Praxis Precision Medicines, Inc. (NASDAQ: PRAX) — Four Shots on Goal, Priced as if Three Are Blockbusters and None Can Miss

An independent equity-research note. Educational and informational; not investment advice.

⚡ Claude’s Take

This block is the author’s own subjective opinion. It is general information, not investment advice. Everything below it (the analytical body) is deliberately position-free and carries no recommendation and no price target — the single opinion in this piece is fenced inside this block.

Verdict: AVOID-HERE / HOLD; not a short. Low-to-medium conviction on direction, high conviction that the margin of safety is gone. Praxis has done something genuinely hard — it ran a positive Phase 3 in essential tremor (a CNS graveyard where every prior branded drug failed), got two NDAs accepted with 2027 launch dates, and re-rated roughly 200-fold off its March-2023 near-death low ($0.80) to a $357 high in May 2026. That is a real accomplishment and it is largely why the stock is near $325 with a ~$9.3B market cap / ~$7.9B enterprise value. But at this price the market is paying the full base-case sum-of-the-parts (~$309/share on my math) for two drugs that have not yet launched a single unit, while crediting almost nothing for the two things that actually happened this quarter: vormatrigine — the largest-TAM asset — missed its Phase 3 primary endpoint on June 1 and POWER2 was paused, and the relutrigine PDUFA slipped three months on a “major amendment.” You are underwriting near-flawless commercialization of a symptomatic, elderly, generic-defended tremor drug (company touts “>$10B peak”; every recent CNS launch comp says $1–3B is the realistic zone) with essentially no discount for the commercialization risk that has defeated every predecessor.

Framing: a de-risked-at-the-top, over-hyped-in-the-tail, low-float idiosyncratic-momentum name — NOT a value or mean-reversion setup, and NOT (yet) a falling knife. The factor tape confirms it: +547% trailing-twelve-month return (annualized), near-zero market beta, R² under 1% — virtually all of the variance is stock-specific binary-catalyst risk, coiled on discrete FDA dates. Directional zone (my view, not a target): I’d want a give-back toward the 200-day area (~$260) and ideally the low-$200s / high-$100s — roughly the rNPV midpoint between my bear (~$122) and base (~$309) — before the risk/reward is balanced; genuine value (paying for the bear, optioning the bull) emerges nearer $120–160. It is not a short: two probable 2027 approvals, a ~$1.45B net-cash fortress with a realistic runway into ~2029–2030 (management’s “into 2028” is conservative), ~15%-of-float short interest that is squeeze-fuel on any positive print, dedicated biotech-crossover anchors (Janus Henderson, Baker Bros.), and a credible strategic-takeout floor make shorting a wall of binary catalysts a good way to be right on thesis and dead on the tape. Two things keep me cautious rather than constructive even at a fair price: not a single insider has bought a share on the open market at any point in the 200x run (they gift and tax-withhold at the peak; a director cashed out under a 10b5-1), and a cluster of late-June wobbles — the relutrigine “major amendment,” a new auditor (EY→KPMG) announced the very same day as that PDUFA slip, and the POWER1 miss — is exactly the kind of thing that reads as noise in a bull tape and as tells in hindsight. Capital allocation, to be fair, is a genuine strength: they raised ~$1.76B into the rally at escalating prices ($42→$260), creating per-share value. Tag: “They cured the near-death; now the market is paying for the cure four times over.” Conviction: low-to-medium. Flips bullish on a clean ulixacaltamide launch that tracks above the XCOPRI/cenobamate benchmark with real Medicare payer coverage — proof the ET market monetizes. Flips more bearish on a relutrigine CRL/narrow label at the December PDUFA, a soft ET launch, or an equity raise that confirms the burn is outrunning the story.

📈 Stock Price Action — Five-Year Event Map

Over the trailing five years PRAX has traced one of the most violent round-trips in biotech: from an IPO near $19–28 (Oct 2020), down through a 2022 collapse to $1.56 and a five-year low of $0.80 on 30-Mar-2023, then a staggering ~200x recovery to a five-year high of $357.13 on 27-May-2026, before easing to ~$325 (10-Jul-2026) — about 9% off the high, with a 52-week range of roughly $37.70 → $357.13. The arc is a near-death pipeline reset (the PRAX-114 depression failures), a genetics-driven rebuild around essential tremor and rare epilepsy, and a momentum melt-up on two positive Phase 3 topline reads and two NDA acceptances. (Price levels: FACT, AZI five-year CSV, unadjusted closes. Attributed drivers: INTERPRETATION, cross-referenced to earnings prints, 8-Ks, and the news feed.)

Cycle narrative. (1) The post-IPO class of 2020 de-rated as early-pipeline risk met a rolling-over biotech tape. (2) 2022 was existential: the PRAX-114 depression-program failures gutted the story and a rate-driven bear market took the stock to $0.80 — a genuine near-death. (3) From that base, a genetics-anchored rebuild around essential tremor (ulixacaltamide) and monogenic epilepsy (relutrigine) drove a ~26x recovery through 2024. (4) In Feb-2025, an Independent Data Monitoring Committee recommended discontinuation for futility on the ulixacaltamide program — the stock halved to ~$33. (5) Praxis pressed on; open-label RADIANT (vormatrigine) and continued DEE data stabilized the tape. (6) The turn was the October-2025 positive Essential3 topline — the first-ever Phase 3 win in ET — which tripled the stock into year-end and unlocked the NDA. (7) 2026 opened with two NDA acceptances, PDUFA dates, a large follow-on raise, and the elsunersen signal, carrying PRAX to a $357 high. (8) Then the largest-TAM asset stumbled: vormatrigine’s POWER1 missed its primary on June 1 and POWER2 was paused, cutting the stock 27% before elsunersen’s Breakthrough designation and relutrigine headlines rebuilt it to ~$325. The tape is now a coiled, catalyst-driven momentum structure sitting just below an all-time high.

1. Executive Summary

Praxis Precision Medicines is a Boston-based, clinical-to-commercial-transition central-nervous-system (CNS) biopharmaceutical company built on a single organizing idea: that many epilepsies, movement disorders, and related CNS conditions share a common root — an imbalance between neuronal excitation and inhibition, often traceable to specific ion-channel genetics — and that precision small molecules (its Cerebrum platform) and antisense oligonucleotides (its Solidus platform) can correct it. The company has no product revenue, ~168 employees, and ~$1.45B of net cash, and it is on the cusp of becoming a commercial company: it carries four late-stage assets and management promotes a headline “>$20 billion aggregate peak-sales” potential across them.

The investable reality is more nuanced, and this memo’s central tension is the gap between that promotional aggregate and a defensible bottom-up estimate. Ulixacaltamide (a selective T-type calcium-channel blocker for essential tremor, ET) is the lead value driver: its Phase 3 Essential3 program hit (Study 1, N=473, placebo-adjusted mADL11 delta ~2.6 points, p<0.0001; Study 2 randomized-withdrawal, 55% vs 33% maintained response), the NDA was accepted on 14-Apr-2026 with no advisory committee, and the PDUFA date is 29-Jan-2027. ET is a genuine first-mover opportunity — propranolol is the only FDA-approved ET drug, ~50 years old — but it is also a symptomatic, elderly, generic-defended market that has been a commercial and clinical graveyard (Sage’s SAGE-324 failed on tolerability). Management’s “>$10B” ET peak is not credible against recent CNS launches (XCOPRI reached only ~$400M annualized five years post-launch; Epidiolex ~$972M across multiple orphan indications); a realistic peak zone is ~$1–3B.

