Catalyst Pharmaceuticals, Inc. [CPRX]

The memo issues no Long, Short, Watchlist or Avoid verdict; it scores Criteria and outputs an analysis. The governing fact of this name is that it is not investable at all: Catalyst Pharmaceuticals was acquired by Angelini Pharma S.p.A. for $31.50 per share in cash, the merger closed on 2026-07-15, Form 25-NSE was filed the same day and Form 15-12G on 2026-07-24. The Tier-1 screen was run on 2026-07-28 against a frozen last-trade tick of $31.49 from 2026-07-14 and reported a +30.8pp valuation margin on a security whose equity had already been extinguished.

Key findings

• CPRX was acquired by Angelini Pharma S.p.A. for $31.50/share in cash; the merger CLOSED on 2026-07-15, Form 25-NSE was filed the same day and Form 15-12G on 2026-07-24. The Tier-1 screen ran on 2026-07-28 — thirteen days after close — against a frozen last-trade tick of $31.49 from 2026-07-14, and reported a +30.8pp valuation margin on an equity that no longer existed.
• ORGANIC vs ACQUIRED: 61.5% of the FY2022-FY2025 revenue growth was acquired or in-licensed. Of $374.8m added, FIRDAPSE (owned) contributed $144.4m (38.5%), FYCOMPA (acquired from Eisai, 2023-01-24) $113.3m (30.2%) and AGAMREE (in-licensed from Santhera, 2023-07-11) $117.1m (31.2%). Same-product FIRDAPSE CAGR is 18.8%, not 40.1%.
• The growth has already stopped: +85.9% (FY23) -> +23.5% -> +19.8% -> +5.6% (Q1-2026), with company FY2026 guidance of $615-645m = +4.4% to +9.5%, issued three months before the screen ran.
• EXCLUSIVITY — the brief's cliff hypothesis is confirmed for the acquired asset and refuted for the core one. FYCOMPA's cliff already happened: '571 patent expired 2025-05-23, '497 expired 2026-07-01, three generic tablets and one generic suspension are on market, revenue -61.3% YoY in Q1-2026, run-rating to ~$55m against $138m in its first year of ownership. FIRDAPSE by contrast has ~9 years of runway: orphan exclusivity lapsed 2025-11-28 but the patent estate held and all four ANDA filers settled (Teva 2035-02-25, Lupin same, Hetero January 2035, Inventia ~2037).
• MECHANISM: the FIRDAPSE upstream royalty expired in January 2026 at seven years from first commercial sale, cutting the blended rate from a maximum 18.5% to 6.0% — up to ~$49m/yr of cost removed, already visible as gross margin 87.3% -> 90.3%. It is a contract clock, not a commercial achievement, and it cannot repeat. Offsetting it, AGAMREE carries 7% + 7% escalating royalties above $100m of sales, so the mix shift is margin-dilutive at the gross line.
• ACCOUNTING QUALITY 1 — 'promoted products', a new and more flattering revenue denominator (FIRDAPSE + AGAMREE, i.e. the portfolio with the declining asset removed), appears for the first time in the 2026-02-25 release after fourteen consecutive quarterly releases at zero. Q1-2026 was headlined '28% YoY increase' on a quarter whose total revenue grew 5.6%. Surfaced generatively by the mention-frequency method.
• ACCOUNTING QUALITY 2 — reported net revenue was purchased with working capital. Catalyst renegotiated its distributor contract to pay REDUCED FEES (which lower the gross-to-net deduction and therefore RAISE reported net revenue) in exchange for slower settlement. DSO went 45.4 -> 40.9 -> 66.0 -> 76.2 -> 78.8 days; AR grew +83.6% YoY against revenue +5.6%; FY2025 operating cash flow FELL to $208.7m from $239.8m while net income rose. The revenue benefit is disclosed in words and NOT quantified by the company.
• The screen's +30.8pp margin is the product of two offsetting parameter errors of very different size: the terminal margin was capped from the actual 43.8% down to 22.9% (pushing required growth UP), and the exit multiple was set at 23.4x — above the 90th percentile of everything this stock ever traded at, and 87% above the 12.49x an informed strategic acquirer actually paid (pushing required growth DOWN, and dominating).
• The framework's own-history 12-month method validated exceptionally well: own-median EV/EBIT of 11.26x on guided FY2026 revenue gives $32.15, within 2.0% of the $31.50 a strategic buyer paid in cash. It also reframes the deal — Angelini's '21% premium' was a premium to a price sitting at the 41st percentile of the name's own multiple history, i.e. a 2% DISCOUNT to its own median fair value.
• Catalyst has never discovered a drug. All three products are licensed or acquired and FY2025 R&D was 2.2% of revenue. In 2025 it reviewed over 100 acquisition targets and completed none, while cash rose from $517.6m to $755.9m. That is the named cause of the downside case, and it is also why the Quality Criteria scores INDETERMINATE rather than PASS.

