Viking Therapeutics

Refresh delta (since 2026-06-18)

What changed

1. Q2 printed a beat on a smaller loss, and — for the first time since the Phase 3 ramp — R&D fell sequentially. Net loss $128.0M / $(1.10) vs $(0.58) a year ago; R&D $115.8M, down 23% from Q1's $150.2M; G&A $16.8M. Consensus was a loss of roughly $(1.15)–$(1.23) depending on the outlet, so a beat of $0.05–$0.13.

2. The board authorized a new $500M buyback — effective July 30, 2026 — against $501.9M of total liquidity. It terminated the unused February-2025 $250M program and doubled it, for a three-year term. The authorization is now 99.6% of every dollar the company has.

3. Simultaneously it terminated the old ATM and signed a new one. Effective 2026-07-29 the 2021/2023 ATM agreement was voluntarily terminated and a New ATM Agreement signed with Stifel, Piper Sandler, Cantor Fitzgerald, Oppenheimer and Canaccord at a 3.0% commission — and the company has no effective shelf to sell off, the 2023 universal shelf having expired 2026-07-26. It also actually used the old ATM for the first time: $34.2M net in H1 2026, 4,591,231 cumulative shares, $40.4M gross left at quarter-end.

4. Three management/board changes inside seven weeks, and the commercial one is the tell. COO Marianne Mancini (in seat since Jan 2021) retires effective 2026-07-31, moving to a consulting agreement running to January 2, 2030 whose sole compensation is continued vesting of her existing options and RSUs. Hubert C. Chen, M.D. was appointed Chief Medical Officer on 2026-06-11, reporting to Lian, over clinical/medical/regulatory. And Dorothy Kelly-Gemmell joined the board as a Class I director effective 2026-08-01 — former Chief Commercial Officer and President of GoodRx, ex-Capsule, ex-AbleTo, career-started at Pharmacia.

5. VK3019 is in the clinic — Viking now has a second clinical asset. The DACRA amylin program the anchor dossier listed as preclinical, IND expected Q1 2026 filed its IND (announced April 2026, one quarter late) and initiated a Phase 1 SAD study in June 2026. It is now named alongside VK2735 in every capital-requirements and R&D-outlook sentence in the filing.

6. Lilly moved the efficacy bar while Viking was enrolling. Retatrutide (triple agonist) reported TRIUMPH-2 and TRIUMPH-3 topline on 2026-07-23: up to 22.6% weight reduction at 80 weeks (12mg; 9mg 21.6%; placebo 3.2%), BLA submission Q1 2027. Separately, Foundayo (oral orforglipron) went to $50/month on Medicare Part D from 2026-07-01.

7. Runway guidance was extended a quarter — on less cash. The Q1 filing said funded "through at least June 30, 2027"; the Q2 filing says "through at least September 30, 2027". On the call management went further and said the balance funds operations "into 2028" — a conflict with the filing that I am not resolving; the 10-Q language is the one that carries liability.

What held

The structural thesis stands, and mostly for the reasons the anchor gave. Single-asset concentration is unchanged — VK2735 is still the company, and VK3019's Phase 1 SAD is a lottery ticket on top of a lottery ticket, not a backstop. The CordenPharma lock is intact and slightly better specified: the prepayments are now disclosed as running 2026 to 2028 and fully creditable against future orders, against dedicated capacity of 100M autoinjectors + 100M vial/syringe + >1B oral tablets annually. The balance sheet is still clean — zero debt, $408.3M equity, conservative gilt-and-IG portfolio. The IP posture is unchanged and still favourable: the ITC win stands, Ascletis filed its opening Federal Circuit brief 2026-06-15, and the district-court action stays stayed. Zero SEC LR and zero AAER in the five years to 2026-08-10, re-verified. And the shape of the bet is unchanged: a binary 2027 VANQUISH topline into two giants with supply and pricing power.

