaways

  • AMM rebalancing happens automatically on every trade — the pool’s invariant formula adjusts token ratios with no input from the LP.
  • When prices shift, the pool systematically accumulates the underperforming asset and sheds the outperforming one, which is the direct cause of impermanent loss.
  • LVR (loss versus rebalancing) is a more precise measure of LP cost than impermanent loss because it captures per-trade value drain from arbitrage, not just start-to-end price movement.

—

What AMM Rebalance Means

An automated market maker (AMM) is a smart contract that holds two or more tokens and lets anyone trade against it using a mathematical formula rather than a traditional order book. The pool does not need buyers and sellers to line up at the same price — the formula sets the price automatically based on the ratio of tokens in the pool.

AMM rebalancing is what happens to that ratio every time a trade goes through. When someone buys ETH from an ETH/USDC pool, they send USDC in and take ETH out. The pool now holds more USDC and less ETH. The formula responds by raising the implied price of ETH. The pool has rebalanced its holdings — fewer ETH, more USDC — to reflect the new state after the trade.

This is not a scheduled event. No cron job, no governance vote, no LP action required. The rebalancing is a side effect of every single transaction.

In a stable market, these adjustments are tiny. In a volatile one, they add up — and they always run in one direction: into the weaker asset, away from the stronger one. For LPs, that direction is the one that costs money.

One terminology trap worth knowing: “AMM rebalance” describes the automatic ratio adjustment inside a pool. It is different from “portfolio rebalancing,” which means deliberately restoring a target asset allocation. The next section covers that distinction directly.

—

How the AMM Rebalancing Mechanism Works

The engine behind AMM rebalancing is the constant product formula: **x * y = k**.

In this formula, x is the quantity of token A in the pool, y is the quantity of token B, and k is a constant the protocol preserves on every trade. When a trader swaps token A for token B, the pool releases some of token B and absorbs some of token A. Because k must stay fixed, the ratio between x and y shifts — and that ratio shift is the rebalance.

A concrete example makes this clear. Suppose a pool holds 10 ETH and 20,000 USDC. The constant k = 10 * 20,000 = 200,000. The implied price of ETH in the pool is 20,000 / 10 = $2,000.

A trader buys 1 ETH from the pool. They send in USDC and take out 1 ETH. The pool now holds 9 ETH. To keep k at 200,000, the pool must now hold 200,000 / 9 = approximately 22,222 USDC. The trader sent in roughly 2,222 USDC (plus fees). The new implied price of ETH is 22,222 / 9 = approximately $2,469.

State ETH in pool USDC in pool Implied ETH price
Before trade 10 ETH 20,000 USDC $2,000
After trade 9 ETH ~22,222 USDC ~$2,469

The ratio has shifted. The pool now holds proportionally less ETH and more USDC. That is the AMM rebalance.

Now introduce the arbitrageur. In most cases, the trade that moved the pool price was itself an arbitrage trade. If ETH’s market price on Binance rises to $2,500 while the pool still shows $2,000, a profit opportunity exists: buy cheap ETH from the pool, sell on Binance. Arbitrageurs do this continuously. Each time one of them executes, the pool adjusts its ratio and implied price upward, eventually matching the external market. The pool’s rebalancing and the market’s price discovery are the same mechanism, playing out trade by trade.

AMM rebalancing is entirely market-driven. No human inside the pool is managing the ratio. Arbitrageurs do the work because it is profitable for them — and that profitability comes, at least in part, from the LP. That is what makes DeFi liquidity farming anything but passive income.

—

AMM Rebalance vs Portfolio Rebalancing — Two Different Things

The same word describes two genuinely different processes. The confusion is common enough to address directly.

AMM pool rebalancing is involuntary and continuous. The protocol adjusts the token ratio on every trade to preserve its invariant. The LP did not choose this and cannot opt out. The rebalancing is a mechanical consequence of how the AMM works.

Portfolio rebalancing is intentional and periodic. An investor sets a target allocation — say, 60% ETH, 40% BTC — and then periodically (or algorithmically) restores that ratio when market prices cause it to drift. The investor decides when to rebalance, what the targets are, and what trades to execute.

Balancer sits at the intersection of both meanings. A Balancer weighted pool sets explicit target weights — for example, 80% ETH and 20% USDC — and uses the AMM invariant to enforce them. When prices shift and the pool composition drifts from those targets, arbitrageurs are incentivized to trade against the pool to restore the weights, earning a profit in the process. From the LP’s perspective, it looks like a portfolio manager enforcing a target allocation. From the protocol’s perspective, it is an AMM formula doing what AMM formulas do.

Feature AMM pool rebalancing Portfolio rebalancing
Trigger Every trade Scheduled or threshold-based
Who acts The AMM formula automatically A person, algorithm, or arbitrageur paid to restore targets
Frequency Continuous (every block with trades) Periodic or conditional
LP cost Impermanent loss, LVR Transaction costs, slippage

Balancer is sometimes marketed as a “self-rebalancing portfolio.” That is accurate in the portfolio sense — the target weights are enforced — but the mechanism is still AMM rebalancing, which carries the same LP costs as any other AMM.