Relutrigine (a persistent-sodium-current blocker for SCN2A/SCN8A developmental and epileptic encephalopathies, DEE) has an NDA under Priority Review with a PDUFA that slipped from 27-Sep to 27-Dec-2026 on a “major amendment” — an orphan asset with high per-patient pricing but a tiny population and an approval package resting on a 16-patient randomized cohort. Vormatrigine (a Nav-preferring sodium modulator for broad focal/generalized epilepsy), the largest-TAM asset, missed its POWER1 Phase 3 primary endpoint on 1-Jun-2026 and POWER2 was paused — a material blow that a spectacular open-label Phase 2 (RADIANT) had masked. Elsunersen (an ASO for SCN2A early-onset DEE) has the cleanest efficacy signal (77% sham-adjusted seizure reduction) and Breakthrough Therapy designation, but the narrowest population.

At ~$325, the equity trades at roughly its base-case risk-adjusted NPV (~$309/share) — crediting clean approvals of ulixacaltamide and relutrigine plus base commercial ramps, with only modest vormatrigine option value and no margin of safety for the ET-commercialization risk. Versus the reference class, PRAX’s ~$7.9B EV is ~3x what Axsome (~$3B) and Neurocrine (~$3–4B) commanded at their launch inflections — before selling a unit. There is no operating moat; the only durable advantages are time-boxed patents and regulatory exclusivity (the ulixacaltamide core composition-of-matter, in-licensed from Purdue, is reportedly near-term, defended by Praxis’s own 2040–2044 formulation/method patents and ~2032 NCE exclusivity). The setup is a wide-tailed binary: a well-funded, genuinely de-risked lead asset wrapped in a promotional aggregate, on a low-float, 15%-short, +547%/year momentum structure. This body takes no position and sets no price target; the discussion below is embedded-expectations and scenario analysis only.

2. Business Overview

Praxis Precision Medicines was founded in 2015 and completed its IPO in October 2020. Its mission is to translate the genetics of epilepsy and related CNS disorders into targeted therapies for conditions defined by an excitation–inhibition (E/I) imbalance in neuronal signaling. The scientific thesis is that ion-channel dysfunction — in sodium channels (Nav1.2/SCN2A, Nav1.6/SCN8A), T-type calcium channels (Cav3), and others — produces a spectrum of hyperexcitability disorders (epilepsies, essential tremor, certain pain and psychiatric conditions), and that correcting the channel activity at its source can be broadly therapeutic. Two proprietary platforms operationalize this: Cerebrum, a small-molecule discovery engine, and Solidus, an antisense-oligonucleotide (ASO) platform.

The company makes no money today. Reported “revenue” has been trivial and non-commercial — $8.6M in FY2024 and $2.4M in FY2023 from collaboration and grant arrangements, and effectively zero in FY2025. Value is entirely prospective and resides in a four-asset late-stage pipeline plus a preclinical genetics portfolio:

• Ulixacaltamide (PRAX-944) — oral, selective T-type calcium-channel blocker for essential tremor. Positive Phase 3 Essential3 program; NDA accepted 14-Apr-2026; PDUFA 29-Jan-2027. The lead commercial asset and ~55% of the risk-adjusted pipeline value.
• Relutrigine (PRAX-562) — oral persistent-sodium-current blocker for SCN2A/SCN8A DEE (severe, early-onset pediatric epilepsies). NDA under Priority Review; PDUFA 27-Dec-2026 (extended from 27-Sep on a major amendment); Orphan and Rare-Pediatric-Disease designations (priority-review-voucher eligible). A broad-DEE registrational study (~200,000-patient U.S. population per management) reads out topline Q4-2026 to support a supplemental NDA.
• Vormatrigine (PRAX-628) — oral Nav-preferring sodium modulator for focal and generalized epilepsy (plus pain optionality). POWER1 Phase 3 missed its primary endpoint (1-Jun-2026); POWER2 paused. Program under “reassessment.”
• Elsunersen (PRAX-222) — intrathecal ASO lowering Nav1.2/SCN2A for SCN2A early-onset DEE. Breakthrough Therapy (22-Jun-2026); registrational EMBRAVE3 topline expected 2027.

Behind these sit preclinical genetic-epilepsy programs (KCNT1/PRAX-020, SYNGAP1, PCDH19/PRAX-080) and collaborations with Ionis, RogCon, Purdue Neuroscience, the Florey Institute, Tenacia Biotechnology (Greater China rights), and — as of 7-Jul-2026 — Remagine Labs (a transdermal ulixacaltamide formulation).

The four assets are not equally weighted, and understanding why is central to the investment case. On a risk-adjusted basis, ulixacaltamide alone carries roughly half the pipeline’s value, relutrigine another quarter, and vormatrigine plus elsunersen and the platform the remainder — so the equity is far more concentrated than the “four late-stage assets” framing implies. Mechanistically the portfolio is a coherent bet on ion-channel modulation: ulixacaltamide dampens the T-type calcium “burst-firing” that drives the tremor circuit; relutrigine and vormatrigine both act on voltage-gated sodium channels (the persistent sodium current and Nav1.6, respectively) to quiet the hyperexcitability underlying seizures; and elsunersen uses an antisense oligonucleotide to lower production of the over-active Nav1.2 channel in gain-of-function SCN2A patients. The shared biology is real and is the intellectual foundation of the platform — but it also means a common failure mode: if sodium-channel modulation proves harder to translate from small open-label studies into large controlled trials than management asserts (as POWER1 just showed for vormatrigine), the read-through risk across relutrigine and elsunersen is not independent. The business model is classic single-company biopharma: burn equity capital to advance a pipeline through the FDA, then attempt to build a U.S. specialty commercial franchise (neurology / movement-disorder specialists and epileptologists) and monetize via branded pricing before patents and exclusivity lapse. There is no recurring revenue, no installed base, and no commercial track record yet — this is a pre-launch, binary, catalyst-driven business. Verdict: a scientifically credible, genetics-first CNS pipeline company whose entire value is forward-looking and whose transition from “clinical-stage” to “commercial” is about to be tested for the first time.

3. Industry Dynamics

Praxis operates across three structurally distinct CNS markets, and conflating them is the single biggest source of the valuation debate.

(1) Broad epilepsy (vormatrigine’s target). The global anti-seizure-medicine (ASM) market is roughly $10.8B (2024), growing only ~4%/year to an estimated ~$15.5B by 2033 — a mature, generic-anchored, slow-growth market. The workhorses (levetiracetam, lamotrigine, gabapentinoids, valproate) are cheap generics; ~3.5M Americans live with epilepsy and roughly a third are drug-resistant. Value has migrated to differentiated adjunctive agents, but even the best of them monetize slowly: SK Biopharm’s cenobamate (XCOPRI) — a genuinely differentiated ASM — reached only ~$400M annualized about five years post-launch. This is a structurally poor market to enter: crowded, price-compressed, and hard to differentiate, as POWER1’s miss just illustrated.