Correction to the anchor dossier (not a delta — the anchor was wrong)

The 2026-06-18 dossier's bear case #1 and Lens 7 described Lilly's oral orforglipron as being at "Phase 3 / filing" with "global regulatory submissions in flight." That was already stale when it was written. Orforglipron was FDA-approved on 2026-04-01 as Foundayo™ and began shipping 2026-04-06 at $149/month self-pay — matching Novo's oral Wegovy, which beat it to market by about three months. Viking's own Q1 10-Q — on disk at the time — already listed "orforglipron (Foundayo™)" among therapies "already approved and marketed for obesity". The anchor understated the single most important competitive fact in the file. Flagging it rather than quietly fixing it, because it changes how much weight the anchor's Lens 12 deserves.

Latest print vs consensus, balance-sheet flags, market reaction (clinical overlay)

The print (Q2 2026, quarter ended 2026-06-30, filed 2026-07-29):

Vs consensus: $(1.10) actual against a Street estimate reported variously as $(1.15), $(1.21) and $(1.23). The outlets conflict and I am not picking a winner — call it a beat of five to thirteen cents, and note that for a company with no revenue an "EPS beat" is a statement about the timing of clinical accruals, not about the business. The signal in the print is the R&D line, not the EPS line.

What the R&D step-down actually means. The YoY bridge shows clinical-study spend +$67.5M offset by manufacturing −$10.3M and preclinical −$4.7M. Sequentially, Q1's $150.2M was the peak: both VANQUISH arms finished enrolling (VANQUISH-1 Nov 2025, VANQUISH-2 Q1 2026), so the enrolment-and-titration cost hump has passed and the trials are now in the cheaper maintenance-dosing phase of a 78-week protocol. This is the burn rolling over, and it is why the company could extend runway guidance on less cash. It reverses when oral Phase 3 starts in Q4 2026.

Balance-sheet flags:

Three flags worth naming, none of them fraud-shaped:

• Payables + accruals have risen $32.9M since year-end while cash fell $203.8M. That $32.9M of working-capital float is real cash the company has not yet paid out — H1 operating burn of $239.1M is flattered by it. Strip it and true H1 economic burn is nearer $272M. The float is a one-time gift; it does not repeat.
• Interest income halved YoY ($9.0M → $4.5M) as the portfolio shrank. As cash burns down, the offset to the loss burns down with it — a compounding drag the P&L will keep showing.
• Prepaid clinical costs jumped $10.9M in the quarter ($4.0M → $15.0M) — advance CRO payments, consistent with the oral Phase 3 gearing up. Cash out ahead of expense recognised.

Market reaction. No violent single-day move on the print. The stock has drifted down roughly 15% over the month into and through the release, a third consecutive weekly decline and an underperformance of the GLP-1 complex. Last ~$34.26 (2026-08-10), inside a 52-week range of $22.96–$43.15 — mid-range, not the lower third the anchor found in June. Implied market capMarket capitalisationThe share price multiplied by the number of shares. What the market says the equity is worth. ≈ $4.0B, essentially unchanged from the anchor's $3.97B: the stock has gone nowhere in eight weeks while burning $101M and watching a competitor print 22.6%.

Guidance: none in the conventional sense — a pre-revenue company gives timing, not numbers. our figures stays header-only, correctly. The only quantified forward statement is the liquidity horizon (through at least 2026-09-30 → 2027-09-30), and it conflicts with the call's "into 2028."

Call-sentiment shift vs prior calls **

The trend across the last four calls, as far as summaries support it:

• Q3 2025 / Q4 2025 — the "can we get it enrolled" era. Language organised around VANQUISH-1 enrolment completion and the CordenPharma supply lock.
• Q1 2026 (Apr) — "enrolment done, next catalyst named." Framed on VANQUISH-2 completion and the Q3 2026 maintenance readout.
• Q2 2026 (Jul 29) — the first call where the vocabulary is commercial rather than operational. Lian's framing is that oral VK2735 could be "the first orally available dual GLP-1/GIP agonist to reach the market" — note the careful hedge: first dual GLP-1/GIP oral, a category of one that exists only because two orals already beat him to the shelf. On the maintenance study he flagged that Viking hopes to break out safety data between the titration phase and the maintenance period, because reducing dose frequency at higher levels is what determines the tolerability profile.

