ENTG FY2025 margins are wrong in both Yahoo and roic.ai

Claim

The quoted ~41% EBITDA margin is roughly 2.5x the real figure of 13–17%. A 41% margin would have made ENTG the most profitable name in its cohort and was a significant part of the case for it — the error pointed in the direction that flattered the thesis.

Why this one matters beyond ENTG

Both vendors were wrong the same way at the same time. The standard control — pull the number from a second source and compare — returns agreement and false confidence. Only the filing is independent.

This is the concrete case behind [[principle-primary-source-beats-vendor]]. Cross-vendor agreement is evidence of a shared upstream source, not of correctness.

Guard

If a margin would make a company a category outlier, treat that as a data-error signal first and an edge second. Check the filing before the number enters any comparison.

What would falsify this

The feeds correcting after the Aug 4 print, which would make this an ingest-timing artifact rather than a persistent error. recheck is set to that date. Even if they correct, the lesson about cross-vendor agreement stands.

History

• 2026-07-30 — found during primary-filing review of the semi-materials sweep.