(2) Rare monogenic DEE (relutrigine, elsunersen). Developmental and epileptic encephalopathies driven by specific mutations (SCN2A, SCN8A, KCNT1, Dravet/SCN1A, CDKL5) are small populations — the SCN2A/SCN8A DEE population is on the order of ~5,000 U.S. patients — but carry orphan economics: seven-year U.S. orphan exclusivity, breakthrough/priority pathways, rare-pediatric-disease priority-review vouchers, and very high per-patient pricing (Marinus/Immedica’s Ztalmy for CDKL5 is ~$105–133K/year). This is a structurally attractive niche on a per-patient basis — but the absolute dollar pools are inherently capped by patient counts, and diagnosis requires genetic testing that gates the addressable population.

(3) Essential tremor (ulixacaltamide) — the outlier. ET is large (≈6.4M U.S. adults meet criteria, ~1.1M carry a documented diagnosis) but sits in the least favorable structural bucket for premium monetization: it is a symptomatic, chronic, largely elderly (Medicare) condition defended by pennies-a-day generics (propranolol is the only FDA-approved agent; primidone and topiramate off-label). The regulatory opportunity is real — no purpose-built ET oral has ever been approved — but the payer and adherence reality is punishing: any branded oral must justify a large premium over generics on a modest symptomatic benefit, in a step-therapy environment, against device/procedure competition (Cala kIQ neuromodulation, MR-guided focused ultrasound, deep-brain stimulation) at the severe end.

Regulatory environment. The FDA is visibly accommodating precision-genetic pediatric epilepsy (Breakthrough designations for relutrigine and elsunersen, priority review, a “simplified/accelerated” ASO path; EMA PRIME) — this lowers approval risk on the DEE assets but not commercial risk. Note the priority-review-voucher program is currently authorized only through 30-Sep-2029 absent reauthorization, a modest policy risk to the relutrigine PRV thesis.

Verdict: structurally mixed, tilting negative for the biggest claimed dollars. The orphan-DEE niche is attractive but dollar-capped; broad epilepsy is a structurally poor, mature market; and ET — where the “>$10B” claim concentrates — is a large but generic-defended, elderly, symptomatic market that is the least favorable of the three for the premium monetization the valuation requires. Marathon’s capital-cycle lens is not the primary tool here (this is a pre-supply, single-company story), but the competitive-entry lens is: the DEE and ASO spaces are actively crowding (UCB, Jazz, SK Bio, Lundbeck’s ~$2.6B Longboard/bexicaserin acquisition, Stoke/Biogen’s zorevunersen for Dravet, Ionis, Encoded/Ultragenyx), which will pressure the very orphan pricing that makes the niche attractive.

4. Competitive Position

There is no operating moat, because there is no operating business yet. For a pre-revenue biopharma, “competitive advantage” reduces to three time-boxed intangibles: patents, regulatory exclusivity, and platform/target precision. Assessed in Greenwald’s taxonomy, Praxis has a weak intangibles position with no reinforcing scale, network, switching-cost, or cost advantage.

Patents — the critical, under-appreciated weakness. The economic life of the lead asset is the whole ballgame, and the picture is mixed. The FY2025 10-K describes an ulixacaltamide patent estate whose issued essential-tremor family expires in 2040, with titration-method (2041), analog-compound (2040), adjunctive-use (2043), and dosage-form (2044) families layered on. However, ulixacaltamide was in-licensed from Purdue Neuroscience, and at least one sell-side analyst (H.C. Wainwright) has flagged that the core composition-of-matter protection is materially nearer-term (~2029) — with the later-dated 2040–2044 families being the weaker, more design-around-able method/formulation claims that generics routinely challenge. Reconciling these is an open question (see the relevant section), but the shape of the risk is clear: for a drug launching in 2027, the durable, hardest-to-challenge protection may be unusually short, backstopped by five-year new-chemical-entity (NCE) exclusivity to ~2032 and secondary patents thereafter. Relutrigine’s composition-of-matter runs to 2039 and elsunersen’s to 2041; the DEE assets also carry seven-year orphan exclusivity from approval.

First-mover in ET — a lead, not a moat. Ulixacaltamide would be effectively the first purpose-built oral for essential tremor, and no competitor has an ET-labeled oral in Phase 3 behind it. That is genuinely valuable while it lasts — but first-mover status in pharma is a head-start, not a durable barrier, and the potentially short core-patent window means the runway to entrench brand and formulary position before generic or branded entrants arrive is narrow.

Platform optionality — real science, not a moat. Cerebrum (small molecule) and Solidus (ASO), plus the genetic-target portfolio and academic/industry collaborations (Ionis, RogCon, Florey, Remagine), give real discovery optionality. But a discovery platform lowers discovery risk only — it does nothing to stop UCB, SK Bio, Stoke, or Ionis from competing in the same genotypes, and the ASO-for-DEE field is already contested. What deteriorates without the intangibles? Everything: strip the exclusivity clocks and there is no scale (a tiny, unbuilt commercial org), no switching costs (prescribers freely substitute ASMs and generics), no network effect, and no cost advantage. Verdict: not a durable competitive advantage — a time-boxed regulatory/first-mover head-start that must be monetized quickly and widely, precisely the hardest thing to do in the generic-defended ET market where the biggest dollars are claimed.

5. Growth History and Forward Opportunities

History. There is no commercial growth history — revenue has been immaterial collaboration/grant income. The relevant “history” is pipeline maturation and its violent translation into equity value: a near-death reset in 2022 (the PRAX-114 depression failures), followed by a genetics-anchored rebuild that produced positive Phase 3 topline in essential tremor (October 2025), positive DEE data (EMBOLD), and two accepted NDAs. In parallel, the company has begun building its first commercial organization — SG&A/G&A roughly doubled year-over-year to ~$28M in Q1-2026 as it hired commercial leadership, a field-force plan, market-access, and manufacturing/inventory for two launches.

Forward opportunities, ranked by defensibility. (1) Ulixacaltamide ET launch (2027) — the largest and best-de-risked opportunity, but with the widest range of commercial outcomes ($1–3B realistic vs. management’s >$10B); the entire thesis is a levered bet on the first-ever commercially successful ET drug. (2) Relutrigine SCN2A/8A DEE launch (2026/27) plus broad-DEE expansion — a smaller but higher-margin orphan annuity; the broad-DEE registrational readout (Q4-2026) is the key expansion catalyst management says could grow the opportunity “several fold.” (3) Elsunersen — the cleanest signal but the smallest population; option value, not a needle-mover on a ~$9B EV. (4) Vormatrigine — was the largest TAM and is now the most impaired; salvageability hinges on a POWER2 redesign around the 30mg dose. (5) Platform/preclinical (KCNT1, SYNGAP1, PCDH19) — long-dated optionality.

