Skip to content

DeFi Liquidity Pools: How AMMs Work and What Liquidity Providers Risk

A factual guide to DeFi liquidity pools, AMM pricing, provider fees, impermanent loss, concentrated ranges and the risks to assess before depositing.

DeFi liquidity pools are smart contracts that hold tokens so users can trade or perform other financial operations onchain. In an automated market maker, or AMM, traders exchange assets against the pool instead of waiting for a specific buyer or seller to accept an order.

That mechanism is useful, but it does not turn liquidity provision into guaranteed “passive income.” A liquidity provider exchanges direct exposure to the deposited tokens for a position whose composition changes with trades. Fees may compensate for some risks, while token losses, price divergence, smart-contract failures, network costs and active position management can exceed the fees earned.

What a DeFi liquidity pool does

A pool contains reserves governed by smart-contract rules. A trader sends one token to the contract and receives another. The contract calculates the amount using its pricing function, available liquidity and fees. After the transaction, the balance of both reserves has changed.

Liquidity providers, commonly called LPs, deposit the assets required by that pool. Their position represents a claim on a share of the reserves and accrued fees according to the protocol’s rules. Older AMMs commonly issue fungible LP tokens. Concentrated-liquidity designs may represent each customized position as an NFT or another accounting record because positions can use different ranges and fee settings.

This is one application within the broader system described in our introduction to decentralized finance. Pools can support decentralized exchanges, lending markets, liquidations, stablecoin conversion and more complex strategies, but each use adds its own contracts and dependencies.

How a constant-product AMM sets prices

The classic Uniswap model uses the invariant x × y = k. Here, x and y are the reserves of two tokens and k is their product. Ignoring fees for a simple illustration, a trade must leave the reserves on the same curve. Taking one asset out requires adding the other.

Suppose a pool holds 10 ETH and 20,000 units of a dollar-denominated token. The reserve ratio implies 2,000 units per ETH, and k is 200,000. If a trader adds enough dollar tokens to remove ETH, ETH becomes scarcer in the pool and its pool price rises. Large trades relative to the reserves move the ratio farther, creating greater price impact.

Price impact is not the same as slippage

Price impact is the change caused by the trade’s size relative to available liquidity. Slippage is the difference between the expected and executed result, which can also reflect other transactions, routing and market movement before confirmation. A slippage tolerance is a transaction constraint, not a promise that a trade is fairly priced.

Arbitrage traders compare the pool with other markets and trade when prices diverge. Their activity can bring the pool price back toward the wider market, but it changes the reserve mix held by LPs. That rebalancing is central to both AMM functionality and liquidity-provider loss relative to simply holding the tokens.

Not every AMM uses the same curve

DesignTypical purposeMain LP consideration
Full-range constant productGeneral volatile token pairsCapital remains available across the full theoretical price range, but much may sit far from the current price
Concentrated liquidityGeneral pairs with user-selected rangesNarrow ranges can earn more fees per unit of active capital but require management and can go out of range
Stable-swap curveAssets expected to trade near parityLow price impact near the peg, with substantial risk if an asset permanently depegs
Weighted or multi-asset poolCustom portfolio weights and more than two assetsExposure and rebalancing depend on the pool’s weights and invariant
Hook- or extension-based poolCustomized fee, oracle or trading logicAdditional code and configuration may add behavior and risk beyond the base AMM

Uniswap’s current developer documentation distinguishes v2 full-range constant-product pools from v3 and v4 concentrated liquidity. Curve’s StableSwap is designed for assets expected to remain near the same value and blends constant-sum-like behavior near the peg with constant-product properties farther away. A pool name alone does not tell you which implementation, fee logic or administrative controls apply; verify the network and contract addresses.

How liquidity providers earn fees

Each swap may pay a fee set by the pool or protocol. Some or all of that fee accrues to active liquidity, and a protocol fee may also apply. The provider’s gross fee revenue depends on trading volume, fee rate, share of active liquidity and time in range. Incentive tokens can add another reward stream, but their value and distribution rules can change.

Net return is more demanding to calculate:

  • Start with fees actually accrued to the position, not an advertised annual percentage.
  • Add incentive tokens at a conservative, realizable value rather than their issue price.
  • Subtract network fees for approvals, deposits, collections, rebalancing and withdrawal.
  • Compare the ending position with the value of holding the original assets over the same period.
  • Account for taxes and recordkeeping obligations in the relevant jurisdiction.
  • Include losses from depegs, token taxes, transfer restrictions or failed transactions.

Transaction fees paid to the blockchain are separate from pool swap fees. Our guide to blockchain gas fees explains why a profitable-looking position can become uneconomic to manage when network demand rises.

Impermanent loss explained accurately

Impermanent loss is the shortfall between the value of an LP position and the value of simply holding the originally deposited assets, measured at the same external prices. It occurs because arbitrage changes the pool’s token balances as relative prices move. The phrase “impermanent” can be misleading: withdrawing realizes the result, and a token pair may never return to its original price ratio.

