Biotech desk — session 2026-09-09

All price/filing-age/countdown from session 2026-09-09. Data degraded: B7, B8 price fetch fail both sources (Nasdaq 400, Yahoo 429) — both already NO_DATA (merger-closed weeks ago), unaffected elsewhere.

Bottom line

4 ACT, 2 SETUP, badly over-committed. A17, A23, A3 hold ACT a 2nd day; B3 joins ACT on a same-day -[redacted] collapse. 6 actionable (ACT+SETUP) would request [redacted] vs [redacted] ceiling — scale every size by 0.59, down from 0.76 yesterday. A17's slide still has no confirmed cause after two sessions — override stands, not a buy. A23 intact thesis, buyable at scaled size. A3 pure technical (catalyst support gone since yesterday's guidance push-out). L7 newly SETUP, no capitulation volume, no explained cause either — same pattern as A17, do not chase. L1 (new WATCH) carries active securities-fraud litigation over a February 2026 cardiac safety signal — treat as impaired, not oversold. L9/A1/B25 vetoes stand, unrefuted. B1 exit still fired.

Big movers (ranked by sigma)

• B3 -11.1%, -4.81σ (typical day 2.3%), gap -1.1% at open, XBI -1.6%, excess -9.5%. Shares down ~40% since May 2026 after Merck's rival oral PCSK9 inhibitor won approval; Q2 revenue $3.7M vs $19.15M a year ago, net loss widened to $64.14M from $17.36M. PREVAIL CVOT interim in 22 days (2026-10-01) is the thing that can still move this.

Signal table — WATCH and above, or carrying a veto

Regime uptrend: XBI [redacted], +17.3pp vs SMA200, 60d +[redacted]. Exposure [redacted] (ACT+SETUP) vs [redacted] ceiling — badly OVER-COMMITTED, scale every size by 0.59.

44 further names at NONE, no veto. 2 NO_DATA (B7, B8 — merger-closed, delete from watchlist.toml, repeat flag unaddressed since 09-04).

What changed since 2026-09-08

• New ACT: B3 (prev WATCH) — same-day -11.1% collapse, capitulation volume confirmed, catalyst 22 days out.
• New SETUP: L7 (prev NONE) — no capitulation volume, ACT blocked on that alone.
• New WATCH: L1, B20, B17, B12, B21, B10, B18 (all prev NONE) — broad idiosyncratic weakening across the B-bucket, sector-relative not sector-wide (regime still uptrend).
• Exposure worsened: 112.0% → 145.0%, scale factor 0.76 → 0.59 — six names actionable now vs four yesterday.
• A17, A23, A3 hold ACT a second day. A19 holds SETUP a second day.
• A3's new 8-K (item 8.01) is a routine updated corporate presentation — checked, not material.
• L9, B25, A1 vetoes stand, unchanged in substance, ages advanced. B1 catalyst_resolved exit still fired.
• New Form 4/144 filings: B3, B19, B24, B4, and seven B13 ownership filings (routine, not reviewed line by line — none material to tier).

Per-name analysis

Triage found 36 qualifying names (ACT/SETUP/WATCH, hard veto, recent_events, new filing, exit flag, or catalyst inside 30d). Capped at 12 deep-dives by (tier, then bucket A before B before lottery), with the three standing veto/exit names pulled forward over lower-priority WATCH-tier B repeats already covered on 09-08. Remaining 24 qualifiers get one line each below.

A17 [ACT, bucket A, max 28.0% x 0.59 = 16.5%]

[redacted], RSI 20.12, capitulation volume YES (1.99x), zone [redacted]-[redacted] in zone, mechanically ACT. Weak conviction (-1) — checked again, still no confirmed cause for a slide now 5 sessions old (-[redacted] over 5d), [redacted] insider selling/120d, RSI in the distress bucket not the discount bucket. Override stands: not a buy. Two sessions of unexplained weakness is worse than one — wait for the cause before sizing.

A23 [ACT, bucket A, max 28.0% x 0.59 = 16.5%]

[redacted], RSI 30.08, capitulation volume YES (1.69x), moderate conviction (+2), in zone. No new bad news — company confirmed [redacted] cash, funded through 2028, presenting at 3 conferences this month. Thesis (oral GLP-1 exposure) intact. Buy: entry [redacted]-[redacted], size [redacted] of allocated capital, invalidation [redacted], catalyst = Body Composition readout ~2026-10-01 (Q4 guided, no month given) ([redacted] of 5d volume short). 20-session technical horizon; the catalyst is what justifies holding past it.