Quality of growth. This is low-quality, binary, event-driven “growth” — not organic compounding. Consensus models a step from ~$4M revenue in 2026 to ~$320M in 2027 on the two launches, but that ramp is entirely contingent on (i) two approvals landing on schedule, (ii) real-world payer coverage and uptake in a Medicare-heavy, generic-defended indication, and (iii) first-time commercial execution with no track record — a setup where sell-side ramps are routinely too optimistic. Verdict: genuinely de-risked at the top (ulixacaltamide approval is probable), overhyped in the middle and tail (vormatrigine impaired; DEE dollars structurally small; the >$20B aggregate is promotional). High-variance growth of low intrinsic quality until a launch actually proves the market monetizes.

6. Financial Quality

For a pre-revenue biotech, “financial quality” is not about margins or returns on capital — there are none — but about the shape and sustainability of the burn, the strength of the balance sheet, and the honesty of the reported picture. On all three, Praxis screens better than the typical clinical-stage name.

The P&L is a widening, deliberate loss. With no product revenue since inception, the income statement is pure investment: R&D of $155M (FY22) → $87M (FY23 trough, after the PRAX-114 discontinuation and pipeline narrowing) → $152M (FY24) → $267M (FY25), re-accelerating ~3x as four assets pushed into pivotal/Phase 3 work; and G&A of ~$56–60M running steady, now inflecting for launch — Q1-2026 G&A of $27.9M roughly doubled year-over-year as the company hired commercial leadership, a field force, market-access, and medical-affairs teams for two potential 2026–27 launches. FY2025 net loss was –$303M; the Q1-2026 quarterly opex run-rate is ~$106M ($78M R&D + $28M G&A). Stock-based compensation of $33.9M (FY25) is a modest ~10% of operating expense and ~14% of cash burn — non-alarming by biotech standards, and down from $41.4M in FY24.

Burn is real, cash-backed, and accelerating. Operating cash burn ran –$111M (FY23) → –$132M (FY24) → –$249M (FY25), and Q1-2026 alone was –$86M (nearly double Q1-2025’s –$53M). The cash-flow-to-net-income ratio (~0.82 in FY25) confirms the loss is overwhelmingly cash, not non-cash write-offs — clean, if expensive. Annualized burn is now ~$345M and rising into the commercial build. There is no revenue offset, and there won’t be a material one until the 2027 launches; consensus models only ~$4M of 2026 revenue stepping to ~$320M in 2027, entirely contingent on approvals and uptake.

The balance sheet is a genuine fortress — the single clearest positive. At 31-Mar-2026 the company held $1.448B in cash, equivalents and marketable securities ($786.9M current + $660.9M long-term), against only ~$1.4M of lease liabilities — net cash ≈ $1.45B, or ~$52/share. Total equity of $1.41B is essentially the cash pile, which is why book value per share (~$50.6) and the resulting P/B (~6.7x, 79th percentile of its own history) are analytically meaningless here: they are a cash multiple, not a business-value signal. Management guides to sufficiency “into 2028,” but the arithmetic is more generous: $1.45B ÷ ~$345M burn ≈ 4.2 years (into ~2030); even stressing burn to $450–500M/year for a full two-drug launch leaves runway into ~2029, before ~$50–60M/year of interest income on the pile. The practical implication is important and differentiating: Praxis does not need to raise money to reach its key PDUFAs and first launches — the financing overhang that usually plagues pre-commercial biotech is, for now, low.

Verdict. The “do economics improve with scale?” test is not applicable pre-revenue; the correct framing is that this is a cash-incineration-until-launch story with an unusually strong, honestly-reported balance sheet. The relevant question — whether $1.45B funds the company to self-sustaining commercial cash flow — is plausibly yes for the first launches, but two approvals plus full commercial infrastructure for essential tremor and epilepsy is expensive, and a clean path to breakeven remains unproven and dependent on payer uptake. High-quality funding; unproven economics.

7. Capital Allocation

Capital allocation is where management has, so far, most clearly earned its keep — and it is a real, differentiating strength in a sector littered with value-destructive dilution.

Use of proceeds and financing discipline. Every dollar has gone into R&D and the pre-commercial build; there has been no M&A, no buyback, and no dividend — all correct for the stage. The story is the timing of the raises. Praxis sold escalating dollar amounts of equity into a rising stock: a January-2024 offering at roughly $42.50/share (~$161.6M net), a March-2024 ATM (~$113.1M net), an October-2025 underwritten offering (~$567M net, closed 20-Oct-2025), and a January-2026 follow-on of 2,212,000 shares at $260.00 (~$540.6M net, priced 6-Jan-2026) — cumulatively ~$1.76B raised from 2024 through January 2026 as the stock ran roughly 10x. Because capital came in far above prior book value, per-share value was created, not destroyed: book value per share climbed from low single digits to ~$50. A team that raised $540M at $260 the same week the stock printed near an all-time high did shareholders a favor. (Note a ~1-for-16 reverse split in late 2023 makes the raw “3M → 28M share” progression look worse than the economic reality; post-split dilution is ~9x, but funded at a steadily rising cost of equity.)

The royalty stack is the one real drag on future economics. Praxis in-licensed key programs and owes meaningful economics out: on the SCN2A/elsunersen franchise, Ionis is owed royalties in the low-20s percent of worldwide net product sales plus development and sublicense milestones, and RogCon a profit-share in the mid-teens percent of net profits — a heavy combined load on any eventual elsunersen gross profit. (Offsetting, Praxis receives milestones/royalties on out-licensed assets via a UCB collaboration — up to $98.5M plus royalties — and a BCPE/Bain collaboration — up to $264M plus royalties.) These inbound/outbound flows matter for the DEE and ASO economics but not for the ulixacaltamide ET franchise, which Praxis owns more cleanly (subject to the Purdue in-license terms).

Incentives. The 2026 proxy shows a reasonable structure — base, performance bonus on pre-set corporate objectives, and equity — with a new PSU tranche vesting on clinical and regulatory milestone attainment (approvals, pipeline readouts), which ties a slice of executive equity directly to the events that drive the thesis rather than to share price alone. The headline “$55.5M CEO Compensation Actually Paid” in 2025 is a mark-to-market artifact of the ~200% stock run, not cash. The peer group was moved up-cap after the run; say-on-pay is annual.

Verdict: management has allocated capital intelligently. Opportunistic, value-accretive raising into strength; no debt; no reckless M&A or premature buybacks; modest SBC; incentives increasingly tied to value-driving milestones. The caveats are the Ionis/RogCon royalty stack on the SCN2A franchise and the absence of any insider open-market buying — management is happy to issue and monetize equity but has not put personal cash in at these prices (see the relevant section). On the core question, though, capital stewardship is a clear positive.

8. Changes and Headwinds — Last Two Years

The last two years contain both the entire bull case and, in the last six weeks, the seeds of the bear case.

The transformational positives (2024 → mid-2026): (1) the October-2025 positive Essential3 Phase 3 topline in essential tremor — the first-ever Phase 3 win in ET — which tripled the stock and unlocked the lead NDA; (2) two NDA acceptances in April 2026 — relutrigine (priority review, 6-Apr) and ulixacaltamide (14-Apr, no advisory committee) — converting the company from clinical-stage to pre-commercial; (3) positive elsunersen EMBRAVE Part A data (77% sham-adjusted seizure reduction, p=0.015) and a Breakthrough Therapy designation (22-Jun-2026); and (4) the ~$1.76B of opportunistic capital raised into the run, over-capitalizing the balance sheet ahead of launch.

