Providing liquidity gets pitched as passive income: deposit two tokens into a pool, collect a share of trading fees, done. The pitch isn't false, exactly. But it describes the revenue side of a position that also has a cost side, and on many pools the cost side is the bigger number. People who deposit based on an advertised APY and withdraw months later are often surprised to find they'd have done better just holding the tokens.
The surprise comes from a mechanism, not bad luck. So before walking through the steps — which are genuinely simple — it's worth understanding what the pool actually does with your deposit, because that's where the outcome is decided.
What a Liquidity Position Actually Is
Most DEXs are automated market makers. There's no order book; instead, a pool holds two assets — say ETH and USDC — and prices trades with a formula. The classic version is constant product: the quantity of asset X times the quantity of asset Y must equal a constant. When a trader buys ETH from the pool, the ETH balance falls and the USDC balance rises, and the formula moves the price along a curve. Every trade pays a fee — 0.3% is the traditional tier, with 0.01% to 1% common depending on the pair — and that fee accrues to the pool, which means to the people who funded it.
That's you, if you deposit. You add both assets at equal value, and you receive LP tokens representing your share of the pool. Your share of every fee from then on is proportional to your share of the pool. So far, so passive.
Here's the part the APY number doesn't show. The pool rebalances against you, continuously, by design. If ETH rises, arbitrageurs buy ETH from the pool until the pool's price matches the wider market — which drains the pool's ETH and fills it with USDC. Your position automatically sells the asset that's going up and accumulates the one that's going up less. When you withdraw, you get back a different mix than you put in, and that mix is always worth less than if you'd simply held the original tokens. The gap is called impermanent loss, though divergence loss is the honest name — it's only "impermanent" if relative prices return to where they started.
The size of the gap is pure math. If one asset doubles against the other, the divergence loss is about 5.7% versus holding. A 4x move costs about 20%. Stable pairs that stay pegged produce almost none — which is why stablecoin pools exist and why their fees are lower. A liquidity position is, in effect, a bet that fee income will outrun divergence loss. Sometimes it does. On volatile pairs, frequently it doesn't.
The Steps, and What Each One Does
The actual process is short. You pick a pool — that's a pair, a fee tier, and a venue. You acquire both assets in roughly equal value, since most pools require balanced deposits. You grant the DEX contract a token approval (this is its own risk surface — approvals are standing permissions, and revoking unused ones is basic hygiene). You deposit, receive your LP position, and from that point your assets are working — being traded against — until you withdraw.
One step deserves more attention than the interfaces give it: pool selection isn't cosmetic. The same pair can exist at multiple fee tiers, and the trade-off is real — lower tiers attract more volume but pay less per trade, higher tiers the reverse. And the venue matters because the pool's contract is where your assets actually live. An audited contract with years of incident-free operation and one deployed last month are different risk objects, whatever the yields say.
Concentrated Liquidity Changes the Job
Newer AMM designs — Uniswap v3 onward, and most serious venues now — let you concentrate liquidity into a price range instead of spreading it across the whole curve. Capital inside an active range earns multiples of what the same capital earns in a full-range position. The cost is that everything sharpens: divergence loss accelerates inside the range, and if price exits your range entirely, you're left holding 100% of one asset and earning nothing until price returns or you reposition.
This turned liquidity provision from a passive deposit into something closer to active market-making. Admittedly, a full-range position in a classic pool still behaves the old way. But the high advertised yields you'll see on aggregators are mostly concentrated positions, and those numbers assume someone is managing the range. Vault protocols exist to do that management for you — for a fee, and with another contract layer added to the risk stack.
Where the Constraints Live
The divergence-loss math is a hard constraint; it falls out of the pricing formula and no interface can remove it. Smart contract risk is structural — pool exploits have drained LPs before and the deposit is only as safe as the contract holding it. There's also a subtler drag that researchers call loss-versus-rebalancing: arbitrageurs profit from every gap between the pool's stale price and the live market price, and that profit comes out of LP returns. It's the formalization of why passive pools systematically underperform what their fee income suggests.
What's changing is mostly responses to that problem. Dynamic fee designs and hook-based pools (Uniswap v4's architecture) try to charge arbitrageurs more than retail flow. Intent-based routing and order-flow auctions are moving some volume off AMM curves entirely. None of this is settled — the LP profitability question is an active research area, which is itself a signal about how often the naive version works.
What Would Confirm or Break This Picture
Confirmation: analytics platforms continuing to show fee income net of divergence loss as the deciding variable between profitable and unprofitable pools, and dynamic-fee pools demonstrably narrowing the arbitrage drag.
Invalidation: a major audited AMM suffering a pricing-formula exploit would break the "the math is the safe part" framing. A sustained period where passive full-range positions on volatile pairs reliably beat holding would mean the divergence-loss model is missing something material.
Timing Perspective
Now: The mechanism is mature and the tooling is good — position calculators and pool analytics make the fee-versus-divergence trade measurable before depositing, which wasn't true in earlier cycles. Next: Watch whether v4-style dynamic fees actually shift returns toward LPs; that's measurable and worth waiting on. Later: If order flow keeps migrating to auction-based routing, the passive-pool fee base shrinks — a structural question, not a this-quarter one.
Boundary
This explains what a liquidity position is and how its profit and loss are generated. It isn't a recommendation to provide liquidity, an endorsement of any venue or vault, or a claim about what any pool will return. Specific fee numbers vary by pool and change over time. Tax treatment of LP positions — which is genuinely messy in several jurisdictions — is outside scope. The static explanation lives here; the tracked version lives elsewhere.