The shift that matters: that last point is management pre-framing how the Q3 data should be read seven weeks before it lands. Pre-framing a readout's analytical lens is what sponsors do when they know the headline number alone will not carry the story. It is not a red flag — the titration/maintenance split is scientifically the right cut — but it is a tell that the endpoint is nuanced rather than clean.

Tone gap, updated. The anchor's observation was that management tone was confident while the tape's was the opposite, and "that gap is the trade." The gap narrowed from the wrong end: the tape stayed flat-to-down while management's language moved from science to commercialisation. Confidence expressed as commercial positioning, from a company with no product and a 2027 readout, is weaker evidence than confidence expressed as data.

New catalysts and ±5% movers **

Catalyst calendar, re-dated:

What moved the stock ±5% in the window — and what conspicuously did not.

The honest finding is that nothing Viking itself did moved the stock. The Q2 beat did not; the $500M buyback authorisation did not; VK3019 entering the clinic did not. What produced the drift was competitor read-through and category re-pricing:

• 2026-07-23 — retatrutide TRIUMPH-2/-3 at 22.6%/80wk. This is the single most consequential price input of the window and it belongs to Lilly. It raises the number VANQUISH-1 must print to be called differentiated, from tirzepatide's ~20% to a triple agonist's 22.6%, and it puts a Lilly BLA into roughly the same Q1-2027 window as Viking's topline.
• 2026-07-01 — Foundayo at $50/month on Medicare Part D. An approved oral GLP-1 at a $50 government-reimbursed price point is the reference price any third oral entrant inherits. This is the bear case's "commoditisation" arriving as a number, not a forecast.
• ~15% drawdown across July into August, three straight down weeks, underperforming GLP-1 peers.

Positioning: short interestShort interestHow many shares have been borrowed and sold by people betting the price falls. ~21% of float — still extraordinary, but down from the 22–28% the anchor found, so some of the squeeze fuel has drained. Sell-side is the mirror image: 17 of 18 Buy, consensus 12-month target $95.50 versus a $34.26 tape. The anchor called the $30-vs-$92 gap "the entire debate"; it is now $34 vs $95.50 — the gap did not close, it widened, which is the same standoff with eight weeks less time on the clock.

Management (new material)

Re-run because three seats changed. The founder-scientist-financier archetype is unchanged (Lian, 14 years, ~2.4% owned, Ligand 8.8%) — see the anchor for the full track record. What is new:

The bench got rebuilt around commercialisation, not science. A new CMO (Chen, June) is a normal late-stage strengthening. A departing COO who ran clinical operations since 2021, seven weeks before the maintenance readout and a year before topline, is not routine — though the terms argue against alarm: Mancini is described as retiring, and her consulting agreement runs to January 2030 covering "clinical project management and clinical operations for all of our drug development programs," compensated solely by continued vesting of existing equity. She keeps working, keeps the options, takes no cash. That is a retained-institutional-knowledge structure, not an exit. It is also, read coldly, a cheap one — the company retains its clinical-ops brain for zero incremental cash and zero new dilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company..

The board addition is the loudest signal in the filing. Dorothy Kelly-Gemmell's career is commercial access and cash-pay distribution — CCO/President of GoodRx, head of commercial at Capsule (digital pharmacy), CCO at AbleTo. You appoint that person when you intend to sell a drug directly to patients who pay cash, in exactly the channel where Foundayo is being priced at $149 self-pay and $50 Medicare. This cuts against the anchor's "built to be acquired" read: a board being assembled for a takeout does not need a GoodRx operator. Either they are genuinely planning a standalone launch, or they are building the optionality to credibly threaten one — which is also how you get a better takeout price.

