Price impact by trade size (this fee tier)
Same pool, same fee tier — only the trade size changes, to show how impact scales non-linearly as size grows relative to pool depth.
| Trade size | % of TVL | Execution price | Price impact |
|---|
Same trade, different fee tiers
Fee tier changes the input haircut but not the underlying curve — impact is dominated by trade-size-vs-depth, not the fee.
| Fee tier | Execution price | Price impact |
|---|
Why the same $25k swap can cost 0.3% on one pool and 8% on another
A constant-product AMM pool holds reserves of two tokens, x (base) and y (quote), such that x·y = k stays constant. When you swap Δy of the quote asset in to buy the base asset, the pool computes new_y = y + Δy·(1−fee), then new_x = k / new_y, and pays you out Δx = x − new_x. Your average execution price is Δy/Δx, always worse than the pre-trade spot price y/x — that gap, expressed as a percentage, is price impact. Because the curve is hyperbolic rather than linear, impact grows faster than trade size: doubling your trade size roughly doubles-plus impact, not exactly doubles it, which is why the size table above bends upward instead of scaling in a straight line.
Pool depth (TVL) is the entire story here. A $25,000 swap against a $2,000,000 pool touches about 1.25% of the reserves and produces mild, often-forgettable impact. The identical $25,000 swap against a $200,000 pool touches 12.5% of reserves and can produce double-digit impact — the same trade, the same fee tier, wildly different outcomes purely because of how much liquidity is sitting on the other side. This is why "what's the price impact" has no fixed answer independent of pool size — it's always a function of trade size relative to depth, never trade size alone.
Price impact is mechanical and known in advance; it isn't the same thing as total slippage, which also includes price movement from other trades executing before yours in the same block or mempool. On a quiet pool the two numbers converge. If you're providing liquidity rather than swapping, the relevant risk is different — see impermanent loss for full-range positions or the concentrated liquidity calculator if you're using Uniswap V3-style ranges, where the same constant-product curve gets compressed into a narrower band.