What is a Liquidity Pool?

In September 2020, a developer named Andre Cronje launched a token called YFI. There was no company, no sales team, no order book full of professional traders quoting prices. People just deposited their coins into a pool of code, and that pool let anyone in the world swap one token for another — instantly, day or night, no permission required. Within weeks, billions of dollars flowed through pools like it.

Think about how strange that is. On the New York Stock Exchange, every trade needs a counterparty: a human or a firm willing to take the other side. If nobody wants to sell you Apple at your price, you wait. Liquidity pools threw that requirement out. You don't trade against a person — you trade against a reserve of tokens locked in a smart contract, and a formula sets the price.

This single idea — replacing the matchmaker with a math equation — is the engine underneath nearly all of decentralized finance. Uniswap, Curve, PancakeSwap, Balancer, SushiSwap: different brands, same core mechanism. Understanding liquidity pools is understanding how DeFi actually moves money. Let's take the machine apart.

Traditional exchanges — NYSE, NASDAQ, Binance — match buyers with sellers through an order book. Someone posts a bid, someone posts an ask, and when prices meet, a trade happens. It works. It's been working since the Dutch East India Company in 1602.

Liquidity pools take a fundamentally different approach. Instead of matching individual buyers and sellers, a pool holds a reserve of two tokens — say ETH and USDC — locked in a smart contract. When you want to swap, you trade against the pool itself. No counterparty needed. No order book. No waiting for someone to take the other side of your trade.

The software that prices and executes those swaps has a name: an Automated Market Maker, or AMM. On a traditional venue, "market makers" are firms that continuously post buy and sell quotes to keep trading smooth — a sophisticated, capital-intensive job. An AMM automates that role with a formula, so the market maker isn't a firm at all. It's a smart contract anyone can deposit into.

Uniswap launched this model in November 2018 and it now processes billions in monthly volume. The entire DEX ecosystem runs on variations of the same idea. As of early 2025, over $10 billion sits in liquidity pools across Ethereum and its L2s alone.

Uniswap's core mechanic fits in a single equation: x × y = k. Here x is the quantity of Token A in the pool, y is the quantity of Token B, and k is a constant that the pool must always preserve.

A pool starts with 10 ETH and 20,000 USDC. Multiply them: k = 200,000. Now you want to buy 1 ETH. After your trade the pool must still satisfy x × y = 200,000. With 9 ETH remaining, the pool needs 200,000 ÷ 9 = 22,222 USDC. The difference — 2,222 USDC — is what you pay. That's a price of $2,222 per ETH, slightly above the $2,000 midpoint, because your trade shifted the ratio against you.

Notice what just happened: the price moved automatically, with no human deciding it. Buy more ETH and the remaining ETH gets scarcer in the pool, so each additional unit costs more. Sell ETH back and the price falls. The curve self-adjusts on every single swap. That gap between the quoted "fair" price and the price you actually pay is called slippage, and it grows with the size of your trade relative to the pool.

Depth is everything. A pool with $10 million in liquidity absorbs a $10,000 swap with negligible price impact — your trade is a rounding error against the reserves. The same $10,000 swap in a $100,000 pool moves the price violently, because you're consuming a meaningful chunk of one side. This is also why AMMs work best for pairs of similar value, like two stablecoins or a token and its wrapped version — small ratio changes mean small slippage.

Here's the question the formula begs: where do the tokens in the pool come from? The answer is liquidity providers (LPs) — ordinary users who deposit their assets so others can trade. Anyone can do it, and that openness is the whole point.

You deposit equal dollar amounts of both tokens — say $5,000 of ETH and $5,000 of USDC — and in return the contract mints you LP tokens. Those LP tokens are a receipt: they represent your percentage share of the pool, and you can burn them at any time to redeem your underlying assets plus everything they've earned. Deposit 1% of a pool's value and you own a claim on 1% of its reserves and 1% of its fees.

And fees are the reason to bother. Every swap pays a fee — typically 0.3% on Uniswap v2, and configurable (0.01%–1%) on Uniswap v3 depending on the pair. That fee gets added straight back into the pool, lifting the value of every LP token. On a pool doing $1 million in daily volume at 0.3%, LPs collectively earn $3,000 per day. Against $10 million of total value locked (TVL), that's roughly 11% APY from fees alone — paid in proportion to your share, with no action required from you.

But there's a catch baked into the mechanics, and it's the single most misunderstood part of being an LP. As prices move, the pool quietly rebalances your position — and you can end up worse off than if you'd simply held the two tokens in your wallet. That phenomenon has a name: impermanent loss. It's important enough to get its own section.

Let's make it concrete with the canonical example. Alice deposits 1 ETH and 100 USDC into a pool. At deposit, ETH is worth $100, so each side is $100 and her stake totals $200. The full pool holds 10 ETH and 1,000 USDC, giving Alice a 10% share.