Capital allocation — the item that moves my assessment. The anchor flagged the unused $250M buyback as "promotional" and unusual. The board doubled it. Effective 2026-07-30, a $500M, three-year repurchase authorisation. Set it against the same filing's own words: total liquidity $501.9M; "we will need to raise additional capital to fund our operations and complete our ongoing and planned clinical trials"; and a new ATM signed on the same day to do exactly that, at a 3.0% commission, off a shelf that does not yet exist because the 2023 one expired four days earlier.

Read those four facts together and the position is: we may buy back up to every dollar we have, while telling you we must sell shares to survive, and we have hired five banks to sell them. Nothing here is improper — an authorisation is not an obligation, the prior one went entirely unused across seventeen months, and the ATM/buyback pair is a legitimate way to signal a price floor to a 21%-short register. But capital-allocation signalling this loud, from a company whose entire remaining job is to fund a 78-week trial to readout, is managing the share register. The anchor's "promotional" read was right and the evidence for it strengthened.

Alignment check from the annual meeting (2026-05-19), which the anchor did not have. Quorum 69.1% of 115.89M shares. Say-on-pay passed but with 12.76M against on 30.05M for — a 29.8% against rate. Director S. Kathryn Rouan drew 21.25M withheld against 21.96M for — a 49.2% withhold; J. Matthew Singleton 33.7% withheld. In a plurality-vote election a near-coin-flip withhold elects the director anyway, but half the voting shares declining to support a sitting director is a governance signal the anchor's "no material red flags" did not capture. Also ratified: CBIZ CPAs P.C. as auditor for FY2026 — successor to Marcum's attest practice, so a firm-name change rather than an auditor switch, and not a flag on its own.

Forensic red flags (new material)

The accounting surface is still small and still clean. Revenue recognition remains n/a — $0 since 2012 inception. Re-checking the anchor's list against the new filing:

• Cash vs earnings. H1 net loss $286.3M vs H1 operating cash burnCash burnHow much more cash goes out than comes in, per period. The clock on a company with no profits. $239.1M. The $47.2M gap is SBC $21.5M plus the $32.9M payables/accruals build, less other items. Reconciles cleanly. But note the direction: the gap is now dominated by working-capital float rather than SBC, and float reverses.
• SBC. $21.5M H1, down from $24.9M in H1 2025 despite headcount growth — G&A SBC alone fell $3.8M. Modest against a $265.9M R&D line. Still honest; still no non-GAAP EPS to flatter.
• Dilution — upgraded from "structural overhang" to "in progress." The anchor said a raise was "when-not-if." It has started: $34.2M net drawn on the ATM in H1 2026, the first use disclosed; shares 114.79M → 116.65M since year-end; $40.4M gross left on the old programme, now terminated; a new ATM with five banks at 3.0% signed 2026-07-29; and a new universal shelf expected to be filed. The filing states plainly that equity issuance is "the most common type of fundraising for companies like ours" and that "the risk of dilution is particularly significant".
• The buyback/ATM contradiction. Covered in Lens 9. As a forensic matter it is a disclosure-consistent contradiction — both sides are fully disclosed in the same document, which is the opposite of concealment. It is a judgement flag, not an integrity flag.
• Internal controls. Disclosure controls concluded effective; the Q1 2026 accounting-system implementation is no longer flagged as under evaluation. No material weakness.
• Off-balance-sheet obligation to watch. The CordenPharma prepayments run 2026–2028 and are fully credited against future orders — so they are prepaid inventory economics, not a liability, but they are a committed cash outflow inside the runway window and the filing still names "the ability of CordenPharma to achieve its prepayment milestones" as a capital-requirement driver. The dollar total is not restated in the Q2 filing; the anchor's $150M over three years figure is and remains the only sourced number.

Regulatory findings (re-verified 2026-08-10): zero SEC Litigation Releases and zero AAERs naming Viking Therapeutics across 2021-08-10 → 2026-08-10, via EDGAR EFTS. Item 1 Legal Proceedings unchanged in substance: Viking remains the plaintiff and prevailing party — ITC affirmance of trade-secret misappropriation by the Ascletis defendants stands, Ascletis filed its opening Federal Circuit brief 2026-06-15, Viking cross-appeals a narrow portion, and the S.D. Cal. action stays stayed pending a final non-appealable ITC determination. No FDA/FTC/DOJ enforcement surfaced. Conclusion: no material regulatory or legal findings against the company as of 2026-08-10.

