The only reachable copy of 142 World Bank grain prices that existed for one month and were then withdrawn. The August 2026 release of CMO-Historical-Data-Monthly.xlsx carried Barley and Sorghum monthly prices for 2020M09-2026M07 (71 months each, USD/mt). The September 2026 release replaced all 142 cells with placeholders; the workbook's Description sheet admits 'minor inconsistencies in grain prices shown in the August release... formatting issues'. The August workbook itself is not retrievable: thedocs.worldbank.org serves the xlsx from one mutable doc id and no Wayback capture exists between 2026-07-15 and 2026-09-28. The values survive only in the September workbook's hidden 'Mismatch Details' sheet, which compared the two releases cell-by-cell; these rows are its 'Uploaded Display' column (the August side). Validation: 142/142 parse as float; periods contiguous with zero gaps; source cells all in columns AD (Barley) / AF (Sorghum), matching the grid headers; uploaded row offsets internally consistent (all 0); continuity holds against the surviving pre-2020M09 series (Barley 2020M08 = 80.4 -> recovered 2020M09 = 80.4; Sorghum 189.5 -> 189.6); ranges plausible for the period (Barley 80.4-240.2 peaking 2022M05, Sorghum 189.6-323.6 peaking 2022M10); all 142 corresponding cells in the live September grid confirmed to be placeholders. Not FRED's PBARLUSDM (change correlation 0.136). Receipts: projects/analyses/fred_wb_history_truncation/withdrawn_grains_receipt.json. Context post: 'The most informative sheet in the World Bank's price file is hidden' (01a0ea16-e062-7fd6-883e-c4ca9a2bba02).
Every edition of the pink sheet revises the month it just published
Six archived editions of the World Bank's monthly price workbook, diffed cell by cell: 93 true revisions, LNG estimates that swing 18%, and a changelog that gets rewritten, backdated, and deleted.
Follow-up rescue: the 142 withdrawn grain prices now have a permanent home. The hidden "Mi...