The fresh headwinds (last six weeks) — the diligence priorities:

• Vormatrigine POWER1 miss (1-Jun-2026). The largest-TAM asset failed its Phase 3 primary endpoint; POWER2 was paused for redesign. Critically, the eye-catching numbers that had built the vormatrigine story came from the open-label Phase 2 RADIANT (56% median reduction, ~22% seizure-free); the placebo-controlled POWER1 did not replicate them — a textbook open-label-to-RCT regression that should also discount the other small, open-label datasets in the pipeline (relutrigine’s 16-patient cohort, elsunersen’s 9-patient EMBRAVE Part A).
• Relutrigine PDUFA slip (29-Jun-2026). The date moved from 27-Sep to 27-Dec-2026 because Praxis itself submitted “additional sensitivity analyses of existing clinical data,” which the FDA classified as a major amendment (automatic 3-month extension). Management frames it benignly (no new studies, no safety/CMC concerns), but a sponsor voluntarily bolstering its efficacy analyses mid-review suggests the FDA raised questions about the robustness of a pivotal signal resting on a very small population. It is not a Complete Response, but it is not nothing.
• Auditor change, EY → KPMG (announced 29-Jun-2026, filed 2-Jul). The 8-K states no disagreements and no reportable events, and prior opinions were clean — but a pre-revenue biotech switching auditors mid-cycle on the very same day it disclosed the PDUFA slip is an optics flag worth noting, even if benign on its face.

Verdict: the two-year arc strengthens the thesis on net (two NDAs, a fortress balance sheet, a genuine ET first), but the last six weeks weaken it at the margin and concentrate the near-term risk — the epilepsy leg is impaired, the first potential approval is delayed and under efficacy scrutiny, and the audit/insider optics argue for caution rather than chasing.

9. Risk Analysis (Risk Matrix)

The dominant risks are commercial (ET uptake) and regulatory (relutrigine label), amplified by a low-float momentum structure that magnifies every binary. The balance sheet mutes the financing risk that would otherwise dominate.

10. Valuation Discussion (Embedded Expectations)

No price target; no recommendation. This section reverse-engineers what the current price implies and stress-tests it with scenario analysis.

With zero revenue, ~$7.9B enterprise value (market cap ~$9.3B less ~$1.45B net cash) is 100% a risk-adjusted call option on the pipeline. Two framings bound the debate.

What the market is paying for. A de-risked, launched CNS specialty franchise a few years from peak typically trades around 2.5–3.5x peak sales on an EV basis (Neurocrine and Axsome mature near 3–4x; orphan assets higher). Applying ~3x to the ~$7.9B EV implies the market is underwriting roughly $2.5–3B of risk-adjusted peak sales. That is only ~13% of management’s promotional “>$20B unrisked aggregate,” so the tape is emphatically not buying the $20B story — but it is ~55% of a skeptical, bottom-up unrisked base (ulixacaltamide $2.0B + relutrigine $0.8B + vormatrigine $1.5B + elsunersen $0.4B ≈ $4.7B). A 55% blended realization requires near-certain approvals and full base-case commercial ramps — including in essential tremor, where no branded drug has ever succeeded commercially.

Scenario sum-of-the-parts (risk-adjusted NPV per share). Valuing each launched asset at a peak-sales multiple capturing ramp, a ~10–11% discount rate, and patent/exclusivity life, then multiplying by approval probability and adding net cash (~$52/share):

The punchline: at ~$325 the stock sits almost exactly on the base case (~$309). The market is fully crediting successful approvals of ulixacaltamide and relutrigine plus base commercial ramps, with only modest vormatrigine option value and no margin of safety for the bear (an ET commercial disappointment or a pivotal setback is a >60% drawdown to ~$122). The bull (~$734) aligns with the surviving high sell-side targets (BTIG $810, RBC $719). Critically, ulixacaltamide is ~55% of base pipeline value — the entire thesis is a levered bet on the first-ever commercially successful essential-tremor drug; if realistic ET peak is $1B rather than $2B, the base falls to ~$237/share.

Versus comps at the same stage, the price is rich. Axsome was worth ~$3B at Auvelity’s approval and Neurocrine ~$3–4B at INGREZZA’s launch — both with a single, clearly-de-risked, first-in-class asset in hand. PRAX at ~$7.9B EV is ~3x those valuations, pre-approval on all four assets, with its broadest asset freshly missed. The DEE-asset M&A comp (Lundbeck’s ~$2.6B Longboard/bexicaserin) brackets relutrigine’s standalone value. On any lens, the price has pulled forward multiple years of near-flawless execution.

A note on the sensitivity that dominates everything. Because ulixacaltamide is ~55% of base pipeline value and its outcome distribution is unusually wide, the whole valuation is hostage to one number: realistic ET peak sales. The company’s “>$10B” and even the mid-range sell-side ($5–10B, Guggenheim) figures imply the stock is deeply undervalued; a skeptic’s $1–1.5B (in line with what XCOPRI and Epidiolex actually achieved) implies it is expensive. The gap is not a modeling nuance — it is the difference between a symptomatic oral behaving like a specialty blockbuster and behaving like every prior branded ET or adjunctive-CNS drug. No amount of DCF precision resolves it; only launch data will. That is why the embedded-expectations framing matters more than any single point estimate: at ~$325 the market has chosen the optimistic side of an unresolved, binary commercial question, and is paying for it in full.

Own-history valuation context. The only clean multiple available — price-to-book at ~6.7x, the 79th percentile of the stock’s own history — is uninformative because book value is ~95% cash; there is no P/E (no earnings) and no P/S (no sales). This is a reminder that conventional value screens are blind here: the entire question is the risk-adjusted NPV of a pipeline, and the stock is “cheap” or “expensive” only relative to one’s assumptions about approval and commercial success, not relative to any reported financial anchor. Embedded-expectations conclusion: the market is underwriting base-case success at a base-case price — a demanding, no-margin-of-safety setup, not a mispricing in either direction.

11. Variant Perception

Consensus / bull narrative (why price targets sit $550–810). Two imminent, low-clinical-risk launches (relutrigine PDUFA 27-Dec-2026; ulixacaltamide 29-Jan-2027); ulixacaltamide as the only Phase-3 asset in a wide-open ET market after Sage’s SAGE-324 and Jazz’s suvecaltamide both failed; orphan pricing on relutrigine; and epilepsy/platform optionality — framed as “the next Neurocrine/Axsome CNS commercial franchise.” The blended sell-side target sits well above the current price.

Strongest bull case. If ulixacaltamide launches into a genuine unmet-need vacuum (propranolol is the only approved drug, poorly tolerated, and 25–55% of patients don’t respond) and captures even a mid-single-digit-billion franchise, plus relutrigine becomes a durable orphan annuity and the broad-DEE expansion (Q4-2026 readout) hits, the sum-of-the-parts clears $700+ — and the “no ET drug has ever been approved” fact flips from a bear point into a first-mover moat. The fortress balance sheet, crossover-fund anchoring (Janus Henderson, Baker Bros.), 15%-short squeeze fuel, and takeover optionality all support the upside skew.

