Most LPs pick a fee tier by vibes, then wonder why a busy-looking position earns almost nothing.
The fee tier is not a setting you tick and forget. It is a price you are quoting to the market for taking on a specific risk. Quote it wrong and you either scare off the volume or get paid too little to hold the bag. Here is how to think about it like someone who wants to keep their capital.
What the fee actually pays you for
A fee tier is compensation. It helps to know what you are being compensated for.
Every time the price of a pair moves, someone can arbitrage your pool back toward the wider market price, and that trade happens at your expense. You end up selling the asset that is rising and buying the one that is falling, a fraction of a second too late, over and over. That is the real job of a liquidity provider: you are the counterparty who is always slightly behind. Impermanent loss is the visible symptom. Your fees are the offset.
So the first rule is blunt. The fee has to be large enough to cover the price moves that happen between arbitrage trades. A pair that barely moves needs almost nothing. A pair that can swing ten percent in an afternoon needs a real cushion, or you are subsidizing trading bots for free.
That one idea, fee income versus expected price movement, is most of the decision. Everything below is refinement.
The trap: a high fee does not mean high income
New LPs assume the 1.00% tier pays five times more than the 0.20% tier. It pays five times more per trade. The catch is that trades do not have to come to you.
Routers send order flow to whatever pool offers the best net price. If the same pair already has a deep pool at a lower fee, that is where the volume goes, and your expensive pool sits there collecting dust. You set a high price, and the market politely walked past you.
The tier that earns is rarely the one with the biggest number. It is the one that captures flow while still covering its risk.
The one number that beats TVL
Total value locked tells you how crowded a pool is, not how much it earns. The number worth staring at is fee yield: fees produced divided by the liquidity sitting in the pool.
Take a single pair, TOKEN/USDC, that happens to live in two tiers at once.
The 0.30% pool holds 2 million in liquidity and trades about 1 million per day. Daily fees are 1,000,000 times 0.003, which is 3,000. Across a year that is roughly 1.1 million against 2 million of liquidity, near 55% before any costs.
The 0.05% pool holds 8 million and trades about 40 million per day. Daily fees are 40,000,000 times 0.0005, which is 20,000. Across a year that is roughly 7.3 million against 8 million, near 91% before any costs.
The lower-fee pool pays the higher yield, because it caught the flow. The higher rate lost to the tighter price. If you had chased the bigger fee number, you would have earned less.
Two honest caveats. These are gross figures, before impermanent loss and gas. On a volatile pair, impermanent loss can eat the whole thing and then some. Fee yield tells you where a dollar earns more, not whether you keep it.
Match the tier to how the pair behaves
Pairs are not interchangeable, so tiers should not be either. The exact percentages available depend on the protocol and chain you are on, so treat these as the common pattern rather than fixed law, and confirm what your venue actually offers.
Stable to stable (a dollar coin against another dollar coin). Price barely moves, so the arbitrage cost against you is tiny, and volume is enormous and ruthlessly price-sensitive. The lowest tier on offer, often 0.01%, usually wins. Anything higher just hands the flow to a cheaper pool.
Pegged or correlated (a staked asset against its base asset, or two wrapped versions of the same coin). Small but real drift, with the occasional depeg scare. A low tier, commonly 0.05%, tends to fit. Not rock bottom, because the peg can snap and you want a little cushion when it does.
Blue chip volatile (a major asset against a stablecoin). Meaningful daily moves, deep and heavily contested volume. The middle tier, often 0.30%, is the usual home. High enough to cover ordinary swings, low enough to stay inside the routing path.
Long tail and fresh launches (small caps, new tokens, thin books). Violent moves, lumpy volume, and traders who often know more than you do. The top tier, often 1.00%, exists for precisely this. If you are early on a token that genuinely trades and has no competing pool, that tier can pay very well. It is also where forgotten liquidity goes to rot, so the volume has to be real, not hoped for.
Check who is already parked there
Before you deposit a single token, pull up every existing pool for the pair. For each one, note the liquidity, the recent volume, and the fee yield. Then ask two questions.
Is there already a deep pool at a lower tier that owns the flow? If yes, opening a higher tier for the same pair on the same route is usually a graveyard. You would be quoting a worse price into a market that already has a better one.
Am I adding to a pool that already earns, or starting a lonely one and hoping volume shows up? Hope is not a fee source.
Range and tier are the same decision
In a concentrated liquidity pool you also choose a price range, and that choice is welded to the tier.
Low-fee stable pools reward very tight ranges. Price sits still, so you concentrate hard and collect dense fees on a narrow band. High-fee volatile pools punish tight ranges. Price wanders out of your band, your fees stop, and impermanent loss keeps working, so you go wider and accept a thinner rate per dollar in exchange for staying active.
Picking a low tier and a wide range on a wild pair is the worst of both worlds: a small rate spread across diluted liquidity. Match your aggression on both dials to the temperament of the pair.
Run this before you deposit
What does this pair actually do in a normal day? Estimate the typical move, not the calm days only. Does the fee cover that move with margin to spare, or am I quietly funding arbitrage? Where is the volume for this pair already trading, and at which tier? Compute fee yield, fees divided by liquidity, for every existing tier. Rank by that, not by size. Does my range match the tier? Tight for calm pairs, wide for wild ones. After a rough guess at impermanent loss, is anything left worth the risk?
The short version
The right fee tier is not the highest number and not the most crowded pool. It is the one where the fee covers the real risk of the pair and still sits where the volume flows. Two dials, volatility and competition, set everything else.
Price your liquidity like you mean it, check the data before you commit, and revisit the choice when the pair starts behaving differently. Pairs change moods. Your tier should be allowed to change with them.
Educational content only, not financial advice. Providing liquidity carries real risk, including impermanent loss and the potential loss of your capital. Do your own research before committing funds.