Three-year forward path + runway-to-catalyst (clinical overlay)

Model integrity — read this before any number below

our model was run and its outputs must not be cited. our model reports:

• "No computed values." 51 formulas, no cached results — every valuation cell is blank.
• "Opening balance sheet sourced: partial" — 7 of 10 lines; Receivables, Inventory and Net PP&E are absent because this company has none.
• Blank share price, blank WACC, blank balance check.
• Seed gaps the builder itself printed: only 2 quarters of history; gross margin seeded at 50%; operating margin at 25%; revenue growth at 8%; capex at 5% of revenue.

Those seeds are applied to a company with $0 of revenue. A 50% gross margin on zero revenue is not a conservative estimate, it is a category error — the workbook is a revenue-driven DCF pointed at a pre-revenue asset. The model is on disk at models/, it has no computed values, and nothing in this lens comes from it. Every figure below is with arithmetic shown.

(1) Does cash reach the catalysts?

• Liquidity 6/30/26: $501.9M, zero debt.
• Filing guidance: funded through at least 2027-09-30. Call guidance: "into 2028". Conflict surfaced, unresolved — the filing controls.
• Trailing burn: H1 operating cash $239.1M → $478M annualised. But Q2's $125.2M sits on a falling opex line, and $32.9M of H1's relief was working-capital float that will not repeat.
• Forward burn, base: Q3 ~$130M (trials in maintenance phase, VK3019 Phase 1 small), Q4 ~$145M (oral Phase 3 initiates) → H2 2026 ≈ $275M, leaving roughly $230M at 12/31/26. At a 2027 run-rate of ~$150M/quarter with oral Phase 3 in full flight, that is spent by late Q2 2027 — before the VANQUISH topline.

Verdict, sharper than the anchor's: cash comfortably reaches the Q3 2026 maintenance readout and the Q4 2026 oral Phase 3 start. It does not plausibly reach the 2027 VANQUISH topline without new capital, and the company has spent the last two weeks building the machinery to raise it (new ATM, five banks, new shelf pending). A raise before the defining readout is now not "when-not-if" — the plumbing is installed. The open question is only whether it comes as ATM dribble into a flat tape, a marketed follow-on into strength after the Q3 data, or a partnership upfront that trades economics for non-dilutive cash.

(2) Three-year loss-per-share path (all, arithmetic shown)

There is no EPS to project in the ordinary sense — there is a loss path, and its shape is the forecast.

Arithmetic for the base FY2026 line: H1 actual net loss $286.3M + H2 estimate $275M opex less ~$8M interest income = ~$553M; ÷ 116.5M weighted shares = $(4.75). That independently lands on the $(4.75) annual figure the sell-side is carrying, which is a reassuring cross-check on the burn assumption rather than a coincidence.

The reading that matters, and it inverts the usual logic: in the bear column the loss per share shrinks to $(1.30) — the best-looking EPS line in the table. It gets there by cancelling the drug. For VKTX a smaller forward loss is not a better outcome; it is the signature of failure. Any screen ranking this name on EPS improvement will rank the catastrophe first. The only forward number with information in it is cash-to-catalyst, and that is covered above.

Bull vs bear, updated **

Bull case — what survived and what got better. The burn has demonstrably peaked (R&D −22.9% QoQ), which converts the anchor's tightest worry — runway into a 2027 readout — from acute to merely serious, and let management extend the funded horizon a quarter on less cash. The pipeline stopped being a single molecule: VK3019 is in Phase 1, a second internally-owned mechanism in the hottest class in medicine, and amylin is precisely where Novo (CagriSema), Lilly, Zealand and AstraZeneca are all pointing. The manufacturing lock is intact and better-specified. And the board addition is a genuine tell: you hire GoodRx's former commercial president when you intend to reach cash-paying patients directly, which is the one channel where a well-tolerated oral dual-incretin could take share without out-spending Lilly's sales force. With 21% of the float short and the Street at $95.50 against a $34 tape, a clean Q3 maintenance readout showing durable weight maintenance on monthly SC dosing — a schedule no incumbent offers — is the kind of differentiated, non-cross-trial datapoint that re-rates a name like this violently.