B3 [ACT, bucket B, max 28.0% x 0.59 = 16.5%]

[redacted], RSI 27.5, capitulation volume YES (1.92x), moderate conviction (+2). Cause found: Merck's rival oral PCSK9 approval (since May) plus a weak Q2 (revenue [redacted] vs [redacted] prior year, loss widened to [redacted]) — idiosyncratic and fundamentals-driven, not noise. In zone [redacted]-[redacted], invalidation [redacted] is only [redacted] away — tightest stop on the desk. Runway 17.72q. PREVAIL CVOT interim in 22 days (2026-10-01, expected) is the thing that reprices this. Two 5.02 officer-departure flags (26d), routine. Buy only for the catalyst, not the chart: entry [redacted]-[redacted], size [redacted], invalidation [redacted] ([redacted] away — expect to be stopped on noise), catalyst PREVAIL interim.

A3 [ACT, bucket A, max 28.0% x 0.59 = 16.5%]

[redacted], RSI 34.04, capitulation volume YES (1.99x), in zone [redacted]-[redacted]. Today's 8-K is routine (updated investor presentation, checked, no new data). 4WARD topline still moved to H1 2028 as of yesterday — no near-term catalyst. Runway 9.73q funds the financing leg independent of any binary. Buy: entry [redacted]-[redacted], size scaled ~[redacted] (below the [redacted] ceiling given no catalyst support), invalidation [redacted], pure 20-session technical bounce — size and hold accordingly ([redacted] of 5d volume short).

L7 [SETUP, bucket lottery, max 5.0%]

$1.73, RSI 34.83, no capitulation volume — stays SETUP. Checked for a cause: nothing found beyond routine Citi conference participation and the already-known CFO retirement (Nov 6, soft flag, 7d). $27.8M-scale insider selling not present here but 58.5% of 5d volume short. Runway 8.91q. Same pattern as A17 — unexplained weakness, do not chase. Not a buy, watching only.

A19 [SETUP, bucket A, max 28.0%]

$33.78, RSI 31.38, no capitulation volume, stays SETUP. Runway 3.89q, cash-only confirmed, thin against a binary. DAYBREAK Phase 3 one-year topline (wet AMD, ~690 enrolled) still guided September 2026, in 21 days — no readout yet, no new news. 17.5 days-to-cover, extreme squeeze fuel either direction. Not a buy — watching into DAYBREAK only.

L1 [WATCH, bucket lottery, max 5.0%]

$4.92, RSI 37.1, weak conviction (+0). Found the cause: Phase 3 HERO trial voluntarily paused Feb 27 2026 over reversible cardiac observations at above-target doses, and multiple securities-fraud class actions now filed (lead-plaintiff deadline Oct 13 2026) alleging the drug was misrepresented as ~99% gut-restricted with no serious adverse events. Stock down 71% from the Feb 2025 IPO price. This is a damaged-to-dead thesis, not an oversold bounce candidate — the RSI/zone signal is noise against active litigation over the drug's safety profile. No position.

L9 [NONE, bucket legacy, max 28.0%] — hard veto stands

$0.579. Listing-deficiency veto (8-K item 3.01, 40d) unrefuted. 3 insiders bought $84,589/120d — the strongest available counter-evidence, still not enough against delisting risk. Zone stale -27%. Runway 3.04q, cash-only confirmed. No position.

B25 [NONE, bucket B, max 28.0%] — hard veto stands

$135.61. Item 3.02 veto (unregistered equity sale, 8d) is a $3B convertible that only dilutes if price triples — mechanical, not urgent, but still a hard veto until refuted by the letter of the rule. +123.1pp vs XBI over 20d already priced in. $35.7M insider selling/120d. Not actionable either way.

A1 [NONE, bucket A, max 28.0%] — 2 hard vetoes stand

$0.6439. NT 10-Q late filing (23d) and 0.85 quarters of cash-only liquidity ($11.24M, confirmed by full XBRL sweep) both unrefuted — a dilutive raise is near-certain. 7 insiders bought $7.87M/120d, still not enough against near-certain dilution. LYTENAVA US launch guided by year-end 2026, $50-75M first-year revenue guided. Falling knife, not a dip. No position.

B1 [NONE, bucket B, max 28.0%] — exit fired again

$74.11. catalyst_resolved: 2026-08-28 ACACIA-HCM conference passed 12 days ago, primary endpoint already met and priced — still unresolved by the user if held. Re-underwrite around MAPLE-HCM PDUFA, Nov 14 2026 (66d, confirmed), a distinct indication (obstructive HCM). Aficamten sNDA (non-obstructive HCM) still guided Q4 2026. No position recommended.

Also flagged — 24 more, one line each

B20(B) $23.84 new WATCH, above zone by 2% (edge case, borderline), 169.64q runway strongest on desk, catalyst inside 90d.

B17(B) $34.18 new WATCH, above zone +2%, bezuclastinib PDUFA (GIST) 82d confirmed, material-agreement 8-K item 1.01 7d ago.

B12(B) $38.75 new WATCH, strong conviction (+3), in zone, EXTREME short crowding 15.1d, catalyst inside 90d.

B21(B) $56.02 new WATCH, insider bought $1,058,700, in zone, CARDIO-TTRansform readout H2 2026 (no near-30d date).

B19(B) $67.57 above zone by 1%, cash-flow positive, catalyst inside 90d, new Form 144 filed.