Now ETH triples to $400 on the wider market. But the pool's formula doesn't know that — its internal price is still based on its ratio. So a gap opens, and arbitrage traders pounce: they buy the now-cheap ETH out of the pool until its internal price matches $400. To satisfy x × y = k at the new ratio, the pool ends up holding roughly 5 ETH and 2,000 USDC. The arbitrageurs pocket the difference — and that profit comes directly out of the LPs' pocket.

Alice withdraws her 10% share: 0.5 ETH + 200 USDC. At $400/ETH that's $200 + $200 = $400. Not bad — until you compare. Had she simply held her original 1 ETH and 100 USDC, she'd have $400 + $100 = $500. Providing liquidity cost her $100, a 20% haircut on her gains. That gap is impermanent loss.

Two things make it counterintuitive. First, it happens whether the price goes up or down — any divergence from your deposit ratio causes it. Second, it's called "impermanent" because if the price ratio drifts back to where you started, the loss evaporates. It only locks in — becomes permanent — the moment you withdraw. The standard divergence-to-loss table is worth memorizing:

• 1.25x price change → ~0.6% loss
• 1.5x → ~2.0% loss
• 2x → ~5.7% loss
• 3x → ~13.4% loss
• 4x → ~20.0% loss
• 5x → ~25.5% loss

Swapping is just the entry point. The same locked-reserve primitive turns out to be the plumbing under a huge slice of DeFi:

• Decentralized trading (DEXs). The original use case — Uniswap, Curve, PancakeSwap — lets anyone swap or list a token instantly, no listing committee required.
• Yield farming and liquidity mining. Protocols hand out their governance tokens as a bonus on top of swap fees to bootstrap deposits. You can also stake your LP tokens elsewhere to stack a second layer of rewards.
• Lending and borrowing. Pooled deposits become the supply that borrowers draw from on platforms like Aave, with interest rates set algorithmically by how much of the pool is utilized.
• Governance. Pooling tokens can aggregate voting power, lowering the threshold for proposals to pass in a decentralized organization.
• Synthetic assets and stablecoins. Pools provide the on-chain collateral and price reference needed to mint synthetic tokens that track real-world assets.
• Insurance and tranching. More advanced designs split a pool's risk and return into tranches, or pool capital to underwrite smart-contract cover.

The common thread: a liquidity pool is a programmable pile of capital. Once you can deposit assets into code that follows fixed rules, you can build a market maker, a money market, a voting bloc, or an insurance fund on the exact same foundation.

Pools are powerful, not safe by default. Honest education means naming the ways LPs actually lose money — and there are several beyond impermanent loss.

• Smart contract risk. Your funds live inside code, and code has bugs. A single flaw in a pool's contract — a reentrancy hole, a math error, a bad upgrade — can let an attacker drain the entire pool in one transaction. Hundreds of millions have been lost this way. Audits reduce the odds; they don't eliminate them.
• Rug pulls. Anyone can spin up a token and a pool. A malicious team can seed a pool, hype it, wait for real deposits, then pull their own liquidity and dump — leaving everyone else holding a worthless token in a near-empty pool. Permissionless listing is a feature and an attack surface.
• Impermanent loss. Covered above, but it bears repeating: on a volatile pair, fees may never catch up to the loss from price divergence. Many first-time LPs are surprised to withdraw less dollar value than they put in, even after collecting fees.
• Slippage and front-running. Large swaps move the price against you, and bots watching the public mempool can sandwich your trade — buying ahead of you and selling after — skimming value via what's called MEV (maximal extractable value).
• Systemic and oracle risk. Pools are composable, so one protocol depends on another. A broken price oracle or a depegged stablecoin can cascade through every pool that relied on it.

GaiaEx uses an order book model, not an AMM liquidity pool. The distinction matters for execution quality. Order books typically deliver tighter spreads and less slippage than AMM pools for large trades, because professional market makers actively manage their quotes rather than relying on a mathematical formula — and because you're not paying an impermanent-loss premium baked into the curve.

That said, liquidity pools and order books coexist in the broader ecosystem, and the smart move is to understand both. Many traders provide liquidity in AMM pools on Uniswap or Curve while trading actively on order-book venues like GaiaEx. The skills transfer directly: knowing pool mechanics lets you read DEX liquidity at a glance, spot arbitrage opportunities when an AMM's formula-driven price drifts from the order-book market price, and accurately estimate the true cost of executing a large swap before you click.

The DeFi ecosystem is interconnected. AMMs and order books are two answers to the same question — how do you price and settle a trade without a trusted middleman taking your money? — and the trader who understands how both work has the edge.