Strongest bear case. (1) ET commercial flop — the graveyard is total; approval ≠ uptake against $4 generics in an elderly, step-therapy market; realistic peak is $1–3B, not $10B. (2) Relutrigine amendment/label risk — approvable ≠ a clean label on a 16-patient package under fresh efficacy scrutiny. (3) Vormatrigine already missed — the large-epilepsy leg is impaired, and the RADIANT→POWER1 regression discredits the other open-label datasets. (4) Zero insider conviction — no open-market buying in a 200x run; monetization and gifting at the peak. (5) Low-float momentum unwind — +547%/year on a 27.9M-share float reverses violently on any binary miss, as June’s −27% showed.

The load-bearing assumptions and what falsifies each:

Factor-positioning read (from the momentum workstream). PRAX is a pure idiosyncratic, binary-catalyst momentum name: +547% trailing-twelve-month return (annualized), 1-year Sharpe 2.73 but a five-year max drawdown of −96.5% (it round-tripped catastrophe), near-zero market beta (0.13–0.33 across factor models) and R² under 1% — virtually all variance is stock-specific catalyst risk, not market or style-factor exposure. This is evidence for where consensus may be offsides: a crowded, un-hedged, event-driven vehicle coiled on discrete FDA dates, precisely the profile that gaps 30–50% on a print in either direction. The tape is not “trending on fundamentals”; it is a sequence of binary bets, and the price already embeds the base case clearing them.

12. Fact vs. Interpretation

13. Open Questions

1. Ulixacaltamide net price, gross-to-net, and Medicare payer coverage — the entire ET thesis (and the >$2.5B models) hinges here, and there is no precedent for a branded ET oral. What step-therapy will PBMs impose against generic propranolol/primidone?
2. The true ulixacaltamide patent life — reconcile the 10-K’s 2040–2044 families against the analyst-flagged ~2029 in-licensed core composition-of-matter and ~2032 NCE exclusivity. How defensible is the franchise at generic entry?
3. What specifically did the FDA question to prompt the relutrigine “major amendment” sensitivity analyses, and does it presage a narrow label or CRL on 27-Dec-2026?
4. The EY→KPMG auditor change — benign process, or is there anything behind a switch announced the same day as the PDUFA slip?
5. Go-forward vormatrigine design post-POWER1 — is the program salvageable at 30mg, or effectively dead?
6. POWER1 disclosure gaps — the exact N, p-value, and placebo-adjusted effect on the missed primary, needed to judge salvageability.
7. Real-world durability of the ulixacaltamide effect — the ~2.6-point mADL benefit is modest and on a subjective scale, with dizziness/cognitive/euphoria TEAEs in an elderly cohort; what is real-world adherence?

14. What Must Be True

Bull case — what must be true, and its falsification test. Ulixacaltamide must both approve (probable) and commercialize into a genuine multi-billion franchise — i.e., neurologists must adopt a branded oral over pennies-a-day generics, payers must reimburse it without prohibitive step-edits, and elderly patients must tolerate and stay on it. Relutrigine must approve on a workable label and the broad-DEE expansion must hit, turning it into a durable orphan annuity. Falsification test: the first four quarters of ulixacaltamide TRx must track above the XCOPRI/cenobamate launch curve with documented Medicare coverage. If the launch tracks below the cenobamate benchmark or payers impose hard step-therapy, the >$2.5B models — and the base-case valuation the stock already reflects — are wrong, and the thesis breaks toward the ~$122 bear.

Bear case — what must be true, and its falsification test. The bear requires that the ET market not monetize at scale (approval without commercial success, like every prior ET drug), and/or that the small/open-label datasets underpinning relutrigine and elsunersen regress in larger controlled trials as vormatrigine’s did, and/or that the relutrigine label disappoints — collapsing the pipeline toward net cash (~$52/share). Falsification test: a clean ulixacaltamide launch that clears a mid-single-digit-billion trajectory and a relutrigine approval on a broad, workable label would refute the bear, validate the platform’s genetics-first approach, and justify the base-to-bull zone. The single most informative near-term data point for both cases is the 27-Dec-2026 relutrigine PDUFA outcome, followed by the early-2027 ulixacaltamide launch metrics.

15. Source Appendix

Primary and secondary sources are consolidated in the Source Appendix below. Key primary sources: Praxis FY2025 Form 10-K (filed 19-Feb-2026); Q1-2026 Form 10-Q (filed 7-May-2026); DEF 14A proxy (30-Apr-2026); the 8-K corpus (relutrigine/ulixacaltamide NDA acceptances 6-Apr / 14-Apr-2026, vormatrigine POWER1 update 1-Jun-2026, elsunersen Breakthrough Therapy 22-Jun-2026, relutrigine PDUFA extension 29-Jun-2026, auditor change 2-Jul-2026); the 424B5 prospectus (6-Jan-2026); Form 4/144 and Schedule 13G corpus; and the Q1-2026 earnings-call transcript (9-May-2026). Quantitative data from EDGAR XBRL (CIK 1689548), ROIC.ai, the AZI price feed and valuation-percentile ranks, and the FactorsToday factor model. Company clinical claims are treated as hypotheses and cross-checked against FDA designations, ClinicalTrials.gov, peer-reviewed epidemiology, and trade press (NeurologyLive, NeurologyAdvisor, Endpoints, Fierce, GlobeNewswire press releases). All URLs and access dates (2026-07-11) are listed in Appendix B.

APPENDIX A — Standard Diligence Questionnaire

Praxis Precision Medicines, Inc. (NASDAQ: PRAX) — as of 2026-07-11

Supplemental to the research memo. Answers grounded in the underlying research; Fact / Interpretation / Assumption labels applied where material.