Bear case — three ways this permanently impairs, updated:

1. The bar moved while Viking was enrolling. VANQUISH-1 was designed against tirzepatide. It will report into a world where retatrutide has printed 22.6% at 80 weeks and filed. Viking's Phase 2 SC gave 14.7% at 13 weeks; the bull requires the untapped slope to carry it well past 20% at 78 weeks. A 17–19% print — a clear clinical success — is now a competitive disappointment. The window between "the trial worked" and "the stock works" has narrowed to a few percentage points of weight loss.
2. Oral commoditisation stopped being a forecast and became a price. Foundayo: approved April 2026, shipping since 2026-04-06, $149/month self-pay, $50/month Medicare Part D since 2026-07-01. Novo's oral Wegovy beat it by three months. Viking's oral starts Phase 3 in Q4 2026 — years from a filing, into a category where the reference price is already set by a government formulary and where its own molecule is a peptide, structurally more expensive to make than Lilly's small molecule. The anchor called this bear point #1 and then mis-described its status; it is now the most concrete risk in the file.
3. The capital story is now a two-front problem. The company must raise before its defining readout, has installed a five-bank ATM to do it, has no effective shelf at the moment of writing, and its board chose that week to authorise buying back 99.6% of total liquidity. If the Q3 maintenance data is ambiguous, the raise happens into weakness at a price the 21% short base gets to set.

Pre-mortem (18 months out, the thesis broke). Q3 2026 maintenance data reads fine but nuanced — weight maintained on q2w, wobbly on monthly, tolerability differences between titration and maintenance that management pre-framed for a reason. No re-rate. The stock grinds to the mid-20s. Viking raises $400–500M via ATM and a follow-on into that tape through H1 2027, taking shares past 145M. VANQUISH-1 reports mid-2027 at ~17% — unambiguously approvable, unambiguously behind retatrutide, which filed in Q1. No acquirer bids, because every large pharma that wanted obesity now owns a program. The oral enters Phase 3 against two approved orals priced at $50 on Medicare. The drug works, the equity does not.

Is it too expensive? No earnings multiples exist and none are invented — n/a, not sourced. The framing is unchanged from the anchor and the arithmetic is now worse: ~$4.0B market cap against $501.9M of cash, i.e. roughly $3.5B of pipeline value for one Phase 3 asset with a raised competitive bar, one Phase 1 asset, and two shelved partnering candidates. That embeds a high probability of Phase 3 success and a takeout premium and a manageable dilution path. Fully valued, not cheap — and eight weeks of nothing happening while a competitor printed 22.6% did not make it cheaper.

Contrarian view — what the market may still be refusing to see. The anchor's contrarian read was that manufacturing plus the SC/oral dual shot is the real differentiator and the likely path is a takeout. I would sharpen it in a different direction: the market is pricing VK2735 as a fourth GLP-1 and ignoring that the Q3 catalyst is about dosing schedule, not weight loss. Every approved and near-approved competitor is daily-oral or weekly-injection. If monthly SC maintenance holds weight, Viking owns a compliance claim in a category whose real-world failure mode is discontinuation — roughly two-thirds of patients off therapy within a year. That is a commercial argument, not an efficacy one, which is exactly why a GoodRx operator just joined the board. The Street's $95.50 target is built on a beat on the efficacy axis; the more interesting outcome is a win on the adherence axis that nobody is modelling.