—

What AMM Rebalancing Does to Your Position

When the AMM rebalances, it systematically sells the asset going up in price and buys the one going down. That is the only direction the math can move.

When ETH rises, the arbitrageur buys ETH from the pool (reducing the pool’s ETH and pushing price up). When ETH falls, the arbitrageur sells ETH to the pool (increasing the pool’s ETH and pushing price down). Either way, the LP ends up with more of the asset that performed worse and less of the asset that performed better. Not bad luck. The formula in action.

The concrete consequence is impermanent loss. An LP deposits 1 ETH and 2,000 USDC when ETH = $2,000. Total deposit value: $4,000. ETH then rises to $4,000. The pool has been selling ETH and buying USDC the entire time. When the LP withdraws, they hold approximately 0.707 ETH and 2,828 USDC — worth roughly $5,656 in total. Simply holding 1 ETH and 2,000 USDC would have returned $6,000. The $344 difference is impermanent loss.

Impermanent loss at common price multiples (relative to entry price):

  • 1.25x move: approximately 0.6% impermanent loss
  • 1.5x move: approximately 2.0% impermanent loss
  • 2x move: approximately 5.7% impermanent loss
  • 3x move: approximately 13.4% impermanent loss
  • 5x move: approximately 25.5% impermanent loss

“Impermanent” does not mean “will definitely reverse.” The loss disappears only if and when price returns to your entry point. If ETH never comes back to $2,000, the loss is permanent at withdrawal. Before depositing into any pool, understand how yield farming positions are actually structured — impermanent loss is often buried in the APY headline.

AMM rebalancing does not bring your position to zero. Even in extreme price movements, the constant product formula always leaves some balance of both tokens in the pool. The worst case is an asymptotic approach to 100% of one asset — not a total wipeout.

Arbitrageurs fund their profits by trading against your stale pool price before it updates. That dynamic is the foundation of exit liquidity — a concept every LP should understand before depositing into a pool where arbitrage volume is high.

—

LVR and AMM Rebalance — Why Impermanent Loss Understates the Real Cost

Impermanent loss is a useful first approximation, but it has a blind spot: it only compares where prices started and where they ended. It ignores everything that happened in between.

Think about what that misses. Suppose ETH starts at $2,000, spends a week bouncing between $1,800 and $2,500, and then closes back at $2,000. Impermanent loss would show zero — start price equals end price. But during that week, arbitrageurs made dozens of profitable trades against the pool, each time exploiting the gap between the AMM’s stale price and the external market. The LP funded every one of those trades.

Loss versus rebalancing (LVR) captures exactly that cost. LVR measures the per-trade value drain that occurs every time an arbitrageur buys a stale AMM price before it updates. The concept was formalized by a16z Crypto researchers Jason Milionis and Anthony Zhang, whose paper quantifying LVR shows that for a constant-product AMM with 5% daily volatility and 30 bps fees, LPs need the pool to turn over at least 10.4% of its total value in daily trading volume just to break even.

The formula underlying LVR is σ²/8, where σ is the daily standard deviation of returns. Higher volatility means more arbitrage opportunities, faster-accumulating LVR, and a harder break-even threshold.

Three things change how you evaluate a pool once you factor in LVR:

  • You can have zero impermanent loss (prices returned to entry) and still have lost real value through LVR accumulated along the way.
  • Stablecoin pairs have very low σ, which is why stablecoin AMM pools tend to be more profitable for LPs despite charging lower fees.
  • Fee income needs to exceed LVR, not just impermanent loss, to consider the position profitable.

LPs do not need to compute LVR themselves. But knowing it scales with volatility changes how you evaluate pairs and fee tiers before depositing.

Hedging LVR with options or perpetuals is possible but adds complexity. Crypto derivatives — specifically futures and options — can offset the directional exposure that AMM rebalancing creates.

—

Concentrated Liquidity and AMM Rebalancing in Uniswap v3 and v4

Classic AMMs like Uniswap v2 distribute LP capital across every possible price from zero to infinity. The LP’s position is always active, always earning fees on trades, and always drifting gradually with the pool’s ratio. AMM rebalancing is passive and continuous.

Uniswap v3 changed this with concentrated liquidity. LPs select a specific price range — say, ETH between $1,800 and $2,500. Within that range, the capital earns fees at a much higher rate than it would in a full-range pool. But when price moves outside the range, the position stops working.

When ETH drops below $1,800, the position converts entirely to ETH. When ETH rises above $2,500, the position converts entirely to USDC. In both cases, the position stops earning fees. It sits idle — 100% in one token — until price re-enters the range or the LP actively repositions.

This is a hard rebalance, not a gradual drift. The LP’s position snaps to single-asset exposure in a discrete step.

