24-Hour Movers (2026-09-08)

Roivant +19% in a single session, with the stock tracking an all-time high — yet the readout was not Roivant's own. The 8-K it filed before the bell points at PHocus, a trial run by its subsidiary Pulmovant: mosliciguat came back positive in pulmonary hypertension associated with interstitial lung disease.

MechanismRoivant is a holding platform that parks each drug candidate inside its own subsidiary. The asset sits at Pulmovant, and the only thing a public investor can actually buy is the parent's single line of stock, so a subsidiary's mid-stage readout is absorbed whole by the parent — which is why the move gapped premarket and was priced at the opening bell. Note what the document is: an Item 7.01 voluntary disclosure. The press release travels as an exhibit, and the filed text itself carries no efficacy figures and no formal endpoint definitions. So this paragraph explains only why the news can move the price and prints no clinical number at all: if you cannot click through to a figure from this card, we do not print it.

Event time

2026-09-08 · premarket

Lag

priced at the bell — not measurable

Novartis -14% and Dyne -16% in the same session, tied together by one disease. The 6-K Novartis filed before dawn says the Phase III HARBOR study of del-desiran missed its primary endpoint in myotonic dystrophy type 1 — and Dyne's most advanced program treats exactly that disease. The document was already on EDGAR; hours later the trade press was still describing Dyne's premarket drop as having no confirmed catalyst.

MechanismThis card has two legs and only one of them is really joined. The Novartis 6-K is the only first-hand document of the day: HARBOR missed its primary endpoint in myotonic dystrophy type 1, and Novartis itself describes that disease in the filing as a progressive neuromuscular condition with no approved treatment options. The failure therefore read as a verdict on the indication rather than on one company, and Dyne, which is betting on the same disease, was sold alongside it. Dyne filed nothing of its own that day; its decline is explained by someone else's readout, not by news of its own. The third name, Structure Therapeutics, sits in this cluster only because it fell on the same day: its 8-K covers topline data from oral small-molecule amylin and GLP-1 weight-management programs and has no overlap with HARBOR or with this indication. We print that on the card rather than let the grouping imply it was knocked down by the same readout.

Event time

2026-09-08 · premarket

Lag

n/a

Solaris +16% in a session, and it was not an earnings day: before the bell the company raised adjusted EBITDA guidance for both Q3 and Q4 and set an outlook for Q1. The catalyst badge on this card points at a generic pre-bell energy sector wrap. That is not the reason; the reason is the guidance item that crossed earlier the same morning.

MechanismA mid-quarter guidance raise is a different object from a raise delivered with earnings. On an earnings day a company hands over a finished quarter and a fresh expectation at once; a standalone revision between reports says that what management saw in the past few weeks was already enough to lift the number without waiting. Here two consecutive quarters were lifted together and a third was given an outlook, and three steps moving in the same direction are harder to dismiss as a one-off than any single one of them. The item crossed before the opening bell, so the repricing was complete at the bell and there is no lag to measure. One more thing to state plainly: the attribution layer tags this as secondary, because the strongest evidence matching in both time and scope was a newswire item rather than a filing by the company itself. The badge reflects the evidence we actually hold, and we do not upgrade it.

Event time

2026-09-08 · premarket

Lag

priced at the bell — not measurable

Zacks report. SMR gapped at the bell and rose 15.3% over 24 hours on 5.9× baseline volume — the news broke pre-market, so the lag cannot be measured.

Event time

2026-09-08 · premarket

Lag

priced at the bell — not measurable