Liquidity Pools and AMMs Explained
Course goal: understand how automated market makers price trades, how liquidity providers earn fees, why impermanent loss occurs, what concentrated liquidity changes, and how to evaluate whether an LP position makes economic sense.
1. Why liquidity pools exist
Markets need liquidity: assets available for buyers and sellers. Traditional exchanges use order books where market makers post bids and offers. Many decentralized exchanges instead use liquidity pools funded by users and smart contracts that quote prices according to mathematical rules.
A liquidity pool might hold ETH and USDC. Traders swap one asset for the other, changing the quantities in the pool. Liquidity providers, or LPs, supply the assets and earn a share of trading fees.
2. Automated market makers
An automated market maker, or AMM, is a smart-contract system that determines prices and executes swaps without a traditional centralized order book. The pool itself becomes the counterparty to traders.
This design makes markets available continuously as long as liquidity remains in the contract. It also makes pricing transparent because the mechanism is encoded in software.
3. Constant-product model
A classic AMM uses the relationship x Ă— y = k, where x and y represent quantities of the two pooled assets. When a trader removes some of asset x, they must add enough of asset y for the relationship to remain consistent after fees.
The resulting curve means price moves as the pool becomes imbalanced. A large trade relative to pool size causes more price impact than a small trade.
4. Arbitrage keeps prices aligned
Suppose ETH trades for $2,500 on large exchanges but a DEX pool briefly values it at $2,450. Arbitrageurs can buy the cheaper ETH from the pool and sell it elsewhere. Their trades change the pool balance and push the DEX price back toward the wider market.
AMMs therefore depend on arbitrage for price alignment. The pool formula does not know the “true” external price on its own.
5. LP shares
When users add liquidity, the protocol tracks their share of the pool. Older AMMs issue LP tokens; newer systems may represent positions differently. When liquidity is withdrawn, the provider receives their proportional share of the current pool assets plus accumulated fees, minus any protocol-specific effects.
Importantly, the asset mix at withdrawal may be very different from the mix originally deposited.
6. Trading fees
Each swap normally pays a fee. Some or all of that fee is distributed to LPs. A pool with high trading volume relative to liquidity can generate attractive fee revenue, but high volume often comes with high volatility or risk.
Fee APY estimates are backward-looking. They can change quickly when trading activity or liquidity moves elsewhere.
7. Impermanent loss intuition
Impermanent loss occurs when an LP position becomes worth less than simply holding the original assets because the AMM continuously rebalances the portfolio as relative prices change.
If ETH rises strongly against USDC, traders buy ETH from the pool. The pool ends up with less ETH and more USDC. The LP benefited from the price rise but holds less of the outperforming asset than a passive holder would have held.
8. Why “impermanent” can be misleading
The loss is called impermanent because if the relative price returns to the original level before withdrawal, the divergence can shrink. But if the LP withdraws while prices remain different, the underperformance becomes realized.
Even without withdrawal, the opportunity cost is economically real. The word should not be interpreted as “not a real loss.”
9. Fees versus impermanent loss
An LP strategy is profitable relative to holding when fee income and incentives outweigh the rebalancing drag and other costs. Highly correlated pairs can experience lower divergence risk than volatile pairs, while exotic token pairs can produce large impermanent loss.
Do not look at APY alone. Compare expected fees with expected volatility and token risk.
10. Stablecoin pools
Stablecoin AMMs often use specialized curves designed for assets expected to trade near the same price. This improves capital efficiency and reduces slippage around the peg.
However, if one stablecoin depegs permanently, the pool can become concentrated in the weaker asset as arbitrageurs remove the stronger one. A “low-volatility” pool can therefore have severe tail risk.
11. Concentrated liquidity
Concentrated-liquidity AMMs allow LPs to deploy capital only within selected price ranges. If price stays inside the range, the position can earn more fees per dollar of capital than a traditional full-range pool.
If price moves outside the range, the position can become entirely one asset and stop earning trading fees until price returns or the LP repositions.