What this name should change in the process

• A screen must verify the security still trades: (a) latest-trade timestamp within N sessions of the screen date, (b) absence of Form 25 / Form 15 in the recent filing index. Two API calls; either alone would have rejected this record. Highest-value control this memo produced.
• data_quality_ok:false must GATE ranking, not annotate it. The screen set the flag on CPRX and ranked it at +30.8pp anyway.
• Trailing volatility must not drive sizing across a structural break: 32.1% trailing vs 1.2% post-announcement is a 27x error in the sizing input, erring toward MORE size.
• PROPOSED RULE: where revenue growth spans a business combination or in-licensing, the demonstrated CAGR must be recomputed on a same-product basis or declared UNIDENTIFIED. Product-level disaggregation is a required disclosure and made this computable directly from the company's own audited tables.
• The exit multiple must be sanity-checked against the name's OWN history. 23.4x sat above this name's all-time p90 of 19.38x; anchoring above a name's own p90 requires an argument and there was none. Largest single contributor to the spurious margin.
• Any quote or trade used for sizing or execution must carry a staleness assertion. Alpaca still returns a tight bid 31.48 / ask 31.51 in 200x200 size on this dead ticker, timestamped ten sessions ago.

Sections

Disclosed limitations

• Consensus estimates not obtained — Alpha Vantage 25/day quota exhausted after one call. Consensus Criteria is INDETERMINATE and blocks nothing. Company guidance used as the near-term base instead.
• Earnings-call transcripts: only 2024Q1 retrieved before the quota cut off. The mention-frequency series is built on 15 quarterly earnings press releases (8-K Ex-99.1), is labelled as such throughout, normalised per 1,000 words, and carries no prepared-remarks/Q&A split. The transcript metric itself is INDETERMINATE.
• Peer Spread not computed. HRMY named as the correct comparator; no figure asserted.
• The revenue benefit of the distributor fee renegotiation is not quantified by the company and is not estimated here.
• The AGAMREE Patent Term Extension is requested, not granted; the 2040 patent horizon is not assumed.
• The FY2025 10-K states the FIRDAPSE orphan-exclusivity expiry as both 2025-11-26 and 2025-11-28 in different sections. Both are reported; the discrepancy is the company's and is not resolved.
• No options chain was pulled — there is no listed underlying. No structure is proposed.
• No Excel model, no forward DCF and no chart pack were produced. Every series is tabulated in full in the Financial Model Notes instead.
• Own-multiple history uses Alpaca IEX daily closes (SIP historical bars returned HTTP 403 on this entitlement). SIP latest-trade and snapshot endpoints corroborate the final price to the cent.
• The no-deal counterfactual target holds the TTM operating margin (44.8%) flat through FY2026. That is an assumption, stated at the point of use, not a forecast.
• Angelini Pharma is private. No post-close information on this asset is or will be publicly available; the name is permanently uncoverable after 2026-07-24.