True-Cost Index, October 2026.

Skontro's dollar series. Each exchange is priced separately on every measured asset, then averaged into one number. 15,167 measurements in 1,988 asset-hour rounds (1,985 rankable), observed 2026-10-01 → 2026-10-04 · 91.7% of scheduled asset-hour rounds so far (period still running).

Basket v2: the 25-asset series (from 2026-08-01). 48 implausible rows (fills more than 0.5% below the reference) never entered the index. Ranking is by net amount received; affiliate relationships never move a number. How “hidden spread” is measured: /true-cost/methodology. Corrections → hello@skontro.com, applied and dated visibly.

Skontro's founding EUR series — the longest continuous record, running on as a continuing service beside the dollar edition. Each exchange is priced separately on every measured asset, then averaged into one number. 16,432 measurements in 2,075 asset-hour rounds (2,059 rankable), observed 2026-10-01 → 2026-10-04 · 95.7% of scheduled asset-hour rounds so far (period still running).

Basket v2: the 25-asset series (from 2026-08-01). * = estimated, not measured: a list-price exchange with no public order book; excluded from the cheapest/dearest rounds. Not assessed = a thin order book or an exchange no longer compared; shown for the record, never ranked. 8 implausible rows (fills more than 0.5% below the reference) never entered the index. 179 thin-book rows (spread above 20% of the order size) are left out of the averages since September. Ranking is by net amount received; affiliate relationships never move a number. How “hidden spread” is measured: /true-cost/methodology. Corrections → hello@skontro.com, applied and dated visibly.