For a simplified fee-free, 50/50 constant-product pool, the relative difference after the external price ratio changes by a factor r is:

2 × √r ÷ (1 + r) − 1

If one token doubles relative to the other, the result is about −5.72% versus holding, before fees and costs. This does not necessarily mean the LP position lost value in dollar terms; it means it is worth less than the hold alternative. Fees may offset the difference, but they are not guaranteed to do so.

Concentrated positions behave differently. A narrow range can magnify fee earning while active, but price movement can convert the position almost entirely into one asset. When the market leaves the selected range, the position stops earning swap fees until price returns or the provider repositions it. Going out of range is not itself a liquidation, but it can leave the provider holding the asset that underperformed across the range.

Risks beyond impermanent loss

RiskWhat can happenWhat to inspect
Token riskA token can depeg, be frozen, add transfer restrictions or become worthlessIssuer, contract, upgrade rights, holders, liquidity and redemption mechanism
Smart-contract riskA bug or exploit can drain or lock fundsExact deployment, audits, age, changes, bug bounty and incident history
Administrative riskPrivileged keys may change fees, pause functions or upgrade contractsOwner functions, multisig signers, timelocks and governance process
Oracle and integration riskManipulated prices or failures in another protocol can cause lossesPrice sources, dependencies, limits and emergency controls
Range-management riskA concentrated position stops earning and becomes single-sidedChosen range, volatility, rebalancing plan and gas costs
Front-end and approval riskA fake site or unlimited token approval can enable theftDomain, contract address, wallet prompt and existing allowances
MEV and execution riskTransaction ordering can worsen executionSlippage, routing, private transaction options and trade size
Legal and tax riskObligations and availability vary by locationLocal rules, records and qualified professional advice

An audit is useful evidence, not insurance. Ethereum’s security documentation notes that deployed contracts can be difficult to patch and that oracle manipulation can exploit protocols using weak price inputs. The CFTC’s DeFi report likewise identifies smart-contract, oracle, governance, liquidity, leverage and concentration risks. Loss recovery may be limited even when the cause is known.

A due-diligence workflow before depositing

  1. Identify the exact deployment. Confirm network, protocol version, pool address and both token contract addresses from official sources.
  2. Understand the assets. Read issuer and token documentation, including minting, freezing, upgrading, redemption and transfer controls.
  3. Read the pool rules. Check fee tier, range model, withdrawal process, protocol fee, reward terms and any lockup.
  4. Inspect control paths. Determine who can upgrade, pause or change the contracts and whether a timelock applies.
  5. Model the position. Estimate results under price divergence, depeg, lower volume and higher gas—not only the displayed yield.
  6. Review security evidence. Look for audits covering the deployed code, bug bounties, monitoring and incident response.
  7. Protect the wallet. Verify every signature and approval. Our guide to public and private keys explains why a legitimate interface never needs your recovery phrase.
  8. Start with a limited amount. Test deposit, fee collection and withdrawal before increasing exposure.
  9. Keep records. Save transaction hashes, starting token amounts, fees, incentives, gas costs and ending balances.

Permissionless pool creation means a familiar ticker or logo is not enough. A newly issued token can copy another name, and its creator may control most of the supply or liquidity. The same issuer and token-rights checks used for initial coin offerings are relevant when an unknown token appears in a high-yield pool.

A simple position example

Imagine depositing equal values of Token A and Token B into a full-range constant-product pool. Trades generate fees, but Token A then rises sharply on external markets. Arbitrageurs buy A from the pool and add B until the pool ratio reflects the market. Your position now holds less A and more B than it began with.

To evaluate the outcome, withdraw or value the current claim, add uncollected fees, subtract every transaction cost, and compare that total with the value of the original quantities of A and B if held untouched. Only then can you tell whether fees overcame divergence loss. An interface’s current annualized yield cannot answer that historical comparison by itself.

Frequently asked questions

Are liquidity-pool fees guaranteed?

No. Fees depend on actual trading and the pool’s rules. A concentrated position earns no swap fees while out of range, and any fees earned can be outweighed by other losses and costs.

Is impermanent loss only a paper loss?

No. It is a comparison with holding, and withdrawing fixes the current reserve mix. Even without withdrawal, the price ratio may never return, and a depegged or failed token can create a much larger permanent loss.

Does more total value locked make a pool safe?

No. Deeper liquidity can reduce price impact, but total value locked does not eliminate contract, token, governance, oracle or legal risks. Verify how the figure is calculated and whether assets are counted elsewhere.

Can I lose more than I deposit?

A basic unleveraged LP position is generally limited to the assets committed, but borrowing, leveraged liquidity, derivatives or linked protocols can create additional liabilities. Understand every contract in the strategy.

Bottom line

DeFi liquidity pools replace an order-book counterparty with smart-contract reserves and algorithmic pricing. That makes continuous onchain trading possible, but providing liquidity is an active financial position—not a savings account. Understand the curve, tokens, range, fees, permissions, dependencies and exit process, then compare the net result with simply holding the assets. This article is educational and not individualized financial advice.

Primary references

Comments

Leave a Reply