METHODOLOGY
Every conclusion traces back to an input and formula
Actual ledger values, verifiable replays, modelled ranges, and unavailable costs are labelled separately.
Ledger net result
Realized P&L minus trading commission, plus funding received, minus funding paid and other identified costs.
Cross-venue replay
Executed notional and Maker/Taker behavior stay fixed. Only a verified target fee schedule is replaced.
Confidence
Ledger coverage, liquidity role, fee assets, unclassified rows, and schedule freshness contribute to a 0–100 score. Below 70, or when the target schedule is more than 30 days unverified, the report shows the cost breakdown only and offers no venue-switch conclusion.
Why the three result types stay separate
Exact values come from records that already exist. A verifiable difference comes from applying a verified schedule to the same executed notional. A modelled range comes from bounds introduced by a missing field. Merging them would imply the modelled part is as reliable as the ledger, so the interface and the formulas keep them apart.
Why current rates never rewrite history
Funding differs every settlement period, and fee schedules change with tier and promotions. Recomputing last month with today's numbers produces a precise-looking figure that cannot be checked. The target venue's historical funding cannot be verified from your source ledger, so it is always listed as excluded.
Boundaries of the replay
Only variables with evidence are substituted: notional and maker/taker split stay fixed, and just the verified target schedule changes. Depth, execution quality, slippage, and product availability are outside the model and are never converted into an amount.