Published July 8, 2026 · by ILCalc
Providing liquidity is sold as passive income, but it is really a trade: you collect fees in exchange for taking on impermanent loss (IL) and gas costs. Whether that trade pays depends entirely on the pool, the volatility, and your time horizon. This guide gives you one clean way to decide — the fees − IL − gas vs HODL framing — plus honest data on how often LPs actually come out ahead, and a checklist you can run before committing a single dollar.
Strip away the marketing and liquidity provision reduces to a single comparison. Your ending value as an LP versus what you would have had if you had simply held the two tokens in your wallet:
Net edge = fees earned − IL$ − gasIf that number is positive, LPing beat holding. If it is negative, you would have been richer doing nothing. IL$ is the dollar gap that opens up between the pool's rebalancing and just holding, and it is driven by price movement. Using the constant-product formula, IL = 2·√k/(1+k) − 1, where k is the relative price change between the two assets. The IL is symmetric — a 2× move up and a 50% move down produce the same drag.
| Price move | k | Impermanent loss | IL on $10,000 |
|---|---|---|---|
| 1.25× / −20% | 1.25 | ≈ 0.6% | ≈ $60 |
| 1.5× / −33% | 1.5 | ≈ 2.0% | ≈ $200 |
| 2× / −50% | 2 | ≈ 5.7% | ≈ $570 |
| 3× / −67% | 3 | ≈ 13.4% | ≈ $1,340 |
| 4× / −75% | 4 | ≈ 20.0% | ≈ $2,000 |
| 5× / −80% | 5 | ≈ 25.5% | ≈ $2,550 |
So a $10,000 position that lives through a 2× divergence carries roughly $570 of IL. To make LPing worth it, the fees over that same window must clear $570 plus whatever you paid in gas to enter and exit. The full derivation, including how to turn a fee APR into a break-even threshold, is in impermanent loss vs trading fees.
Here is the part most yield dashboards won't show you: multiple independent on-chain studies of Uniswap V3 found that roughly half of liquidity providers earned less than if they had simply held their tokens, once impermanent loss is netted against fees. That is not a fringe result — it is the base rate. LPing is not free money; it is a strategy with real dispersion between winners and losers.
Why do so many lose? Because headline APRs advertise gross fees and quietly ignore IL. A pool showing "40% APR" can still be a net loss if the pair diverges hard. The LPs who win tend to concentrate in pools where the two assets move together, where fee volume is genuinely high, and where they are not paying gas to constantly rebalance.
Uniswap V3 lets you concentrate capital into a price range, which multiplies your fee earnings — but it multiplies IL by roughly 2–4× for typical ranges, and far more for very tight ones. The trap is subtle: once price exits your chosen range, the position becomes 100% one asset and stops earning fees entirely. You are now holding the loser of the pair with none of the income that justified the risk.
Tight range = more fees while in-range, more IL, and total fee shutoff once you're out.Concentrated liquidity is an active job, not a passive one. If you are not prepared to monitor and re-range, a wide range or a plain V2-style full-range position is usually the more honest choice. Techniques for damping this are covered in how to avoid impermanent loss.
k=1, growing only quadratically. A full 10% depeg (k=0.90) is still only about 0.14% IL. Almost any fee income clears that bar.Staking pays a steadier yield with no impermanent loss — you deposit one asset and earn a predictable return, typically low single digits for ETH and higher on some chains. The tradeoffs are different, not absent: your capital may be locked for an unbonding period, some networks impose slashing, and you take on validator or protocol risk. Staking is the better fit when you want smooth, defensible yield on a single asset you already hold. LPing can out-earn it, but only in the right pool and with active attention.
IL% ÷ (days/365). If you expect 5.7% IL over 90 days, you need roughly 23% fee APR just to tie holding.Run the numbers in the ILCalc calculator →
Usually only in stablecoin or pegged pools to start. Those have near-zero IL, so a beginner can learn how fees, gas, and position management work without a volatile pair quietly draining the position. Avoid tight-range V3 and volatile pairs until you can compute break-even confidently.
It depends only on price divergence. A 1.25× move is about 0.6%, a 2× move about 5.7%, and a 3× move about 13.4% — and the loss is symmetric for moves up or down. Pegged pairs stay near zero. Estimate your expected divergence, read the IL off the table, and compare it to expected fees.
Not entirely, but you can shrink it dramatically by choosing correlated or pegged pairs, using wider ranges, and holding through short-term noise rather than exiting at a divergence peak. The trade-offs and mitigation tactics are detailed in how to avoid impermanent loss.