Build a swap-cost waterfall without overlapping components

A cost waterfall works only when each component covers a different part of the path from benchmark to final net value. Start with quantities or values that reconcile exactly, then split a component further only when the evidence supports that split.

A waterfall that adds up

Use hypothetical values expressed in one comparison currency. The pre-trade reference values the input at 1,000. The accepted quote is worth 994. The actual received output is worth 993 at the same valuation basis. Separate payments total one.

Final net value is 992. The three components sum to the full eight-unit difference without overlap.

Do not add explanatory percentages twice

The six-unit reference-to-quote gap can contain fees, size effects and an initial reference mismatch. Adding a displayed pool fee and impact amount to that six without removing their embedded contributions overstates the total.

Impact and quote-to-fill slippage describe different comparisons. Define their denominators and timing before assigning them to the waterfall.

Leave uncertain attribution unallocated

You can know the total gap without knowing its exact causes. Label an unresolved component “reference-to-quote difference” instead of inventing separate fee and liquidity amounts that happen to sum to the total.

Keep reference prices fixed across the stages when measuring execution. If you also want to show later market valuation, add that as a separate subsequent stage. A price move after receipt should not silently alter the execution-loss components. The waterfall's purpose is transparent reconciliation, not a claim that every shortfall was avoidable.