Devil's advocate **

Where the short seller is now stronger than in June:

• "Best-in-class efficacy" is a claim with an expiry date, and it expired on 23 July. The bull's number is a 13-week Phase 2 curve. The comparator is now an 80-week Phase 3 at 22.6% from a company that can fund ten more trials out of petty cash. Every month Viking spends in a 78-week protocol, the incumbent's label gets better.
• Watch what the board buys, not what it says. A pre-revenue company with $501.9M and an explicit "we will need to raise additional capital" does not authorise a $500M buyback because it has spare cash. It does it because the stock is 21% short, retail-held, and drifting — and an authorisation is free. The prior $250M authorisation was never used across seventeen months. The tell is not the buyback, it is that they doubled a facility they never touched, on the same day they hired five banks to sell stock. That is share-register management dressed as conviction.
• The COO left seven weeks before the catalyst. The consulting structure is benign-looking and the retirement framing is probably accurate. But the clinical-operations leader since 2021 stepping out immediately before a readout she ran, with a new CMO six weeks in seat, is exactly the personnel pattern a short deck builds a slide around — and it is unfalsifiable in either direction until the data prints.
• Half the voting shares declined to back a sitting director, and nearly 30% voted against pay. Small, but the anchor's clean-governance read is no longer quite clean.
• The interest-income treadmill. Other income halved YoY. As cash burns, the subsidy to the loss burns with it — each quarter of delay costs more than the burn line alone shows.
• Assumptions that must all hold for $34, let alone $95.50: (a) VANQUISH magnitude competitive with a triple agonist, not just tirzepatide; (b) a takeout bid from a shrinking pool; (c) a raise that does not happen into weakness; (d) an oral that reaches market before the $50-Medicare price becomes the ceiling; (e) monthly maintenance dosing that actually holds. Five conditions, and the anchor listed four.
• The single scenario that permanently impairs, restated. Not a VANQUISH efficacy miss — that is priced to a degree. It is a tolerability or SAE signal that emerges specifically in the maintenance phase, at the dose-frequency reductions the Q3 study is testing. Management has already told the market it wants to break titration and maintenance safety apart. For a third or fourth entrant into a category with cheap, approved, government-reimbursed incumbents, a safety asterisk is fatal to both the launch and the takeout simultaneously. Plausibility: low-to-moderate. Consequence: total.

Where the short seller got weaker: the burn peaked, the balance sheet is still debt-free, the pipeline widened into the clinic, and the runway guidance went up. A short thesis resting on "they run out of money" has less to work with than it did in June.

Lenses carried by reference

• Lens 1 · Company Overview — unchanged, see the previous dossier. (Amend one detail: HQ suite is now 9920 Pacific Heights Blvd Suite 500 per the July 8-KForm 8-KA filing for something that happened between reports and matters enough to tell shareholders now. cover, reflecting the completed office move; ROU assets $4.0M and a $3.9M long-term lease liability now sit on the balance sheet.)
• Lens 2 · Supply Chain — unchanged, see anchor. CordenPharma remains the sole disclosed chokepoint; prepayment window now stated as 2026–2028 and fully creditable against orders.
• Lens 3 · Competitive Advantages — unchanged in structure, see anchor, but read it against the Lens 12 correction: the "data/efficacy moat" is measured against a bar that moved on 2026-07-23.
• Lens 4 · Segments — n/a and confirmed n/a. The Q2 filing adds an explicit segment note: one reportable segment, CODM is the CEO, who allocates on loss-from-operations and its effect on cash. our figures stays empty by design.
• Lens 7 · Catalyst calendar + mechanism comps — catalyst half re-run in Lens 8; comp half unchanged, see anchor. Peer multiples remain n/a (no revenue, no earnings, on either side of the comp).
• Lens 14 · Management Questions — unchanged, see anchor. Three to add, all from this filing: (a) Why authorise $500M of buyback against $501.9M of liquidity on the same day you sign a new ATM with five banks? (b) What does the maintenance study need to show on monthly SC dosing specifically, and what would make you drop that arm? (c) Does Dorothy Kelly-Gemmell's appointment mean the base case is now a standalone cash-pay launch rather than a partnership?