12. Active management
Concentrated liquidity turns passive LPing into a more active strategy. Providers must choose ranges, monitor price and decide when to rebalance. Frequent rebalancing can add gas costs and adverse-selection risk.
The highest displayed APY may assume a narrow range that requires much more management than a broad, conservative range.
13. Tick ranges
Concentrated AMMs divide prices into discrete ranges or ticks. LPs choose lower and upper bounds. Liquidity only participates in trades while the market price is within those bounds.
The position therefore behaves somewhat like an automated trading strategy that continuously sells one asset as it rises and buys it as it falls within the range.
14. Adverse selection
LPs can lose to informed or fast traders when prices move on external markets before the pool adjusts. Arbitrage restores the correct price but captures value from the pool during the adjustment. This is part of the economic cost LPs accept in exchange for fees.
15. MEV
On-chain trade ordering can create maximal extractable value opportunities. Sandwich attacks and arbitrage can affect traders and LPs. Some AMM designs and private-routing systems attempt to reduce these effects, but they remain part of decentralized market structure.
16. Smart-contract risk
LPs deposit assets into code. A contract exploit can result in total loss independent of market performance. Audits, formal verification and long operating history can reduce but not eliminate the risk.
17. Token risk
If one side of the pair collapses toward zero, the pool can end up heavily concentrated in that failing asset. Fee income is unlikely to compensate for catastrophic token failure.
Liquidity provision should therefore begin with the question: “Would I be comfortable owning both assets?” If not, the pair may be inappropriate regardless of APY.
18. Incentive-token risk
Protocols may boost yields by distributing governance tokens. If the reward token falls sharply, the realized yield can be much lower than advertised. Emission-heavy incentives can also attract temporary liquidity that disappears when rewards end.
19. LP accounting
Track returns in a consistent base currency and compare them with a relevant benchmark. A position that earns 20% in fees but underperforms holding the underlying tokens by 10% has a different outcome from the headline fee number.
Include gas, bridge costs, token incentives and any tax obligations when calculating net returns.
20. Simple scenario
Imagine supplying $5,000 of ETH and $5,000 of USDC. Trading volume generates $400 of fees over several months. ETH rises substantially, and the AMM rebalances so the withdrawn portfolio is worth $12,500. If simply holding the original assets would have been worth $13,300, the LP earned money in absolute terms but underperformed hold by $800. The fees did not fully offset impermanent loss.
21. LP due-diligence checklist
- Do I understand both assets?
- How deep and mature is the pool?
- What is the fee tier?
- How volatile is the pair?
- Is liquidity concentrated?
- Where does advertised yield come from?
- What smart-contract audits exist?
- Are token incentives sustainable?
- What bridge or oracle dependencies exist?
- What benchmark will I use to judge performance?
22. Knowledge check
- What role does arbitrage play in an AMM?
- Why can an LP underperform simply holding?
- What happens when a concentrated position moves out of range?
- Why can a stablecoin pool still be high risk?
- Why is fee APY alone insufficient?
Answers: arbitrage aligns pool price with wider markets; AMMs rebalance away from the outperforming asset; out-of-range liquidity may stop earning fees and become one-sided; depegs can concentrate the pool in the weak asset; fees must be compared with divergence, token and contract risk.
23. Practical exercise
Choose one ETH/stablecoin pool and one stablecoin/stablecoin pool. Record liquidity, 24-hour volume, fee tier and any incentive rewards. Estimate annualized fees using recent volume, then list three reasons the estimate may not persist. For the ETH pair, describe what happens to the asset mix if ETH doubles. For the stablecoin pair, describe what happens if one token trades at $0.80.
24. Key takeaways
Liquidity provision is market making, not a savings account. LPs earn fees for offering inventory to traders and accept rebalancing, volatility, contract and token risk in return. The correct benchmark is often what you would have earned by simply holding the assets.
Next lesson: Crypto Scams: Rug Pulls, Phishing and Honeypots →
Educational content only. DeFi liquidity provision can result in substantial or total loss.