Why two interfaces can report different price-impact percentages

Two impact percentages can disagree without the underlying receiving amounts disagreeing. Before comparing them, identify each screen's benchmark, price direction, fee treatment and rounding convention.

Separate the possible measurements

In a hypothetical example, a spot-rate extrapolation suggests 1,000 output tokens. A fee-free size calculation gives 992, and the final fee-inclusive quote gives 989. A field measuring the curve-only shortfall reports 0.8%; a field measuring the full quote gap reports 1.1%.

Those figures describe different boundaries. The latter should not automatically be interpreted as pure own-trade price impact. Nor should its additional 0.3 percentage points be assumed to be a separately payable fee without checking the quote.

Check the benchmark and units

One interface can use a pool-local reference while another uses an external token valuation. A stale reference creates another difference. Output-per-input and input-per-output percentage changes are also not numerically identical.

Uniswap's glossary relates price impact to the difference between mid price and execution price caused by trade size. For an interface comparison, inspect how the specific field implements that concept.

Compare the actionable amounts

Use final quoted output, minimum output or maximum input, and separately paid costs to assess the offer. If those match but the percentages differ, investigate display methodology rather than declaring one route economically worse.

Keep enough precision to see whether rounding alone explains the gap. A field showing “less than 0.1%” and another showing 0.04% may be consistent.

If the methodology is undocumented, report that the impact labels are not directly comparable. Do not reverse-engineer a precise fee from them without corroborating amounts. A clear comparison can still rank matched net outputs while acknowledging that the explanatory percentage fields use unresolved conventions.