B15(B) $101.06 in zone, weak conviction, $27.8M insider selling/120d, Prader-Willi trial completes 2026-10-31.

B10(B) new WATCH, $86.93 idiosyncratic -16.2pp vs XBI, $418.5M insider selling/120d, EXTREME short crowding 16.0d, SKYLINE/SKYWAY readouts 21-22d out.

B18(B) new WATCH, $58.98 above zone by 4%, catalyst inside 90d, no veto.

A12(A) $7.54 above zone +19%, moderate conviction, cytisinicline NDA resubmission catalyst 22d.

A5(A) $5.73 catalyst 0 days (SOHO PV abstracts today, already priced per prior-session move), moderate conviction, no position into it.

A22(A) $38.78 idiosyncratic -13.3pp vs XBI, above zone +4%, EMBARQ CSU readout 22d, moderate conviction.

A18(A) $43.39 above zone +6%, HCM Phase 3 init guided 22d, moderate conviction.

L12(lottery) $4.34 idiosyncratic -67.6pp vs XBI, weak conviction, LUCIA Phase 3 readout 22d after LUGANO missed — this is the whole remaining thesis, no position.

B24(B) $41.62 cash-flow positive, above zone +15%, new Form 4, nothing actionable.

B4(B) $546.66 above zone +11%, moderate conviction, officer-departure soft flag, new Form 144.

B16(B) [redacted] cash-flow positive, PDUFA in 17 days (2026-09-26, confirmed): zilurgisertib for FOP, Priority Review — size for the outcome if held, no position recommended.

A11(A) $6.52 strong conviction (+3), insider bought $118,250, above zone +8%, conference catalyst 22d.

B13(B) $59.51 EXTREME short crowding 15.9d, above zone +7%, 8 new ownership filings today (routine), officer-departure soft flag.

L6(lottery) $8.73 idiosyncratic -18.9pp vs XBI, insider bought $949,381, readout both 21-83d out, runway only 4.4q.

A6(A) $13.95 above zone +21%, cash-only runway confirmed 16.69q, SLS009 AML readout 22d, no veto.

A20(A) $17.54 strong conviction, $103.6M insider cluster buy, HARMONI-3 readout 22d, PDUFA Nov 14 confirmed, above zone +15%.

A9(A) $22.73 strong conviction, insider bought $19,999,990, above zone +16%, A9 data 22d.

B7(B), B8(B) both NO_DATA, both merger-closed weeks ago — delete from watchlist.toml, repeat flag, unaddressed since 09-04.

13 further names one-lined by brief.py (L11, A25, A4, B5, L2, A2, B23, B6, A24, A13, A16, A14, A7) carry only soft flags or routine notes. 13 names (L5, L8, L3, A10, A21, B22, A15, L4, B9, B11, L10, B14, B2) have nothing to say.

Catalyst calendar (next 90 days)

New today: A23 Body Composition readout added (2026-10-01, expected, Q4 guided no month). New today: A24 lonvo-z BLA PDUFA 2027-03-10 (confirmed) — outside the 90-day window, first dated catalyst this name has carried. Both appended to catalysts.toml.

Dilution watch

• A17 +220.3% shares since 2023-05-11 (largest single-name jump among today's ACT names). B10 +143.9% since 2024-03-01. L9 +168.2% since 2023-11-09. L7 +105.3% since 2023-11-02, plus a new S-3 shelf (28d).
• B3 +33.9% since 2024-02-28 — modest relative to the ACT/SETUP cohort, dilution is not the reason for today's drop.
• A1 shares down 6.5% since 2024-02-14 — raise hasn't landed, still coming, on 0.85q runway.
• A3 shares down 40.7% since 2023-11-09 — buyback/reverse-split territory, not a dilution risk.
• B25 dilution technically nil (+4.7% since 2023); the $3B convertible only bites if price triples.
• L1 +0.9% since 2025-05-14 — the collapse here is a safety/litigation story, not a share-count story.

Risk notes

Exposure badly over-committed: [redacted] vs [redacted] ceiling (worse than yesterday's [redacted]) — A17, A23, B3, A3, L7, A19 all firing SETUP/ACT the same session is a concentration event, not six independent ideas. If taking more than one, scale every size by 0.59, and remember these fail together in a sector drawdown even where today's moves are idiosyncratic (A17, B3, A3, L7 all screen idiosyncratic). Regime uptrend (XBI +17.3pp vs SMA200), so the downtrend's strong-conviction-only ACT gate is not in force — A23, B3, A3 clear ACT on moderate conviction that would not clear in a downtrend. A17 and L7 are both unexplained multi-day slides — the biggest open questions on the desk; re-check both before any position, and do not treat "oversold" as a reason on its own. L1's litigation risk is a new, durable overhang — its RSI/oversold reading should not be read as a discount while the safety allegations are live. B7/B8 watchlist hygiene overdue, unaddressed since 09-04, delete both.