RRC · Desk conversation

The 18:00 snapshot shows the three broad benchmarks higher and volatility lower. The separate breadth read is only partial, so I would call the backdrop firm, not uniformly broad.

The day's public news is cross-current: France's 10-year yield crossed 5%, Hormuz flow estimates topped 14.5 million barrels a day but exclude bypass pipelines, and the IEA reported sharp gains in strategic minor metals. None gives us a verified link to this candidate set.

Breadth means the count of stocks rising versus falling in the sampled list. The intraday proxy leaned positive, but its end-of-day layer was not captured; one stale quote and one missing quote are part of that record. It is a session hint, not proof of full-market participation.

RRC has the clearest event-and-tape match. The event record shows +15.20% on 5.13 times average volume; the 17:40 observation was +16.92%. Price was just above its $52.41 VWAP. VWAP is the day's average traded price weighted by volume, so that says the move held around its main trading level for that session, not that it will persist.

The pre-cut SEC filing documents an all-stock EQT acquisition agreement; a secondary summary reports 0.987 EQT shares for each RRC share. In an all-stock deal the value moves with EQT's shares and the deal still has to close. I support Raw Try, not Try of the Day.

RRC's 20-session return was 41.38%, ahead of the XLE proxy by 29.21 points. The active model puts the chance of a positive return at the fifteenth follow-up session at 52.10%, but its adverse-MAE P10 is −24.30%. MAE is the worst dip from entry; P10 is a modeled severe-tail estimate, not a floor. Coverage is 19 of 23 features, with SEC inputs marked context-only. I support Raw, not Try.

RRC was 5.8 ATR above its 21-day average, and earnings are due tomorrow. ATR measures a recent typical daily price range; this distance is nearly six such ranges, so a sharp pullback could erase a substantial part of the move. I support Raw because the follow-up is testable, but that risk keeps me from Try.

I dissent and would call Caught. The deal may be real, but RRC has already repriced sharply, earnings are next, and the next 15 sessions depend on EQT's value as well as completion. I do not think the remaining path merits publication.

The cutoff is 18:02:39 UTC. I excluded the 18:22 options captures and FRMM's trajectory row retrieved after the cutoff. RRC's agreement record predates the cut, and its next check is traceable; I support Raw.

TTMI deserves credit for holding above VWAP, but its 20-session return lagged XLK by 25.78 points and the Q3-results summary is secondary. SAIQ's letter of intent is not recognized revenue; FRMM's reported GPU amount does not establish utilization or financing. LU remains below VWAP with much weaker relative performance.

XNCR is the strongest counterexample on relative behavior: it beat the healthcare proxy over 20 sessions, but the packet lacks the clinical outcome figures and the last pre-cut observation was below VWAP. XTND's reported backlog is pending work, not revenue; its tape was well below VWAP. I remain on Caught.

Reopen: RRC remains a testable Raw Try, but this setup is unusually unlike the model’s comparison cases. Treat its session-15 return and downside estimates with added caution.

Does a 7.3 typicality percentile mean a 7.30% chance of success?

No. It is a similarity percentile, not return odds: higher values mean a setup looks more typical of the model's comparison cases. RRC's 7.3rd percentile is unusually low, so its session-15 probability and downside estimates fit this case less well than the 19-of-23 feature coverage alone suggests. The adverse-tail estimate remains material, and the Raw Try rests on the documented agreement and observable checks, not the model.

That is the material correction. The sealed filing, tape and follow-up conditions are unchanged; the model-fit caveat changes how readers should weigh those estimates, not the selected candidate or the Raw Try resolution.