General

What thoughtful questions have other investors asked about this company? The sell-side Q1-2026 call (Piper, TD Cowen, Raymond James, Guggenheim, Needham, Jefferies, BTIG, Deutsche Bank, OpCo, LifeSci) clustered on: (1) POWER1 expectations and how to read placebo-adjusted seizure reduction (asked before the June miss — instructive in hindsight); (2) what supports relutrigine working in the broad DEE population beyond the tiny SCN2A/8A cohort; (3) ulixacaltamide real-world compliance/retention given first-few-weeks tolerability; (4) manufacturing/supply readiness for two simultaneous launches; and (5) the ex-US strategy (management is deliberately US-first, citing pricing policy). The unasked-but-critical question is payer coverage and gross-to-net for a branded ET oral against generics.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Not applicable — there are no earnings; the company has never generated product revenue and posts widening losses (FY2025 net loss –$303M). Driven by external environment or internal actions? Internal: the loss reflects deliberate R&D and pre-commercial investment, not a cycle. How stable are revenues? No revenues; “revenue” to date is immaterial collaboration/grant income. Outlook for products/services? Two potential launches in 2026–27 (relutrigine, ulixacaltamide); consensus models ~$4M (2026) → ~$320M (2027), entirely approval- and uptake-contingent. How big will this market be? Three markets: broad epilepsy (~$10.8B, +4%/yr, mature); rare DEE (small populations, orphan pricing, growing); essential tremor (~6.4M prevalent / ~1.1M diagnosed, generic-defended). Growing but with very different monetization profiles; largely domestic near-term.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More — the DEE and ASO spaces are actively crowding (UCB, Jazz, SK Bio, Lundbeck/bexicaserin, Stoke/Biogen, Ionis, Encoded). How profitable is the business (ROIC, ROE)? Deeply negative — this is a cash-burning pre-revenue biotech; ROIC/ROE are meaningless. How profitable is the industry? Bimodal — commoditized generic ASMs earn little; differentiated orphan/branded CNS assets earn high margins (90%+ gross) if they achieve scale, which few do quickly. Can the business be easily understood? The thesis (genetics-first ion-channel modulation) is understandable; the value depends on hard-to-forecast clinical, regulatory, and commercial binaries. Undermined by foreign low-cost labor? No — value is IP and regulatory exclusivity. Do brands matter? Somewhat — prescriber trust and formulary position matter in specialty neurology, but the core protection is patents/exclusivity, not brand. Nature of competition? Clinical differentiation, regulatory exclusivity, and payer access. Switching costs? Low — prescribers freely substitute ASMs and generics; no lock-in. [Interpretation] No durable moat — a time-boxed regulatory/first-mover head-start.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The pipeline itself — ~$7B+ of the equity’s value is off-balance-sheet optionality (in-process R&D is expensed, not capitalized). Off-balance-sheet liabilities? Royalty/milestone obligations owed out (Ionis low-20s% net-sales royalty, RogCon mid-teens% profit-share on the SCN2A franchise; milestone commitments across collaborations) — real future economic drags not on the balance sheet. Operating leases are minimal (~$1.4M). How conservative is the accounting? Straightforward for a biotech — losses are cash-backed (CF-to-NI ~0.82), SBC modest (~10% of opex). The one optics item is the EY→KPMG auditor change (29-Jun-2026), disclosed with “no disagreements.” How CapEx-hungry? Minimal (–$56K FY2025) — this is an R&D/clinical-trial spend model, not a capital-asset model.

Capital Allocation & Management

How much FCF does the business generate? Negative — FCF was –$249M (FY2025) and worsening; ~$345M annualized burn. How does management use FCF / what is the philosophy? N/A (no FCF); capital is raised via equity and deployed into R&D + commercial build. Philosophy has been opportunistic, value-accretive equity issuance into strength. Significant acquisitions recently? None (only small collaboration equity stakes; the “deals” are in-/out-licenses). Buying back shares? No (correct for the stage). Issuing large amounts of new shares to insiders? Routine annual grants; SBC modest. Broadly, ~$1.76B raised from public/institutional investors 2024→Jan-2026 at escalating prices ($42→$260/share), roughly tripling the share count but creating per-share book value. Compensation policy? Base + performance bonus + equity, with a 2026 addition of PSUs vesting on clinical/regulatory milestones (good alignment); 2025 CEO “Compensation Actually Paid” of $55.5M is a mark-to-market artifact of the run. Motivations of management? Mixed signal — incentives increasingly tied to value-driving milestones, but zero insider open-market buying in a 200x run, with tax-withholding and gifting at the peak and a director’s ~$3.6M 10b5-1 sale.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No — ordinary NASDAQ common stock (US filer). Dividend policy? None; none expected — capital is reinvested. How profitable is the business? Unprofitable by design (pre-revenue). Is net income diverging from cash from operations? No material divergence — losses are overwhelmingly cash (CF-to-NI ~0.82); the small gap is SBC and working-capital timing, not accrual games. Valuation anchors: no P/E (no earnings), no P/S (no sales); P/B ~6.7x is a near-meaningless cash multiple (book ≈ net cash ~$52/share). EV ~$7.9B is 100% risk-adjusted pipeline NPV; current price (~$325) sits on the base-case rNPV (~$309/share).

Risks & Downside

What factors would cause the stock to decline? A soft ulixacaltamide launch (the dominant risk); a relutrigine CRL or narrow label at the 27-Dec-2026 PDUFA; further vormatrigine failure; a read-through where other small/open-label datasets regress in controlled trials; a dilutive raise after a miss; or a low-float momentum unwind (the June POWER1 –27% is the template). Risk of catastrophic loss? Real but not base-case — simultaneous ET commercial failure + relutrigine CRL + vormatrigine death would strip the pipeline toward net cash (~$52/share), a >80% drawdown; the 2022 collapse to $0.80 proves the tail exists. Chance of a total loss? Low near-term — the ~$1.45B net-cash fortress (runway into ~2029–2030) makes insolvency implausible over the forecast horizon; a total loss would require sustained multi-asset failure and cash exhaustion.

Recent News & Events

Has the business environment changed recently? Yes, materially and in both directions. Positives (2025–26): the first-ever positive Phase 3 in essential tremor (Oct-2025), two NDA acceptances (Apr-2026), elsunersen Breakthrough Therapy (Jun-2026), and ~$1.76B raised. Negatives (last six weeks): vormatrigine POWER1 Phase-3 miss + POWER2 pause (1-Jun-2026); relutrigine PDUFA slip to 27-Dec-2026 on a “major amendment” (29-Jun-2026); and an EY→KPMG auditor change (2-Jul-2026). Significant acquisitions? None. Change in accounting policies? None disclosed beyond the auditor change (stated as no disagreements). Recent changes — new markets, facilities, management? Building a first-ever US commercial organization (new COO Megan Sniecinski, 2026; commercial leadership and field-force hiring; dual-source drug-substance manufacturing for ulixacaltamide); Remagine collaboration (transdermal ulixacaltamide, 7-Jul-2026); Greater-China rights licensed to Tenacia.

APPENDIX B — Source Appendix

Praxis Precision Medicines, Inc. (NASDAQ: PRAX) — as of 2026-07-11

All non-obvious facts in the memo trace to the sources below. Company clinical/commercial claims are treated as hypotheses and cross-checked against filings, FDA actions, epidemiology, and independent data. Accessed 2026-07-11 unless noted.

