Corp-action forensics

measured, never reportedsample data

A stock splits and the token is supposed to follow. Sometimes it doesn't — a price left at the wrong scale, a multiplier never flipped. We line every venue up against the licensed reference feed and publish the verdicts, mishandlers first.

Mishandled — the receipts

GMEbGME · backedSPLIT MISS

reverse split 4:1 on 2026-06-15 — price still at pre-split scale — split not applied. Price sits at 0.25× the reference, and no on-chain split was declared.

Every verdict

Underlying	Session	Split	Venue	Type	Price ratio vs ref	Declared	Verdict
NVDA	2026-05-25	10:1 forward	aster	perp	0.998×	none	tracked the split
NVDA	2026-05-25	10:1 forward	avantis	perp	1.001×	none	tracked the split
NVDA	2026-05-25	10:1 forward	bitget	spot	0.999×	10:1	tracked the split
NVDA	2026-05-25	10:1 forward	bybit	spot	0.999×	10:1	tracked the split
NVDA	2026-05-25	10:1 forward	dinari	spot	0.998×	10:1	tracked the split
NVDA	2026-05-25	10:1 forward	hl	perp	1.001×	none	tracked the split
NVDA	2026-05-25	10:1 forward	ondo	spot	1.005×	10:1	tracked the split
NVDA	2026-05-25	10:1 forward	ostium	perp	1×	none	tracked the split
NVDA	2026-05-25	10:1 forward	variational	perp	0.998×	none	tracked the split
SMCI	2026-04-27	10:1 forward	ostium	perp	1.003×	none	tracked the split
GME	2026-06-15	4:1 reverse	aster	perp	1.003×	none	tracked the split
GME	2026-06-15	4:1 reverse	backed	spot	0.25×	none	price still at pre-split scale — split not applied
GME	2026-06-15	4:1 reverse	hl	perp	0.999×	none	tracked the split

Split + halt detection today; dividend / ex-date forensics land with a corporate-actions reference feed. Derived without issuer cooperation, like everything here.