Feature Uniswap v2 (full range) Uniswap v3/v4 (concentrated)
Rebalancing behavior Gradual, continuous Snaps to single asset at range boundary
Fees when out of range Always earning Zero — position is idle
Who manages rebalancing No action needed LP must reposition, or use a vault manager

Uniswap v4 introduces hooks — programmable smart contract extensions that run before or after a swap. Vault managers like Arrakis Finance, Gamma Strategies, and Bunni use hooks to automatically detect when a position has drifted out of range and reposition it without requiring LP action.

Automated vault management is not free. The vault contract pays gas on every reposition, and most vault managers charge a performance fee on earned yield. LPs considering liquid staking tokens in concentrated liquidity pools — LSTs like stETH are a common pairing choice — should know that LST repricing relative to ETH can trigger range exits more frequently than expected. That makes vault management particularly relevant for these pairings.

Concentrated liquidity amplifies both the upside (higher fee capture per dollar deployed) and the downside (idle capital at range boundaries, gas costs from repositioning). The AMM rebalancing mechanism still applies inside the range — it just becomes irrelevant outside it, where the position has already fully converted.

—

Frequently Asked Questions

Does AMM rebalancing happen automatically or do I need to do anything?

Pool-level AMM rebalancing is fully automatic — it happens every time any trader swaps against the pool. As a liquidity provider you do not press a button or set a schedule. The formula enforces the ratio on every transaction. What is not automatic in Uniswap v3/v4 is managing your price range: if your position drifts out of range, you need to manually reposition it or use a vault manager to do it on your behalf. In classic full-range AMMs like Uniswap v2 or Balancer, no range management is needed.

How does AMM rebalancing cause impermanent loss?

Every time the pool rebalances in response to a price change, it moves the LP’s portfolio toward the asset that lost relative value. If ETH rises from $2,000 to $4,000, the pool sells ETH and buys USDC to keep the invariant balanced — so the LP ends up holding less ETH than when they entered. On withdrawal, those fractions of ETH they no longer hold represent the gap between what the pool returned and what simply holding would have returned. That gap is impermanent loss. It only becomes permanent at withdrawal.

What is LVR and how is it different from impermanent loss in an AMM rebalance context?

Impermanent loss compares your final LP value to simply holding your starting assets outside the pool. LVR (loss versus rebalancing) measures something more granular: the per-trade loss each time an arbitrageur buys a stale AMM price before it updates to reflect the market. Impermanent loss ignores how many times the price moved during your LP period. LVR captures that price-path cost. In practice, LVR tends to be higher than impermanent loss in volatile markets because it counts every arbitrage interaction, not just the net start-to-end price move.

Does AMM rebalancing happen more often when prices move faster?

Yes. Rebalancing is driven by trades, and fast-moving markets attract more arbitrageurs who exploit the price gap between the pool and external markets. More arbitrage transactions mean more rebalancing events per block. This is why LVR scales with price volatility: higher volatility means more arbitrage opportunities, faster-accumulating LP losses from rebalancing, and a higher fee volume needed to break even.

Can I avoid losses from AMM rebalancing entirely?

Not entirely. If you provide liquidity to any standard AMM, some rebalancing loss is unavoidable — it is the cost of offering continuous liquidity. You can reduce the impact by choosing high-volume, low-volatility pairs (such as stablecoin-to-stablecoin pools where price divergence is minimal and fees accumulate on high volume), using concentrated ranges only on assets you understand well, or using protocols with dynamic fees that adjust to volatility. Full avoidance means not providing liquidity to AMMs at all.

—

Where To Start

If you are approaching AMM rebalancing for the first time and considering a liquidity position, these steps give you a practical foundation:

  1. Start with stablecoin pools. USDC/USDT or similar pairs have near-zero volatility, which means AMM rebalancing barely moves the ratio and impermanent loss stays minimal. Fee income is more likely to outpace LVR here than in volatile pairs.
  1. Understand the pool’s fee tier before depositing. Higher fee tiers exist to compensate LPs for higher-volatility pairs. A 0.05% fee on ETH/USDC is designed for tight, liquid conditions. A 1% fee is designed for volatile or exotic pairs. Neither tier eliminates LVR, but the fee structure tells you what the protocol expects about the pair’s volatility.
  1. If you are using Uniswap v3 or v4, decide whether you will manage your range manually or use a vault manager. Manual management requires you to monitor the position and pay gas to reposition. Vault managers automate this but charge a performance fee.
  1. Check daily trading volume against total value locked (TVL) before depositing. A rough guide: daily volume should be at least 5-10% of TVL for fees to have a realistic chance of covering LVR costs on a moderate-volatility pair.
  1. Approach any fee APY estimate skeptically if it does not account for impermanent loss and LVR. Many AMM interfaces show fee APYs calculated from recent volume without adjusting for the losses that rebalancing creates. The net return can be meaningfully lower — and sometimes negative — than the headline number suggests.