Primary — SEC filings (EDGAR, CIK 0001689548)

• Form 10-K (FY2025), filed 2026-02-19 — business/pipeline, patent estate (ET family to 2040; titration 2041; analog 2040; adjunctive 2043; dosage form 2044; relutrigine comp-of-matter 2039; elsunersen 2041), Essential3 design (Study 1 N=473; Study 2 N=238), risk factors, license terms (Ionis, RogCon, Purdue, UCB, BCPE/Bain, Tenacia), PRV program authorized through 2029-09-30. https://www.sec.gov/Archives/edgar/data/1689548/000168954826000029/prax-20251231.htm
• Form 10-Q (Q1-2026), filed 2026-05-07 — balance sheet ($1.448B cash/investments; ~$1.4M leases; equity $1,409.96M; 27.86M shares), “no revenue from product sales since inception,” runway “into 2028.” https://www.sec.gov/Archives/edgar/data/1689548/000168954826000047/prax-20260331.htm
• DEF 14A proxy, filed 2026-04-30 — compensation structure, 2026 milestone-vesting PSUs, CEO “Compensation Actually Paid” $55.5M (2025), NEOs, up-cap peer group.
• 424B5 prospectus, filed 2026-01-06 — January 2026 follow-on: 2,212,000 shares at $260.00 ($575.1M gross / ~$540.6M net; +331,800 over-allotment). https://www.sec.gov/Archives/edgar/data/1689548/000110465926001823/tm261686-2_424b5.htm
• 8-K, 2026-04-06 — relutrigine NDA accepted (Priority Review, PDUFA 2026-09-27); elsunersen EMBRAVE Part A positive (77% placebo-adjusted reduction, p=0.015).
• 8-K, 2026-04-14 — ulixacaltamide NDA accepted, PDUFA 2027-01-29, no advisory committee.
• 8-K / PR, 2026-06-01 — vormatrigine POWER1 missed primary endpoint; 50%-responder secondary met; POWER2 paused. https://www.globenewswire.com/news-release/2026/06/01/3304714/0/en/praxis-precision-medicines-provides-vormatrigine-program-update.html
• 8-K / PR, 2026-06-22 — elsunersen Breakthrough Therapy Designation (SCN2A-DEE). https://www.globenewswire.com/news-release/2026/06/22/3315249/0/en/
• 8-K / PR, 2026-06-29 — relutrigine PDUFA extended to 2026-12-27 (major amendment; company-submitted sensitivity analyses; no new studies/safety/CMC). https://www.globenewswire.com/news-release/2026/06/29/3319308/0/en/
• 8-K, 2026-07-02 — auditor change, Ernst & Young → KPMG (effective 2026-06-29; “no disagreements / no reportable events”).
• 8-K, 2025-10-20 — October-2025 underwritten offering (~$567M net). https://www.sec.gov/Archives/edgar/data/1689548/000110465925100488/tm2528989d1_8k.htm
• Form 4 / Form 144 corpus (2025–2026) — no open-market purchases (code P) by any insider; director-grant clusters (2026-06-10/12); Jan-2026 employee grants + F-code tax-withholding at ~$288–293; director Jill DeSimone Form 144 (2026-06-24, ~$3.6M / 11,600 sh, 10b5-1); CEO Souza gift of 23,121 sh (code G, 2026-07-08).
• Schedule 13G corpus (2026) — Janus Henderson 12.3%, Baker Bros. 6.3%, FMR/Fidelity 5.4%, Vanguard Capital Management 5.09%, Orbis 5.0%.
• Q1-2026 earnings-call transcript, 2026-05-09 (via ROIC.ai) — management framing of two launches, “>$20B aggregate / >$10B ET peak” claims, POWER1/EMBOLD/EMBRAVE commentary, runway “into 2028,” Q1 opex ~$106M / burn $86M.

Primary — clinical / regulatory

• ClinicalTrials.gov: Essential3 (ulixacaltamide ET), EMBOLD/EMERALD (relutrigine DEE), POWER1/POWER2 (vormatrigine focal epilepsy), RADIANT (vormatrigine Ph2), EMBRAVE (elsunersen SCN2A).
• FDA designations: ulixacaltamide Breakthrough Therapy (Dec-2025); relutrigine Orphan Drug + Rare Pediatric Disease (SCN2A/8A DEE, Dravet); elsunersen Breakthrough Therapy + Rare Pediatric Disease, EMA PRIME.

Secondary — company clinical claims (cross-check)

• Praxis Q1-2026 corporate update, GlobeNewswire 2026-05-07. https://www.globenewswire.com/news-release/2026/05/07/3289917/0/en/
• EMBOLD (relutrigine SCN2A/8A) positive results, GlobeNewswire 2025-12-04. https://www.globenewswire.com/news-release/2025/12/04/3200330/0/en/
• EMBRAVE Part A (elsunersen) results, GlobeNewswire 2026-04-06. https://www.globenewswire.com/news-release/2026/04/06/3268372/0/en/
• Ulixacaltamide NDA/PDUFA, NeurologyAdvisor. https://www.neurologyadvisor.com/news/ulixacaltamide-nda-accepted-for-essential-tremor-pdufa-date-set/
• Ulixacaltamide mechanism/data summary, Wikipedia. https://en.wikipedia.org/wiki/Ulixacaltamide
• Vormatrigine RADIANT Phase 2, NeurologyLive. https://www.neurologylive.com/view/praxis-vormatrigine-significant-ability-reduce-seizure-incidence-phase-2-radiant-study
• POWER1 update, StockTitan. https://www.stocktitan.net/news/PRAX/praxis-precision-medicines-provides-vormatrigine-program-3k6c2t22gthx.html
• IP / patent commentary (H.C. Wainwright; ~2029 core composition-of-matter flag), Investing.com. https://m.investing.com/news/analyst-ratings/hc-wainwright-reiterates-praxis-precision-stock-rating-on-patent-news-93CH-4709518

Secondary — industry, epidemiology, comps

• Essential-tremor prevalence (Tremor Journal / PMC). https://pmc.ncbi.nlm.nih.gov/articles/PMC4137360/
• ET treatment landscape (propranolol only approved; non-response rates), US Pharmacist. https://www.uspharmacist.com/article/essential-tremor-a-common-disorder-with-limited-treatments
• SAGE-324 essential-tremor failure (Medscape / ClinicalTrials NCT04305275).
• Ztalmy (ganaxolone) orphan pricing, Pharmaphorum. https://pharmaphorum.com/news/fda-clears-marinus-ztalmy-as-first-drug-for-rare-epilepsy
• Lundbeck / Longboard (bexicaserin) ~$2.6B acquisition, Pharmaphorum. https://pharmaphorum.com/news/lundbeck-drops-epilepsy-drug-26bn-longboard-bid
• Stoke Therapeutics zorevunersen (Dravet) timelines, Stoke IR. https://investor.stoketherapeutics.com/news-releases/
• Epilepsy-drug market size, Straits Research / Mordor Intelligence.
• XCOPRI/cenobamate revenue trajectory (SK Biopharm disclosures); Epidiolex FY2024 revenue (Jazz / Fierce Pharma).
• Axsome (AXSM) valuation at Auvelity approval, Motley Fool. https://www.fool.com/investing/2026/07/02/is-axsome-therapeutics-stock-a-millionaire-maker-h/
• Neurocrine INGREZZA launch economics, SEC 8-K (2017). https://www.sec.gov/Archives/edgar/data/0000914475/000119312517247407/d409947dex991.htm
• Analyst price targets / rating actions (RBC, Needham, TD Cowen, BTIG), Investing.com / MarketBeat.
• Short interest (~15.1% of float, ~6.8 days to cover), MarketBeat. https://www.marketbeat.com/stocks/NASDAQ/PRAX/short-interest/

Quantitative data feeds

• SEC EDGAR XBRL — multi-year P&L, cash flow, share count (authoritative).
• Public financial databases — income statement, balance sheet, cash flow, enterprise value (cross-checked and reconciled to filings).
• Public price history — 5-year daily price series.
• Public factor model — factor loadings (near-zero market beta), risk-adjusted return history (trailing-12-month return well above +500% annualized; 5-year max drawdown near −96%).